A multi-layer vibrating screening apparatus and a control system thereof

Abstract

The present application relates to the technical field of vibration screening equipment control, in particular to a multi-layer vibration screening equipment and a control system thereof; the control system of the multi-layer vibration screening equipment collects acceleration data of each vibration disc through a vibration sensor data acquisition module and performs signal conditioning optimization. Frequency domain analysis is performed through an improved Welch algorithm to accurately identify the working conditions of the vibration layer and improve the screening efficiency. A 12-class working condition classifier is constructed in combination with a deep belief network, the system can accurately identify faults and automatically adjust the working state of the equipment to ensure the screening accuracy. An adaptive control decision module is adopted to adjust the vibration frequency, amplitude and inclination in real time, adapt to different working conditions, reduce manual intervention and improve the screening efficiency; the model predictive control is used to automatically adjust the parameters to ensure the safe operation of the equipment and further improve the stability and reliability of the equipment.

Description

Technical Field

[0001] This invention relates to the field of vibration screening equipment control technology, specifically to a multi-layer vibration screening equipment and its control system. Background Technology

[0002] Vibrating screening equipment is widely used in industries such as mining, metallurgy, building materials, and chemicals for material screening and grading. With the increasing automation and intelligence of industry, traditional vibrating screening equipment has gradually revealed some limitations, such as the inability to diagnose equipment faults in real time, low screening efficiency, and difficulty in adjusting operating conditions. Traditional vibrating screening equipment mainly relies on manual intervention and cannot dynamically adjust equipment parameters when facing various working conditions, thus affecting the stability and screening effect. Furthermore, existing equipment often lacks precise fault detection and intelligent control methods, leading to untimely detection of equipment faults and increased maintenance costs and downtime.

[0003] Traditional equipment typically cannot automatically adjust vibration frequency, amplitude, and tilt angle according to different operating conditions. When operating conditions change, manual intervention is required to adjust the equipment, increasing labor costs. Furthermore, in some complex operating conditions, manual adjustment is insufficient to meet the requirements of efficient screening.

[0004] To address the aforementioned issues, it is necessary to propose a multi-layer vibration screening device and its control system. Summary of the Invention

[0005] The purpose of this invention is to solve the problems existing in the background art, and to propose a multi-layer vibration screening device and its control system.

[0006] The objective of this invention can be achieved through the following technical solutions:

[0007] A multi-layer vibration screening device and its control system, comprising a multi-layer vibration screening device control system and a multi-layer vibration screening device.

[0008] In a first aspect, the present invention provides a multi-layer vibration screening equipment control system, including a vibration sensing and data acquisition module, a spectrum analysis and feature extraction module, an operating status assessment and fault diagnosis module, an adaptive control decision module, and a power drive and execution module.

[0009] The vibration sensing and data acquisition module is responsible for accessing the triaxial accelerometers of each vibration layer, acquiring the vibration acceleration data of each vibration layer, converting the mechanical vibration into a processable digital signal, and performing signal conditioning, anti-aliasing filtering and digital transmission to provide the system with high-precision vibration sensing data.

[0010] The vibration layers are sequentially numbered to obtain the identification symbols for each vibration layer.

[0011] At preset time intervals, the triaxial accelerometers distributed in each vibration layer are accessed to capture the raw three-dimensional vibration acceleration data of the screen body at a preset sampling rate. The raw data is then subjected to DC offset removal, sliding weighting filtering, and anti-aliasing filtering to eliminate environmental interference.

[0012] The specific process of performing DC offset removal is as follows:

[0013] Using the original vibration acceleration data as input parameters, the original acceleration data is processed by DC offset using a preset DC offset formula to obtain the DC offset-removed acceleration data.

[0014] As a preferred embodiment of the present invention, sliding weighted filtering and anti-aliasing filtering are performed based on the DC offset removal processing results. The specific process is as follows:

[0015] The DC offset removal processing results are weighted using a sliding window calculation based on a preset formula to obtain the weighted filtering result and the transfer function of the anti-aliasing filter. The transfer function is then transformed from the frequency domain to the time domain to obtain its time-domain representation. Substituting the time-domain representation and the weighted filtering result into the filtering formula, anti-aliasing filtering is performed to obtain the filtered acceleration data for each vibration layer.

[0016] The spectrum analysis and feature extraction module acquires the filtered acceleration data corresponding to each vibration layer and performs in-depth analysis to extract the contained spectral information. A 2048-point Fast Fourier Transform is performed using an improved Welch algorithm to convert the time-domain signal into a frequency-domain power spectral density distribution. A local maximum search algorithm automatically identifies resonance peaks and calculates five key feature parameters in parallel: resonance frequency, RMS vibration value, kurtosis, spectral entropy, and envelope ratio. Key feature vectors are generated from these key feature parameters and sent to the operational status assessment and fault diagnosis module.

[0017] The specific process of performing a 2048-point Fast Fourier Transform using the improved Welch algorithm is as follows:

[0018] The filtered acceleration data corresponding to each vibration layer is obtained. The filtered acceleration data corresponding to each vibration layer is divided into several data groups according to the preset window size of 2048. The overlap ratio of each data group is 50%, that is, each time the window is slid, there are 1024 data points in the previous data group and the next data group that completely overlap.

[0019] Perform a 2048-point Fast Fourier Transform on each data set to obtain the frequency domain signal corresponding to each data set.

[0020] As a preferred embodiment of the present invention, key feature parameters are extracted from the frequency domain signal, and the specific process is as follows:

[0021] The power spectral density of each vibration in each data group is obtained by squaring the frequency domain signal corresponding to each data group and then averaging the result.

[0022] The frequency corresponding to the local maximum value of the power spectral density is extracted from the frequency domain signal and denoted as the resonant frequency of the vibration layer.

[0023] Using the filtered acceleration data corresponding to each vibration layer as the calculation parameter, and substituting it into the preset formula, the effective vibration value of each vibration layer is calculated.

[0024] Using the effective value of vibration and the filtered acceleration data corresponding to each vibration layer as calculation parameters, the kurtosis of each vibration layer is calculated by substituting them into a preset formula.

[0025] Using power spectral density as the calculation parameter, the spectral entropy of each vibration layer is calculated using a preset formula.

[0026] Using power spectral density as the calculation parameter, the envelope ratio of each vibration layer is calculated using a preset formula.

[0027] As a preferred embodiment of the present invention, the resonance frequency, effective vibration value, kurtosis, spectral entropy and envelope ratio corresponding to each vibration layer are obtained, forming a key feature vector, and sent to the operation status assessment and fault diagnosis module.

[0028] The operational status assessment and fault diagnosis module constructs a 12-class classifier based on a deep belief network (DBN), including typical states such as normal screening, material accumulation, and screen damage. After receiving key feature vectors, they are input into the operational status classifier. Through forward propagation calculation using three hidden layers, the confidence probability distribution of each state is output, realizing operational status identification and performance assessment.

[0029] The 12-class work condition classifier based on Deep Belief Network (DBN) is as follows:

[0030] It includes an input layer, a hidden layer, and an output layer.

[0031] The input layer receives key feature vectors from each vibration layer; the hidden layer performs probability prediction distribution of key feature vectors through forward propagation classification; and the output layer outputs the confidence probability distribution of each state, including 12 operating conditions: normal screening, material accumulation, screen damage, material overload, uneven vibration, motor overheating, abnormal vibration frequency, unbalanced load, bearing damage, electrical fault, abnormal vibration amplitude, and transmission system fault.

[0032] The final output value obtained from the forward propagation classification of the hidden layer is the confidence probability distribution corresponding to the 12 working conditions.

[0033] As a preferred embodiment of the present invention, the working condition of each vibration layer is determined according to the working condition classifier, and the specific process is as follows:

[0034] The key feature vectors of each vibration layer are output to the input layer, and the confidence probability distributions of the 12-class working condition classifier based on Deep Belief Network (DBN) are obtained. The maximum value of all confidence probability distributions is taken. If the maximum value is greater than 90%, it is determined that the vibration layer matches the preset working condition corresponding to the maximum value of the confidence probability distribution, and the corresponding working condition judgment signal is output, including normal screening signal, material accumulation signal, screen breakage signal, material overload signal, uneven vibration signal, motor overheating signal, abnormal vibration frequency signal, unbalanced load signal, bearing damage signal, electrical fault signal, abnormal vibration amplitude signal, or transmission system fault signal.

[0035] The generated operating condition judgment signal is output to the adaptive control decision module and the power drive and execution module.

[0036] The adaptive control decision module determines the working state of each vibration layer based on the working condition judgment signal of each state and matches the corresponding adaptive control decision.

[0037] For vibration layers where the working condition judgment signal is a normal screening signal, no unnecessary operations are performed;

[0038] For vibration layers where the working condition determination signal is a material accumulation signal, an adaptive optimization algorithm is used to adjust the vibration frequency, amplitude, and tilt angle.

[0039] For vibrating layers whose operating condition judgment signal is a screen breakage signal, stop their vibration screening operation;

[0040] For vibration layers whose operating condition judgment signal is a material overload signal, stop their vibration screening operation;

[0041] For vibration layers where the working condition judgment signal is a non-uniform vibration signal, an adaptive optimization algorithm is used to adjust the vibration frequency, amplitude, and tilt angle.

[0042] For vibration layers whose operating condition judgment signal is a motor overheating signal, stop their vibration screening work.

[0043] For vibration layers where the operating condition judgment signal is an abnormal vibration frequency signal, vibration frequency adjustment is performed based on an adaptive optimization algorithm.

[0044] For vibration layers whose operating condition judgment signal is an unbalanced load signal, stop their vibration screening work;

[0045] For vibration layers where the operating condition judgment signal is a bearing damage signal, stop the vibration screening work.

[0046] For vibration layers whose operating condition judgment signals are electrical fault signals, the vibration screening work should be stopped.

[0047] For vibration layers whose operating condition judgment signal is an abnormal vibration amplitude signal, amplitude adjustment is performed based on an adaptive optimization algorithm;

[0048] For vibration layers where the operating condition judgment signal is a transmission system fault signal, the vibration screening work should be stopped.

[0049] The identifier of the vibration layer that enters the vibration frequency, amplitude and tilt angle adjustment based on the adaptive optimization algorithm is sent to the power drive and execution module.

[0050] The power drive and execution module adjusts the vibration frequency, amplitude, and tilt angle based on an adaptive optimization program. It acquires the vibration frequency, amplitude, and tilt angle of each vibration layer corresponding to the received number, performs optimization analysis on these parameters, and sends the analysis results to the directional magnetic field drive module (controlling the vibration frequency), the hydraulic servo system (controlling the amplitude), and the electric actuator (controlling the tilt angle of the vibration layer), respectively, thus completing the adjustment of the vibration frequency, amplitude, and tilt angle control parameters for each vibration layer.

[0051] The adaptive optimization program is as follows:

[0052] Obtain the vibration frequency, amplitude, and tilt angle of the vibrating layer.

[0053] If the working condition judgment signal of the vibration layer is a material accumulation signal or uneven vibration, then the vibration frequency, amplitude and tilt angle are adjusted to the preset maximum values.

[0054] If the working condition judgment signal of the vibration layer is an abnormal vibration frequency signal, then the d / q axis flux linkage and d / q axis current of the vibration motor in the directional magnetic field drive module of the vibration layer are obtained and input into the vibration motor torque adjustment model to calculate the adaptive torque of the vibration motor of the vibration layer. The torque is then output to the directional magnetic field drive module of the vibration layer as the maximum control value of the vibration torque to complete the subsequent vibration screening operation.

[0055] If the operating condition judgment signal of the vibration layer is an abnormal vibration amplitude signal, the eccentric mass, eccentricity, vibrating plate, and total mass of the material of the vibration layer are obtained and input into the amplitude adjustment model to calculate the adaptive amplitude of the hydraulic servo system for controlling the vibration amplitude of the vibration layer. The adaptive amplitude of the hydraulic servo system for controlling the vibration amplitude of the vibration layer is sent to the hydraulic servo system of the vibration layer as the maximum control value of the vibration amplitude to complete the subsequent vibration screening operation.

[0056] Secondly, the present invention provides a multi-layer vibration screening device, comprising several vibration layers, each vibration layer having a vibrating disk, a directional magnetic field drive module, a hydraulic servo system for controlling the amplitude, and an electric push rod machine for controlling the tilt angle of the vibration layer.

[0057] Each vibrating layer has a vibrating disc with a preset aperture size, and the aperture radius of each vibrating disc decreases as the number of layers increases. The vibrating disc in the top layer has the largest aperture radius, and the vibrating disc in the bottom layer has the smallest aperture radius.

[0058] The vibrating plate is used to carry the material, and material particles smaller than the screen hole size pass through the screen holes and enter the vibrating plate of the next vibrating layer.

[0059] The directional magnetic field drive module drives the vibratory feeder to vibrate at a corresponding frequency.

[0060] The hydraulic servo system that controls the amplitude limits the amplitude of the vibratory feeder.

[0061] The electric push rod mechanism that controls the tilt angle of the vibration layer drives the vibration layer to rotate, thus generating the tilt angle.

[0062] Compared with the prior art, the beneficial effects of the present invention are:

[0063] 1. The multi-layer vibration screening equipment control system of this invention collects acceleration data of each vibrating disc through a vibration sensor data acquisition module, and performs signal conditioning processing such as DC offset removal, sliding weighted filtering, and anti-aliasing filtering on the data. After signal optimization, the system can extract key spectral features and perform frequency domain analysis through an improved Welch algorithm, thereby accurately identifying the working conditions of the vibration layer and optimizing screening efficiency. By constructing a 12-class working condition classifier through a deep belief network, it can accurately identify different fault modes and automatically adjust the working state of the equipment, ensuring efficiency and accuracy in the screening process;

[0064] 2. This invention employs an adaptive control decision module, which dynamically adjusts the system based on real-time vibration layer condition information. For example, in cases of material accumulation or uneven vibration, the vibration frequency, amplitude, and tilt angle are automatically adjusted to preset maximum values; conversely, in cases of abnormal vibration amplitude, the amplitude is adjusted to ensure effective screening and stable equipment operation. In this way, the equipment can automatically adapt to changing working environments, improving screening efficiency and reducing manual intervention.

[0065] 3. This invention utilizes a deep belief network to monitor and diagnose the working status of the multi-layer vibrating screening equipment in real time. It can immediately stop the corresponding vibrating screening operation when faults such as screen breakage, motor overheating, or unbalanced load occur, protecting the equipment from further damage. Furthermore, through an optimization strategy based on model predictive control, the equipment can automatically adjust control parameters when problems arise, preventing equipment damage and maintaining the stability of the screening effect, further improving the safety and reliability of the equipment. Attached Figure Description

[0066] To facilitate understanding by those skilled in the art, the present invention will be further described below with reference to the accompanying drawings:

[0067] Figure 1 This is a system block diagram of the present invention. Detailed Implementation

[0068] The technical solution of the present invention will be clearly and completely described below with reference to 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.

[0069] First embodiment:

[0070] Please see Figure 1 As shown, a multi-layer vibration screening equipment control system includes a vibration sensing and data acquisition module, a spectrum analysis and feature extraction module, an operating status assessment and fault diagnosis module, an adaptive control decision module, and a power drive and execution module.

[0071] The vibration sensing and data acquisition module is responsible for accessing the triaxial accelerometers of each vibration layer, acquiring the vibration acceleration data of each vibration layer, converting the mechanical vibration into a processable digital signal, and performing signal conditioning, anti-aliasing filtering and digital transmission to provide the system with high-precision vibration sensing data.

[0072] At preset time intervals, the triaxial accelerometers distributed in each vibration layer i are accessed to capture the raw three-dimensional vibration acceleration data a(i,t) of the screen body at a preset sampling rate. The raw data is then preprocessed with DC offset removal and sliding weighted filtering to eliminate environmental interference. Here, i is the vibration layer number, i=1,2,...,n; n is the total number of vibration layers.

[0073] The specific process of signal conditioning and anti-aliasing filtering is as follows:

[0074] The raw acceleration data a(i,t) is processed to remove DC offset, using the following formula:

[0075]

[0076] Where T is the total number of sampling points. This is the acceleration data after DC offset removal processing.

[0077] It should be noted that the purpose of DC offset removal is to eliminate persistent offsets in the signal caused by equipment, environment, or other factors, causing the signal to fluctuate around zero and avoid affecting subsequent processing. Without DC offset removal, the DC component of the signal may mislead the analysis results, leading to inaccurate frequency distribution in spectrum analysis and affecting the accuracy of feature extraction and fault diagnosis. Furthermore, DC offset may also cause inaccurate amplitude calculations, thus affecting the system's response and control performance.

[0078] Furthermore, based on the results of the DC offset removal process... The specific process for performing sliding weighted filtering and anti-aliasing filtering is as follows:

[0079] By using a preset formula:

[0080] Calculate the results of DC offset processing The results of the moving average filtering The transfer function H(f) of the anti-aliasing filter. Here, M is the preset filter window size, and k is the time index of the filter window, k=1,2,...,M; where... This represents the DC offset processing result at time tk; where m is the order of the low-pass filter, fc is the preset cutoff frequency, and f is the frequency variable.

[0081] Furthermore, filtering is performed based on the transfer function H(f) of the anti-aliasing filter. The transfer function H(f) is transformed from the frequency domain to the time domain to obtain its time-domain representation function h(τ). Substituting the time-domain representation function h(τ) into the filtering formula:

[0082]

[0083] Anti-aliasing filtering was performed to obtain the filtered acceleration data for each vibration layer. .

[0084] It should be noted that anti-aliasing filtering is performed to prevent aliasing during signal sampling, remove frequency components above the Nyquist frequency, remove high-frequency noise caused by the mutual influence of vibration layers, and retain effective low-frequency information in the signal, thereby providing accurate input data for further system analysis.

[0085] The spectrum analysis and feature extraction module acquires the filtered acceleration data corresponding to each vibration layer and performs in-depth analysis to extract the spectral information contained therein. An improved Welch algorithm is used to perform a 2048-point Fast Fourier Transform, converting the time-domain signal into a frequency-domain power spectral density distribution. A local maximum search algorithm automatically identifies resonance peaks and calculates five key feature parameters in parallel: resonance frequency f_res, RMS vibration value, kurtosis Kur, spectral entropy, and envelope ratio. Key feature vectors are generated from these key feature parameters and sent to the operational status assessment and fault diagnosis module.

[0086] Obtain the filtered acceleration data corresponding to each vibration layer. Based on the improved Welch algorithm, a 2048-point Fast Fourier Transform is performed to conduct a frequency-to-time domain transformation to obtain frequency domain data. Key feature parameters are then extracted from this data, including resonance frequency f_res(i), effective vibration value RMS(i), kurtosis Kur(i), spectral entropy Spe(i), and envelope ratio Env(i). The specific process is as follows:

[0087] Filtered acceleration data corresponding to each vibration layer The data is divided into several groups according to a preset window size of 2048. The overlap ratio of each data group is 50%, meaning that 1024 data points completely overlap between the previous and next data groups when the window is slid. Each data group is denoted as wj(i,t), where j is the data group number (j=1,2,...,m), and m is the total number of data groups. Each data group has a length of 2048, meaning each data group contains 2048 filtered acceleration data points corresponding to each vibration layer. .

[0088] Furthermore, a 2048-point Fast Fourier Transform is performed on each data group to obtain the frequency domain signal Xk(f) corresponding to each data group.

[0089] Furthermore, key feature parameters are extracted from the frequency domain signal. The specific process is as follows:

[0090] The power spectral density Pk(f) of each vibration i in each data group is obtained by squaring the frequency domain signal Xk(f) corresponding to each data group and then averaging the result.

[0091] Extract the frequency corresponding to the local maximum value Pk(f)max of the power spectral density from the frequency domain signal Xk(f), and denote it as the resonant frequency f_res(i) of the i-th layer.

[0092] By using a preset formula: Calculate the effective vibration value RMS(i) for each vibration layer i;

[0093] By preset formula Calculate the kurtosis Kur(i) of each vibration layer i;

[0094] By preset formula Calculate the spectral entropy Spe(i) for each vibration layer i.

[0095] By preset formula Calculate the envelope ratio Enc(i) of each vibration layer i.

[0096] Among them Let f_res(i) be the power spectral density from 0 to the resonance frequency f_res(i) in data set j; where f_res(i) is the power spectral density. Let f_res(i) be the power spectral density from the resonance frequency f_res(i) to the maximum frequency f_max in data set j.

[0097] Furthermore, the resonance frequency f_res(i), vibration effective value RMS(i), kurtosis Kur(i), spectral entropy Spe(i), and envelope ratio Env(i) corresponding to each vibration layer i are obtained to form a key feature vector M(i) = {f_res(i), RMS(i), Kur(i), Spe(i), Env(i)}, and sent to the operation status assessment and fault diagnosis module.

[0098] It should be noted that the improved Welch algorithm is a method that improves the accuracy and reliability of spectrum estimation by dividing the signal into multiple overlapping windows and performing FFT on each window.

[0099] The operational status assessment and fault diagnosis module constructs a 12-class classifier based on a deep belief network (DBN), including typical states such as normal screening, material accumulation, and screen damage. After receiving key feature vectors, they are input into the operational status classifier. Through forward propagation calculation using three hidden layers, the confidence probability distribution of each state is output, realizing operational status identification and performance assessment.

[0100] The 12-class work condition classifier based on Deep Belief Network (DBN) is as follows:

[0101] It includes an input layer, a hidden layer, and an output layer.

[0102] The input layer receives the key feature vector M(i) of each vibration layer i; the hidden layer performs probability prediction distribution of the key feature vector M(i) through forward propagation classification; the output layer outputs the confidence probability distribution of each state, including 12 working conditions: normal screening, material accumulation, screen damage, material overload, uneven vibration, motor overheating, abnormal vibration frequency, unbalanced load, bearing damage, electrical fault, abnormal vibration amplitude, and transmission system fault confidence probability distributions: Q1(i), Q2(i), Q3(i), Q4(i), Q5(i), Q6(i), Q7(i), Q8(i), Q9(i), Q10(i), Q11(i), and Q12(i).

[0103] The forward propagation classification formula for the hidden layer is as follows:

[0104]

[0105] Where d is the working condition classification index, d=1,2,...,12; corresponding to 12 working conditions. Specifically, d=1 represents normal screening; d=2 represents material accumulation; d=3 represents screen damage; d=4 represents material overload; d=5 represents uneven vibration; d=6 represents motor overheating; d=7 represents abnormal vibration frequency; d=8 represents unbalanced load; d=9 represents bearing damage; d=10 represents electrical fault; d=11 represents abnormal vibration amplitude; and d=12 represents transmission system fault.

[0106] Where c is the total number of layers in the forward propagation, and h1, h2, ..., hc-1 are the output values ​​of each layer; , , ..., The pre-defined weighting factors for each layer; among them , , ..., Preset bias factors for each layer.

[0107] The specific values ​​of the weighting factor and bias factor are obtained through model training and system parameter tuning, where σ is the activation function, and .

[0108] The working condition of each vibration layer i is determined based on the working condition classifier. The specific process is as follows:

[0109] The key feature vectors M(i) of each vibration layer i are output to the input layer, and the confidence probability distributions of the 12 working condition classifiers based on the deep belief network DBN are obtained: Q1(i), Q2(i), Q3(i), Q4(i), Q5(i), Q6(i), Q7(i), Q8(i), Q9(i), Q10(i), Q11(i) and Q12(i).

[0110] Take the maximum value of all confidence probability distributions. If the maximum value is greater than 90%, it is determined that the vibration layer i matches the preset working condition corresponding to the maximum value of the confidence probability distribution. Output the corresponding working condition judgment signal, including normal screening signal, material accumulation signal, screen damage signal, material overload signal, uneven vibration signal, motor overheating signal, abnormal vibration frequency signal, unbalanced load signal, bearing damage signal, electrical fault signal, abnormal vibration amplitude signal, or transmission system fault signal.

[0111] The generated operating condition judgment signal is output to the adaptive control decision module and the power drive and execution module.

[0112] The adaptive control decision module determines the working state of each vibration layer based on the working condition judgment signal of each state and matches the corresponding adaptive control decision.

[0113] For vibration layer i, where the working condition judgment signal is a normal screening signal, no unnecessary operations are performed;

[0114] For vibration layer i, where the working condition determination signal is a material accumulation signal, the vibration frequency, amplitude, and tilt angle are adjusted based on an adaptive optimization algorithm.

[0115] For vibration layer i, whose working condition judgment signal is a screen breakage signal, stop its vibration screening operation;

[0116] For vibration layer i, whose working condition judgment signal is a material overload signal, stop its vibration screening operation;

[0117] For the working condition judgment signal being a non-uniform vibration signal, an adaptive optimization algorithm is used to adjust the vibration frequency, amplitude, and tilt angle.

[0118] If the operating condition judgment signal is a motor overheating signal, stop its vibration screening operation;

[0119] For vibration layer i, where the working condition judgment signal is an abnormal vibration frequency signal, vibration frequency adjustment is performed based on an adaptive optimization algorithm.

[0120] For vibration layer i, whose working condition judgment signal is an unbalanced load signal, stop its vibration screening work;

[0121] For vibration layer i, where the working condition judgment signal is a bearing damage signal, stop its vibration screening work;

[0122] For vibration layer i, where the operating condition judgment signal is an electrical fault signal, stop its vibration screening operation;

[0123] For vibration layer i, where the working condition judgment signal is an abnormal vibration amplitude signal, amplitude adjustment is performed based on an adaptive optimization algorithm;

[0124] For vibration layer i, where the working condition judgment signal is a transmission system fault signal, stop its vibration screening work.

[0125] The number i of the vibration layer i, which is subject to vibration frequency, amplitude and tilt angle adjustment based on the adaptive optimization algorithm, is sent to the power drive and execution module.

[0126] The power drive and execution module adjusts the vibration frequency, amplitude, and tilt angle based on an adaptive optimization program. It acquires the vibration frequency, amplitude, and tilt angle of each vibration layer i corresponding to the received number i, performs optimization analysis on these parameters, and sends the analysis results to the directional magnetic field drive module (controlling the vibration frequency), the hydraulic servo system (controlling the amplitude), and the electric actuator (controlling the tilt angle of the vibration layer), respectively, thus completing the adjustment of the vibration frequency, amplitude, and tilt angle control parameters for each vibration layer.

[0127] The adaptive optimization program is as follows:

[0128] Obtain the vibration frequency fi, amplitude Ai, and tilt angle θi of vibration layer i.

[0129] If the working condition judgment signal of vibration layer i is a material accumulation signal or uneven vibration, then the vibration frequency fi, amplitude Ai and tilt angle θi are adjusted to the preset maximum values.

[0130] If the condition determination signal of vibration layer i is an abnormal vibration frequency signal, then the d / q axis flux linkage of the vibration motor in the directional magnetic field drive module is obtained. and Obtain d / q axis current and And input the vibration motor torque adjustment model:

[0131]

[0132] The adaptive torque Te of the vibration motor in vibration layer i is calculated and output to the directional magnetic field drive module of vibration layer i, serving as the maximum control value of the vibration torque for subsequent vibration screening. Here, P represents the number of pole pairs of the motor in vibration layer i.

[0133] If the working condition judgment signal of vibration layer i is an abnormal vibration amplitude signal, obtain the eccentric mass me, eccentricity r, and total mass Mt of the vibrating disk and material of vibration layer i, and input them into the amplitude adjustment model:

[0134]

[0135] The adaptive amplitude A of the hydraulic servo system for controlling the vibration amplitude of vibration layer i is calculated. Here, ks is the preset vibratory plate stiffness, λ is the preset frequency ratio, ω is the preset standard vibration frequency, and ζ is the preset standard damping ratio.

[0136] The adaptive amplitude A of the hydraulic servo system controlling the amplitude of vibration layer i is sent to the hydraulic servo system of vibration layer i as the maximum control value of vibration amplitude to complete the subsequent vibration screening operation.

[0137] Second embodiment:

[0138] A multi-layer vibration screening device includes several vibration layers, each vibration layer having a vibrating plate, a directional magnetic field drive module, a hydraulic servo system for controlling the amplitude, and an electric push rod machine for controlling the tilt angle of the vibration layer.

[0139] Each vibrating layer contains vibrating discs with pre-defined sieve apertures, and the radius of the sieve apertures on each vibrating disc decreases as the number of layers increases. The vibrating discs in the top layer have the largest sieve aperture radius, while those in the bottom layer have the smallest.

[0140] The vibrating plate is used to carry the material, and material particles smaller than the screen hole size pass through the screen holes and enter the vibrating plate of the next vibrating layer.

[0141] The directional magnetic field drive module drives the vibratory feeder to vibrate at a corresponding frequency.

[0142] The hydraulic servo system that controls the amplitude limits the amplitude of the vibratory feeder.

[0143] The electric push rod mechanism that controls the tilt angle of the vibration layer drives the vibration layer to rotate, thus generating the tilt angle.

[0144] It should be understood that the terms “comprising” and “including” used in this disclosure and claims indicate the presence of the described features, integrals, steps, operations, elements and / or components, but do not exclude the presence or addition of one or more other features, integrals, steps, operations, elements, components and / or collections thereof.

[0145] It should also be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the disclosure. As used in this disclosure and claims, the singular forms “a,” “an,” and “the” are intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in this disclosure and claims means any combination and all possible combinations of one or more of the associated listed items, and includes such combinations;

[0146] The preferred embodiments of the present invention disclosed above are merely illustrative of the invention. These preferred embodiments do not exhaustively describe all details, nor do they limit the invention to any specific implementation. Clearly, many modifications and variations can be made based on the content of this specification. This specification selects and specifically describes these embodiments to better explain the principles and practical applications of the invention, thereby enabling those skilled in the art to better understand and utilize the invention. The invention is limited only by the claims and their full scope and equivalents.

Claims

1. A control system for a multi-layer vibration screening device, comprising a vibration sensing and data acquisition module, a spectrum analysis and feature extraction module, an operating status assessment and fault diagnosis module, an adaptive control decision module, and a power drive and execution module, characterized in that: The vibration sensing and data acquisition module is responsible for accessing the triaxial accelerometers of each vibration layer to obtain the vibration acceleration data of each vibration layer. The specific process for obtaining the vibration acceleration data of each vibration layer is as follows: Each vibration layer is sequentially numbered to obtain a serial number for each vibration layer. At preset time intervals, the triaxial accelerometers distributed in each vibration layer are accessed according to the serial number of each vibration layer. The raw three-dimensional vibration acceleration data of the screen body is captured at a preset sampling rate. The raw data is then processed by DC offset removal, sliding weighted filtering, and anti-aliasing filtering to obtain the filtered acceleration data corresponding to each vibration layer. The spectrum analysis and feature extraction module acquires the filtered acceleration data corresponding to each vibration layer. This data is then divided into several data groups according to a preset window size of 2048. A 2048-point Fast Fourier Transform is performed on each data group to obtain the corresponding frequency domain signal. Five key feature parameters are extracted from the frequency domain signal, including: resonance frequency, effective vibration value, kurtosis, spectral entropy, and envelope ratio. A local maximum search algorithm automatically identifies resonance peaks, and the five key feature parameters are calculated in parallel to generate key feature vectors, which are then sent to the operation status assessment and fault diagnosis module. The specific process for extracting five key feature parameters from the frequency domain signal is as follows: The power spectral density of each vibration in each data group is obtained by squaring the frequency domain signal corresponding to each data group and then averaging the result. The frequency corresponding to the local maximum value of the power spectral density is extracted from the frequency domain signal and denoted as the resonance frequency of that layer. Using the filtered acceleration data corresponding to each vibration layer as the calculation parameter, and substituting it into the preset formula, the effective vibration value of each vibration layer is calculated. Using the effective value of vibration and the filtered acceleration data corresponding to each vibration layer as calculation parameters, the kurtosis of each vibration layer is calculated by substituting them into a preset formula. Using power spectral density as the calculation parameter, the spectral entropy of each vibration layer is calculated using a preset formula. Using power spectral density as the calculation parameter, the envelope ratio of each vibration layer is calculated using a preset formula; The operation status assessment and fault diagnosis module constructs a 12-class classifier based on a deep belief network (DBN), including typical states such as normal screening, material accumulation, and screen damage. After receiving the key feature vector, it is input into the operation condition classifier. Through forward propagation calculation with three hidden layers, the confidence probability distribution of each state is output to realize operation condition identification and performance assessment. The adaptive control decision module determines the working state of each vibration layer based on the working condition judgment signal of each state and matches the corresponding adaptive control decision. The power drive and execution module adjusts the vibration frequency, amplitude, and tilt angle based on an adaptive optimization program; it acquires the vibration frequency, amplitude, and tilt angle of each vibration layer corresponding to the received number, performs optimization analysis on them, and sends the analysis results to the directional magnetic field drive module that controls the vibration frequency, the hydraulic servo system that controls the amplitude, and the electric actuator that controls the tilt angle of the vibration layer, respectively, to complete the adjustment of the vibration frequency, amplitude, and tilt angle control parameters of each vibration layer.

2. The control system for a multi-layer vibration screening device according to claim 1, characterized in that, The specific process of performing DC offset removal, moving weighted filtering, and anti-aliasing filtering is as follows: Using the original vibration acceleration data as input parameters, the original acceleration data is processed to remove DC offset using a preset DC offset formula to obtain the DC offset-removed acceleration data. The DC offset processing results are weighted by a sliding window using a preset formula to obtain the sliding weighted filtering result and the transfer function of the anti-aliasing filter. The transfer function is transformed from the frequency domain to the time domain to obtain its time domain representation function. The time domain representation function and the sliding weighted filtering result are substituted into the filtering formula to perform anti-aliasing filtering and obtain the filtered acceleration data corresponding to each vibration layer.

3. The control system for a multi-layer vibration screening device according to claim 1, characterized in that, The specific process of performing a 2048-point Fast Fourier Transform is as follows: The filtered acceleration data corresponding to each vibration layer are divided into sliding windows, with an overlap ratio of 50% between adjacent data groups, meaning that 1024 data points completely overlap between the previous and next data groups.

4. The control system for a multi-layer vibration screening device according to claim 1, characterized in that, The 12-class work condition classifier based on Deep Belief Network (DBN) is as follows: It includes an input layer, a hidden layer, and an output layer; The input layer receives key feature vectors from each vibration layer; the hidden layer performs probability prediction distribution on the key feature vectors through forward propagation classification; and the output layer outputs the confidence probability distribution of each state, including 12 operating conditions: normal screening, material accumulation, screen damage, material overload, uneven vibration, motor overheating, abnormal vibration frequency, unbalanced load, bearing damage, electrical fault, abnormal vibration amplitude, and transmission system fault. The final output value obtained by the forward propagation classification of the hidden layer is the confidence probability distribution corresponding to the 12 working conditions. The working condition of each vibration layer is determined based on the working condition classifier. The key feature vectors of each vibration layer are output to the input layer, and the confidence probability distributions of the 12-class working condition classifier based on the deep belief network (DBN) are obtained. The maximum value of all confidence probability distributions is taken. If the maximum value is greater than 90%, it is determined that the vibration layer matches the preset working condition corresponding to the maximum value of the confidence probability distribution, and the corresponding working condition judgment signal is output, including normal screening signal, material accumulation signal, screen breakage signal, material overload signal, uneven vibration signal, motor overheating signal, abnormal vibration frequency signal, unbalanced load signal, bearing damage signal, electrical fault signal, abnormal vibration amplitude signal, or transmission system fault signal.

5. The control system for a multi-layer vibration screening device according to claim 1, characterized in that, The specific process of matching the corresponding adaptive control decision is as follows: For vibration layers where the working condition judgment signal is a normal screening signal, no unnecessary operations are performed; For vibration layers where the working condition determination signal is a material accumulation signal, an adaptive optimization algorithm is used to adjust the vibration frequency, amplitude, and tilt angle. For vibrating layers whose operating condition judgment signal is a screen breakage signal, stop their vibration screening operation; For vibration layers whose operating condition judgment signal is a material overload signal, stop their vibration screening operation; For vibration layers where the working condition judgment signal is a non-uniform vibration signal, an adaptive optimization algorithm is used to adjust the vibration frequency, amplitude, and tilt angle. For vibration layers whose operating condition judgment signal is a motor overheating signal, stop their vibration screening work. For vibration layers where the operating condition judgment signal is an abnormal vibration frequency signal, vibration frequency adjustment is performed based on an adaptive optimization algorithm. For vibration layers whose operating condition judgment signal is an unbalanced load signal, stop their vibration screening work; For vibration layers where the operating condition judgment signal is a bearing damage signal, stop the vibration screening work. For vibration layers whose operating condition judgment signals are electrical fault signals, the vibration screening work should be stopped. For vibration layers whose operating condition judgment signal is an abnormal vibration amplitude signal, amplitude adjustment is performed based on an adaptive optimization algorithm; For vibration layers whose operating condition judgment signals are transmission system fault signals, stop their vibration screening work; The identifier of the vibration layer that enters the vibration frequency, amplitude and tilt angle adjustment based on the adaptive optimization algorithm is sent to the power drive and execution module.

6. The control system for a multi-layer vibration screening device according to claim 1, characterized in that, The adaptive optimization program is as follows: Obtain the vibration frequency, amplitude, and tilt angle of the vibrating layer; If the working condition judgment signal of the vibration layer is a material accumulation signal or uneven vibration, then adjust the vibration frequency, amplitude and tilt angle to the preset maximum value; If the working condition judgment signal of the vibration layer is an abnormal vibration frequency signal, then the d / q axis flux linkage and d / q axis current of the vibration motor in the directional magnetic field drive module of the vibration layer are obtained and input into the vibration motor torque adjustment model to calculate the adaptive torque of the vibration motor of the vibration layer. The torque is then output to the directional magnetic field drive module of the vibration layer as the maximum control value of the vibration torque to complete the subsequent vibration screening operation. If the working condition judgment signal of the vibration layer is an abnormal vibration amplitude signal, obtain the eccentric mass, eccentricity, vibrating plate and total mass of the material of the vibration layer, input them into the amplitude adjustment model, calculate the adaptive amplitude of the hydraulic servo system for controlling the vibration amplitude of the vibration layer; send the adaptive amplitude of the hydraulic servo system for controlling the vibration amplitude of the vibration layer to the hydraulic servo system of the vibration layer as the maximum control value of the vibration amplitude to complete the subsequent vibration screening operation.

7. A multi-layer vibrating screening device, used to implement the control system of the multi-layer vibrating screening device according to any one of claims 1-6, characterized in that, It includes several vibration layers, each vibration layer having a vibrating plate, a directional magnetic field drive module, a hydraulic servo system for controlling the amplitude, and an electric push rod machine for controlling the tilt angle of the vibration layer; Each vibrating layer has a vibrating disc with a preset aperture, and the aperture radius of each vibrating disc decreases as the number of layers increases; the vibrating disc in the uppermost layer has the largest aperture radius, and the vibrating disc in the lowermost layer has the smallest aperture radius. The vibrating plate is used to carry the material, and material particles smaller than the screen hole size pass through the screen holes and enter the vibrating plate of the next vibrating layer. The directional magnetic field drive module drives the vibratory feeder to vibrate at a corresponding frequency; The hydraulic servo system that controls the amplitude limits the amplitude of the vibratory feeder; The electric push rod mechanism that controls the tilt angle of the vibration layer drives the vibration layer to rotate, thus generating the tilt angle.

Images