Screening performance control method of airflow microporous screen

Abstract

The invention relates to a screening performance control method of an airflow microporous screen, belongs to the technical field of screening efficiency control, and solves the problems that a calculation method of the screening efficiency of the airflow microporous screen in the prior art is high in cost, poor in efficiency, low in accuracy and not suitable for complex working conditions. Obtaining first predicted screening efficiency of the airflow microporous screen to be detected, and if the first predicted screening efficiency does not meet an expected target, optimizing the initial design parameters by adopting a genetic algorithm; in the operation process of the to-be-detected airflow microporous screen, for each moment, executing the following operations: if the difference value between the second predicted screening efficiency and the actual screening efficiency is greater than a preset threshold value, optimizing control parameters of a feeding speed controller and a vibration frequency controller based on the second predicted screening efficiency and the actual screening efficiency, therefore, the to-be-detected airflow microporous screen can reach the second predicted screening efficiency. The screening performance control method of the airflow microsieve is high in precision, high in efficiency, low in cost and wide in application range.

Description

Technical Field

[0001] The present invention relates to the technical field of screening efficiency control, and particularly to a method for controlling the screening performance of an air-flow microporous sieve. Background Art

[0002] An air-flow microporous sieve is a device for achieving fine classification of materials based on the principle of air-flow dynamics, abandoning the traditional operating principle of gravitational potential energy. Through the kinetic energy of high-speed air flow, powder particles are fully diffused and sprayed onto the sieve mesh with sufficient kinetic energy, thereby achieving rapid classification, and is particularly suitable for screening ultra-fine powders, light or easily agglomerated materials.

[0003] In the prior art, the methods for studying the screening efficiency of air-flow microporous sieves mainly include three methods: experimental testing, numerical simulation, and theoretical analysis. For experimental testing, it refers to measuring the particle size distribution of materials before and after screening through the actual operation of the equipment and calculating the screening efficiency. However, this method has problems such as high cost, long cycle, and poor repeatability (greatly affected by material characteristics (such as humidity and static electricity)); for numerical simulation, it refers to using computational fluid dynamics (CFD) and discrete element method (DEM) to simulate the air-flow screening process, analyze the particle motion trajectory, air-flow distribution, and screening efficiency. However, this method has the disadvantages of model simplification, difficulty in fully restoring the actual working conditions (such as particle shape and electrostatic effect), and large computational amount; for theoretical analysis, it refers to establishing a mathematical model based on screening dynamics and fluid mechanics theories to predict the screening efficiency. However, this method has the disadvantages of limited accuracy, difficulty in comprehensively considering complex working conditions (such as particle agglomeration and sieve mesh blockage), and narrow application range (the model is usually for specific materials or equipment).

[0004] Therefore, it is necessary to provide a method for controlling the screening performance of an air-flow microporous sieve with high accuracy, high efficiency, low cost, and wide application range. Summary of the Invention

[0005] In view of the above analysis, an embodiment of the present invention aims to provide a method for controlling the screening performance of an air-flow microporous sieve to solve the problems of high cost, poor efficiency, low accuracy, and inapplicability to complex working conditions of the existing calculation methods for the screening efficiency of air-flow microporous sieves.

[0006] An embodiment of the present invention provides a method for controlling the screening performance of an air-flow microporous sieve, including:

[0007] Obtain the initial design parameters of the air-flow microporous sieve to be tested, and input the initial design parameters into the target screening performance prediction model to obtain the first predicted screening efficiency corresponding to the initial design parameters;

[0008] If the first predicted screening efficiency meets the expected goal, the initial design parameters are used as the operating parameters of the airflow microporous sieve to be tested. If it does not meet the expected goal, the genetic algorithm is used to optimize the initial design parameters, and the optimized design parameters are used as the operating parameters of the airflow microporous sieve to be tested;

[0009] During the operation of the airflow microporous sieve to be tested, for each moment, the following operations are performed: Obtain the feeding speed and vibration frequency in the operating parameters at the current moment, input the feeding speed, the vibration frequency, and other parameters in the operating parameters into the target screening performance prediction model to obtain the second predicted screening efficiency corresponding to this moment, and calculate the actual screening efficiency at this moment based on the current actual screening results. If the difference between the second predicted screening efficiency and the actual screening efficiency is greater than the preset threshold, the control parameters of the feeding speed controller and the vibration frequency controller are optimized based on the second predicted screening efficiency and the actual screening efficiency, so that the airflow microporous sieve to be tested can reach the second predicted screening efficiency.

[0010] Based on a further improvement of the above method, the target screening performance prediction model is trained based on a historical operating parameter sample data set. Each piece of data in the historical operating parameter sample data set includes the particle size distribution ratio, water content, sieve hole size, sieve hole shape, sieve surface inclination angle, air flow velocity, vibration frequency and amplitude, feeding speed, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, tail blade length, and screening efficiency.

[0011] Based on a further improvement of the above method, optimizing the control parameters of the feeding speed controller and the vibration frequency controller based on the second predicted screening efficiency and the actual screening efficiency includes: If the difference is greater than the preset critical value, the vibration frequency is adjusted based on the improved grey wolf algorithm, and the feeding speed remains unchanged; If the difference is less than the preset critical value, the feeding speed is adjusted based on the improved grey wolf algorithm, and the vibration frequency remains unchanged; where the preset critical value is greater than the preset threshold.

[0012] Based on a further improvement of the above method, the improved grey wolf algorithm includes:

[0013] A1: Initialize the wolf pack, set the population size and the first preset number of iterations. Each individual {K p , K i , K d} in the population represents a possible solution. Set the initial values of the convergence factor a, the coefficient vector A, and the coefficient vector C. Among them, K p , K i , K d are the control parameters of the feeding speed controller or the control parameters of the vibration frequency controller;

[0014] A2: Calculate the fitness function value for each wolf, and the fitness function is:

[0015] Fitness = (STEF(t) - SSEF(t)) 2 ,

[0016] wherein, STEF(t) is the third predicted screening efficiency corresponding to the moment t, SSEF(t) is the second predicted screening efficiency corresponding to the moment t, and t is the running moment of the airflow microporous sieve to be measured;

[0017] The third predicted screening efficiency refers to inputting the difference between the control parameters corresponding to this wolf, the second predicted screening efficiency and the actual screening efficiency at the moment t into the feed rate controller to obtain the feed rate or inputting it into the vibration frequency controller to obtain the vibration frequency, and then inputting the feed rate or vibration frequency and other parameters in the running parameters at the moment t into the target screening performance prediction model to obtain;

[0018] A3: Sort the wolf pack according to the fitness value, and select the top three as the leading wolves;

[0019] A4: In each iteration, update the positions of non-leading wolves through the positions of the leading wolves, the convergence factor, the coefficient vector A and C; among them, the convergence factor a is obtained through the following method:

[0020]

[0021] wherein, a1 and a2 are constants, σ is an adjustment factor, max(Fitness) is the maximum value of the fitness function, and std(Fitness) is the standard deviation of the fitness function values of all wolves in this iteration; the coefficient vector A is obtained through the following method:

[0022] A = 2a * r1 - a + γ * (max(Fitness) - avg(Fitness)),

[0023] wherein, r1 is a vector randomly generated within the range of [0, 1], γ is the learning rate, and avg(Fitness) is the average value of the fitness function; the coefficient vector C is obtained through the following method:

[0024]

[0025] wherein, d is the current iteration number, D is the first preset iteration number, and β is a constant;

[0026] A5: Calculate the fitness function value of the new position and compare it with the original position. If the new position is better than the original position, update the position of this non-leading wolf, otherwise, keep the position of this non-leading wolf;

[0027] A6: Repeat steps A2 - A5 until the first preset number of iterations is reached, and set the α wolf at this time as the PID control parameter of the microporous sieve for the airflow to be measured.

[0028] Based on a further improvement of the above method, after obtaining the historical operating parameter sample dataset, it is also necessary to perform data cleaning operations on the historical operating parameter sample dataset. The data cleaning operations include: removing outliers using the dynamic threshold method, supplementing missing values using a small BPNN auxiliary model, or normalizing data using a sliding window.

[0029] Based on a further improvement of the above method, the initial design parameters include: particle size distribution ratio, water content, sieve hole size, sieve hole shape, sieve surface inclination angle, air flow velocity, vibration frequency and amplitude, feeding speed, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, tail blade length;

[0030] Optimizing the initial design parameters using the genetic algorithm includes:

[0031] B1: Initialize the population, set the population size N, and each individual in the population includes: particle size distribution ratio, water content, sieve hole size, sieve hole shape, sieve surface inclination angle, air flow velocity, vibration frequency and amplitude, feeding speed, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, tail blade length;

[0032] Perform binary encoding on the particle size distribution ratio, water content, sieve hole size, sieve hole shape, sieve surface inclination angle, air flow velocity, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, and tail blade length, and perform Gray code encoding on the vibration frequency and amplitude, and feeding speed;

[0033] B2: Calculate the fitness value of each individual according to the fitness function, where the fitness function is:

[0034]

[0035] y z is the third screening efficiency obtained by inputting the z-th individual into the target screening performance prediction model, is the screening efficiency corresponding to the initial design parameters, z = 1, 2, 3,..., N;

[0036] B3: Selection operation, select individuals with higher fitness from the current population as parents based on the tournament algorithm for generating the next generation;

[0037] B4: Crossover operation, for the parameter group with binary encoding, use the single-point crossover method, and for the parameter group with Gray code encoding, use the multi-point crossover method;

[0038] B5: Mutation operation. Select 10% of the individuals with the highest fitness from the population for local search. For each selected individual, generate several neighborhood solutions in its vicinity. Search for solutions better than the current solution among the neighborhood solutions. If a better solution is found, replace the current solution and put the individuals after local search back into the population;

[0039] B6: Generate a new population. Combine the parent and offspring individuals, select a preset number of individuals according to the fitness values as the new generation population, and determine whether the iteration stop condition is reached. If it is reached, use the optimal individual in the current population as the optimized operating parameters.

[0040] For a further improvement based on the above method, the neighborhood solutions are generated in the following way:

[0041]

[0042] where x new_p is the newly generated neighborhood solution, x old is the current individual, represents rounding up, q is the current iteration number, and Q is the second preset iteration number.

[0043] For a further improvement based on the above method, the target screening performance prediction model is a BP neural network model, and a multi-stage training strategy is adopted to train the BP neural network model; the activation function of the BP neural network model is:

[0044] f(x) = max(0, x) + αmin(0, x),

[0045] where α is an adjustable parameter, 0 < α < 0.1, and the loss function is:

[0046] L = λ * MSE + (1 - λ) * HuberLoss,

[0047] where MSE is the mean square error, HuberLoss is the Huber loss, and 0 < λ < 1.

[0048] For a further improvement based on the above method, in the warm-up stage of the multi-stage training strategy, the learning rate of each iteration increases by 0.0001 compared with the previous iteration.

[0049] For a further improvement based on the above method, collect the intermediate data during the operation of the airflow microporous sieve to be measured. After the screening process ends, retrain the target screening performance prediction model based on the intermediate data to improve the stability of the target screening performance prediction model.

[0050] Compared with the prior art, the present invention can achieve at least one of the following beneficial effects:

[0051] 1. The present invention provides a method for controlling the screening performance of an air flow microporous sieve. Based on the predicted screening efficiency and the actual screening efficiency, the output of the feed rate PID controller and the vibration frequency PID controller is controlled, realizing the combination of the fast prediction advantage of the target screening performance prediction model and the real-time adjustment advantage of PID control. It not only realizes the accurate prediction of the screening efficiency of the air flow microporous sieve by the target screening performance prediction model, but also uses PID control to adjust parameters in real time according to the prediction results, thereby providing a screening performance control method with high accuracy and high efficiency.

[0052] 2. The present invention provides a method for controlling the screening performance of an air flow microporous sieve. Based on the improved grey wolf algorithm, the optimization of PID control parameters is realized. It not only improves the accuracy of obtaining PID control parameters, but also optimizes the mechanism of parameter self-adaptive adjustment of the PID controller in the face of complex scenario changes, reduces costs and energy consumption, and improves the usage efficiency. Further, through the optimization of the convergence factor and coefficient vector in the grey wolf algorithm, the grey wolf algorithm can explore the solution space more widely in the initial stage, avoid falling into the local optimum prematurely, and increase the chance of finding the global optimum. It can approach the optimum solution more efficiently during the search process, reduce unnecessary calculations, thereby accelerating the convergence speed. In addition, the optimized parameters can make the algorithm perform more stably when facing different problems, reduce the dependence on the initial conditions or problem characteristics, and improve the robustness of the algorithm, so that the grey wolf algorithm can find the optimum solution faster, more accurately and more stably.

[0053] 3. The present invention provides a method for controlling the screening performance of an air flow microporous sieve. After calculating the screening efficiency based on the current initial design parameters, if the screening efficiency requirement is not met, the genetic algorithm is used to obtain the initial design parameters that meet the screening efficiency. Therefore, the present invention reduces the difficulty of adjusting the initial design parameters. It does not require relevant professionals to adjust according to professional knowledge, and only relies on the provided genetic algorithm to quickly obtain the initial design parameters matching the current air flow microporous sieve, and this method is applicable to any complex scenario. Further, by combining neighborhood solution optimization in the mutation operation of the genetic algorithm, high-quality solutions can be quickly found through efficient local search, which can not only improve the search efficiency and solution quality of the algorithm, but also enhance the robustness and global search ability of the algorithm, thus showing better performance in solving complex optimization problems.

[0054] In the present invention, the above technical solutions can also be combined with each other to achieve more preferred combination schemes. Other features and advantages of the present invention will be described in the subsequent description, and some advantages can be made obvious from the description, or understood by implementing the present invention. The purpose and other advantages of the present invention can be realized and obtained through the content specifically pointed out in the description and the drawings. Brief Description of the Drawings

[0055] The drawings are only for the purpose of showing specific embodiments and are not considered as a limitation of the present invention. Throughout the drawings, the same reference signs denote the same components;

[0056] Figure 1 It is an exemplary diagram of a method for controlling the screening performance of an air flow microporous sieve in an embodiment of the present invention. Detailed Embodiments

[0057] The preferred embodiments of the present invention will be specifically described below with reference to the drawings. The drawings form a part of this application and are used together with the embodiments of the present invention to explain the principles of the present invention, rather than to limit the scope of the present invention.

[0058] A specific embodiment of the present invention discloses a method for controlling the screening performance of an air flow microporous sieve, as Figure 1 shown, including:

[0059] S1: Obtain the initial design parameters of the air flow microporous sieve to be measured, and input the initial design parameters into the target screening performance prediction model to obtain the first predicted screening efficiency corresponding to the initial design parameters.

[0060] The factors affecting the screening efficiency of the air flow microporous sieve mainly include the following aspects: (1) Material characteristics, including: particle shape, particle size distribution, moisture content, friction characteristics, fluidity, etc.; (2) Equipment parameters, including: sieve hole size and shape, sieve mesh opening rate, air flow velocity, equipment motion state, tooth-shaped blade parameters, tail blade parameters, etc.; (3) Operating conditions, including: material layer thickness, feeding uniformity, environmental factors, etc.; (4) Process parameters, including: air flow velocity, feeding method, screening time, etc. The design parameters finally selected by the present invention based on production experience and feasibility include: particle size distribution ratio, moisture content, sieve hole size, sieve hole shape, sieve surface inclination angle, air flow velocity, vibration frequency and amplitude, feeding speed, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, tail blade length.

[0061] The target screening performance prediction model is a BP neural network model, and a multi-stage training strategy is adopted to train the BP neural network model; the activation function of the BP neural network model is:

[0062] f(x) = max(0, x) + αmin(0, x),

[0063] where α is an adjustable parameter used to enhance the adaptability of the model to different data patterns, 0 < α < 0.1. Exemplarily, the value of α is 0.01, and the loss function is:

[0064] L = λ * MSE + (1 - λ) * HuberLoss,

[0065] where MSE is the mean squared error, HuberLoss is the Huber loss, 0 < λ < 1, MSE is used to measure the average error between the model prediction value and the true value, and the Huber loss has better robustness to outliers. Exemplarily, λ is 0.5.

[0066] The target screening performance prediction model is trained based on a historical operating parameter sample data set. Each piece of data in the historical operating parameter sample data set includes the particle size distribution ratio, water content, sieve hole size, sieve hole shape, sieve surface inclination angle, air flow velocity, vibration frequency and amplitude, feeding speed, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, tail blade length, and screening efficiency. After obtaining the historical operating parameter sample data set, it is also necessary to perform a data cleaning operation on the historical operating parameter sample data set. The data cleaning operation includes one or more of the following: removing outliers using the dynamic threshold method, supplementing missing values using a small BPNN auxiliary model, or normalizing the data using a sliding window.

[0067] Exemplarily, the collected historical operation parameter sample data is cleaned. The dynamic threshold method based on statistical distribution and equipment physical characteristics is used to judge and remove outliers. The missing values are initially inferred according to the physical principle and working mode of the equipment, and then accurately filled by a small BPNN auxiliary model. The dynamic adaptive normalization method is adopted. According to different stages and working conditions of the equipment operation, the sliding window technology is used to set different normalization ranges for data in different time periods to complete data preprocessing. Specifically, high-precision pressure sensors, flow sensors and other measuring devices are selected and installed at key measuring points of the air flow microporous sieve to detect air flow pressure, gas flow rate, sieve body inclination angle, etc. According to the operation characteristics of the equipment, the data acquisition frequency is set to ensure that the subtle changes during the equipment operation can be captured and the data is transmitted in real time. The dynamic threshold method based on statistical distribution and equipment physical characteristics is used to judge outliers. According to the historical operation data and physical principle of the equipment, the normal value range of each parameter is determined. For example, according to the design specifications and actual operation experience of the air flow microporous sieve, the normal range of the sieve body inclination angle is [-4°, 4°]. Then, by calculating the mean and standard deviation of the data and combining the dynamic threshold factor, the threshold range is adjusted in real time. For the data points outside the threshold range, they are judged as outliers and marked. Finally, the median filtering or the interpolation method based on neighborhood data is used to replace the outliers to ensure the accuracy of the data. The missing values are initially inferred according to the physical principle and working mode of the equipment. If the feeding speed data is missing, it can be initially estimated according to the feeding speed in the previous and subsequent time periods and the operation state of the equipment (such as whether it is in the stable operation stage). Then, a small BPNN auxiliary model is used for accurate filling. This auxiliary model takes other relevant parameters (such as disturbance frequency, sieve body inclination angle, etc.) as inputs and the parameter corresponding to the missing value as the output. Through the training of a large amount of historical data, the auxiliary model can accurately predict the missing values. During the training process, the mean square error is used as the loss function, and the Adam optimization algorithm is used for model training until the model converges. The dynamic adaptive normalization method is adopted. According to different stages and working conditions of the equipment operation, the sliding window technology is used to set different normalization ranges for data in different time periods. In the equipment startup stage and stable operation stage, different normalization parameters are set respectively due to different change ranges and characteristics of the parameters. Specifically, for each parameter, its maximum and minimum values are calculated within the sliding window, and then the data is normalized to the [0, 1] interval. The formula is: where x is the original data, x min and x max are the minimum and maximum values within the sliding window respectively, and x norm is the normalized data.

[0068] Optionally, a BP neural network model is constructed using Python's TensorFlow deep learning framework. The number of input layer nodes is determined according to the parameter quantity of the air flow microporous sieve, with a total of 12 nodes, corresponding to the particle size distribution ratio, water content, sieve hole size, sieve hole shape, sieve surface inclination angle, air flow velocity, vibration frequency and amplitude, feeding speed, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, and tail blade length respectively. The hidden layer is initially set to 2 layers, with each layer containing 10 nodes, and can be adjusted subsequently according to the model performance. The output layer is set to 1 node, using the Sigmoid function, and the output result is the screening efficiency. The weights and biases of the neural network model are adjusted by combining the Adam optimization algorithm and the learning rate annealing technique until the model converges to achieve the expected prediction accuracy. During the training process, the loss value and accuracy of the model on the test set are continuously monitored. When the loss value converges below 0.01 and the accuracy reaches over 97%, it is considered that the model has achieved the expected prediction accuracy.

[0069] The multi-stage training strategy improves the training efficiency and generalization ability of the model by adjusting the learning rate and model parameters in stages. This strategy usually includes three stages: warm-up, acceleration, and fine-tuning, each with its specific goals and methods. Specifically, in the warm-up stage, the learning rate is set to 0.01, the batch size is 16, and the number of training iterations is 200 times to enable the model to initially adapt to the data characteristics. In the acceleration stage, the learning rate is adjusted to 0.001, the batch size is increased to 32, and the number of training iterations is increased to 500 times to accelerate the convergence speed of the model. In the fine-tuning stage, the learning rate is further reduced to 0.0001, the batch size remains 32, and the number of training iterations is 300 times to finely adjust the model and improve the prediction accuracy of the model.

[0070] Exemplarily, in the warm-up stage, the learning rate of each iteration increases by 0.0001 compared to the previous iteration.

[0071] After obtaining the target screening performance prediction model, the initial design parameters of the air flow microporous sieve to be measured are input into the target screening performance prediction model to obtain the first predicted screening efficiency corresponding to the initial design parameters.

[0072] Exemplarily, in order to further improve the prediction accuracy and stability of the target screening performance prediction model, new operation data (i.e., intermediate data) of the airflow microporous sieve to be measured is collected every certain period of time (e.g., 15 minutes), added to the training dataset, and the target screening performance prediction model is retrained and optimized. Specifically, new operation data is collected every fixed period of time, the newly collected data is added to the training dataset, and at the same time, the earliest same amount of data is deleted to ensure the timeliness of the training dataset and the stability of the data volume. After the screening process ends, the target screening performance prediction model is retrained using the updated training dataset. During the retraining process, the model structure remains unchanged, and the same training strategy, loss function, and optimization algorithm as the initial training are adopted. Through retraining, the model can adapt to the changes in the equipment operation conditions, improving the prediction accuracy and stability of the model.

[0073] S2: If the first predicted screening efficiency meets the expected target, the initial design parameters are used as the operation parameters of the airflow microporous sieve to be measured; if it does not meet the expected target, the genetic algorithm is used to optimize the initial design parameters, and the optimized design parameters are used as the operation parameters of the airflow microporous sieve to be measured.

[0074] The initial design parameters include: particle size distribution ratio, water content, sieve pore size, sieve pore shape, sieve surface inclination angle, air flow velocity, vibration frequency and amplitude, feeding speed, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, and tail blade length.

[0075] Optimizing the initial design parameters using the genetic algorithm includes:

[0076] B1: Initialize the population, set the population size N, and each individual in the population includes: particle size distribution ratio, water content, sieve pore size, sieve pore shape, sieve surface inclination angle, air flow velocity, vibration frequency and amplitude, feeding speed, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, and tail blade length.

[0077] The particle size distribution ratio, water content, sieve pore size, sieve pore shape, sieve surface inclination angle, air flow velocity, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, and tail blade length are encoded in binary, and the vibration frequency and amplitude, and feeding speed are encoded in Gray code.

[0078] B2: Calculate the fitness value of each individual according to the fitness function, where the fitness function is:

[0079]

[0080] y z is the third screening efficiency obtained by inputting the z-th individual into the target screening performance prediction model. is the screening efficiency corresponding to the initial design parameters, where z = 1, 2, 3, ..., N.

[0081] B3: Selection operation. Based on the tournament algorithm, select individuals with higher fitness from the current population as parents for generating the next generation.

[0082] B4: Crossover operation. For the parameter group with binary encoding, use the single-point crossover method; for the parameter group with Gray code encoding, use the multi-point crossover method. Such encoding methods can better adapt to the characteristics of different parameters and improve the search efficiency of the genetic algorithm.

[0083] B5: Mutation operation. Select 10% of the individuals with the highest fitness from the population for local search. For each selected individual, generate several neighborhood solutions in its vicinity, search for a solution better than the current solution among the neighborhood solutions. If a better solution is found, replace the current solution and put the individuals after local search back into the population.

[0084] The neighborhood solutions are generated in the following way:

[0085]

[0086] where x new_p is the newly generated neighborhood solution, x old is the current individual, denotes rounding up, q is the current iteration number, and Q is the second preset iteration number.

[0087] B6: Generate a new population. Combine the parent and offspring individuals, select a preset number of individuals as the new generation population according to the fitness values, and determine whether the second preset iteration number is reached. If it is reached, take the optimal individual in the current population as the optimized operating parameters.

[0088] Preferably, the second preset iteration number is 1000.

[0089] In the genetic operations, the crossover operation adopts different methods according to the characteristics of different groups of parameters. For the parameter group with binary encoding, the single-point crossover method is used; for the parameter group with Gray code encoding, the multi-point crossover method is used. The mutation operation introduces a local search mechanism. During the mutation process, a small-range random perturbation is performed on each parameter, and a local search algorithm is used to find a better parameter value. After multiple rounds of genetic operations, the optimal parameter combination is obtained. Apply the optimal parameter combination to production.

[0090] After this step, it can be confirmed whether the current initial design parameters can achieve the expected goal. If they can, the initial design parameters are directly used as the operating parameters of the airflow microporous sieve to be tested. If they do not meet the expectations, adjustments are made. The adjustment basis of the present invention is based on the initial design parameters, reducing the workload of designers and improving the debugging efficiency, and finding the design parameters that meet the expected goal faster and better. Exemplarily, the expected goal is that the screening efficiency reaches 96%.

[0091] It can be understood that the implementation of initialization operations, crossover operations, mutation operations, etc. in the genetic algorithm is common knowledge in the art, and the present invention does not make specific limitations here.

[0092] S3: During the operation of the airflow microporous sieve to be tested, for each moment, the following operations are performed: Obtain the feeding speed and vibration frequency in the operating parameters at the current moment, input the feeding speed, the vibration frequency, and other parameters in the operating parameters into the target screening performance prediction model to obtain the corresponding second predicted screening efficiency at this moment, and calculate the actual screening efficiency at this moment based on the current actual screening result. If the difference between the second predicted screening efficiency and the actual screening efficiency is greater than the preset threshold, optimize the control parameters of the feeding speed controller and the vibration frequency controller based on the second predicted screening efficiency and the actual screening efficiency, so that the airflow microporous sieve to be tested can reach the second predicted screening efficiency.

[0093] Optimizing the control parameters of the feeding speed controller and the vibration frequency controller based on the second predicted screening efficiency and the actual screening efficiency includes: If the difference is greater than the preset critical value, preferentially adjust the vibration frequency based on the improved grey wolf algorithm, and keep the feeding speed unchanged; if the difference is less than the preset critical value, preferentially adjust the feeding speed based on the improved grey wolf algorithm, and keep the vibration frequency unchanged; where the preset critical value is greater than the preset threshold. It can be understood that when the difference is greater than the preset critical value, the vibration frequency is preferentially adjusted because the vibration frequency has a more direct impact on the screening efficiency. If the difference is less than the preset critical value, the feeding speed is preferentially adjusted to maintain the continuity of production.

[0094] Exemplarily, the preset threshold is 0.5, and the preset critical value is 1.2.

[0095] The improved grey wolf algorithm includes:

[0096] A1: Initialize the wolf pack, set the population size and the first preset number of iterations. Each individual {K p , K i , K d} in the population represents a possible solution. Set the initial values of the convergence factor a, the coefficient vector A, and the coefficient vector C. Among them, K p , K i , Kd It is a control parameter of the feed rate controller or a control parameter of the vibration frequency controller.

[0097] A2: Calculate the fitness function value of each wolf, and the fitness function is:

[0098] Fitness = (STEF(t) - SSEF(t)) 2 ,

[0099] wherein, STEF(t) is the third predicted screening efficiency corresponding to the moment t, SSEF(t) is the second predicted screening efficiency corresponding to the moment t, and t is the running moment of the airflow microporous sieve to be measured.

[0100] The third predicted screening efficiency refers to inputting the difference between the control parameter corresponding to this wolf, the second predicted screening efficiency and the actual screening efficiency corresponding to the moment t into the feed rate controller to obtain the feed rate or inputting it into the vibration frequency controller to obtain the vibration frequency, and then inputting the feed rate or the vibration frequency and other parameters in the running parameters at the moment t into the target screening performance prediction model to obtain. Among them, the feed rate u1 is:

[0101]

[0102] The vibration frequency u2 is:

[0103]

[0104] The deviation e is:

[0105] e = SSEF(t) - ASEF(t), ASEF(m) is the actual screening efficiency, is the control parameter corresponding to the feed rate PID controller, is the control parameter corresponding to the vibration frequency PID controller.

[0106] A3: Sort the wolf pack according to the fitness value, and select the top three as the leading wolves.

[0107] A4: In each iteration, update the positions of non-leading wolves through the positions of leading wolves, the convergence factor, the coefficient vector A and C; among them, the convergence factor a is obtained through the following method:

[0108]

[0109] wherein, a1, a2 are constants, σ is an adjustment factor, max(Fitness) is the maximum value of the fitness function, and std(Fitness) is the standard deviation of the fitness function values of all wolves in this iteration; the coefficient vector A is obtained through the following method:

[0110] A = 2a * r1 - a + γ * (max(Fitness) - avg(Fitness)),

[0111] where r1 is a vector randomly generated within the range of [0, 1], γ is the learning rate, and avg(Fitness) is the average value of the fitness function. Specifically:

[0112] A w = 2a * r 1w - a + γ * (max(Fitness) - avg(Fitness)),

[0113] A w is the w-th element in the coefficient vector A, r 1w is the w-th element in the vector r1, where w = 1, 2, 3,..., N. N usually represents the dimension of the problem. Exemplarily, in the present invention, w = 1, 2, 3. The coefficient vector C is obtained in the following manner:

[0114]

[0115] where d is the current iteration number, D is the first preset iteration number, and β is a constant. It can be understood that substituting the corresponding elements in r1 into the above formula can obtain the values of the corresponding elements in the coefficient vector C.

[0116] Preferably, the value of the constant a1 is set to 0.5, the value of the constant a2 is set to 1.8, the value of the adjustment factor σ is 0.8, the learning rate γ is 0.6, and the first preset iteration number is 800.

[0117] A5: Calculate the fitness function value of the new position and compare it with the original position. If the new position is better than the original position, update the position of this non - leading wolf; otherwise, retain the position of this non - leading wolf.

[0118] A6: Repeat steps A2 - A5 until the first preset iteration number is reached, and set the leader α - wolf at this time as the PID control parameters of the airflow microporous sieve to be measured.

[0119] It can be understood that the setting of the initial values of the convergence factor a, the coefficient vector A, and the coefficient vector C in the grey wolf algorithm, as well as the implementation of operations such as position update, are all well - known common knowledge in the art, and the present invention does not make specific limitations here.

[0120] Compared with the prior art, a method for controlling the screening performance of an air flow microporous sieve provided in this embodiment controls the output of the feeding speed PID controller and the vibration frequency PID controller based on the predicted screening efficiency and the actual screening efficiency, combining the fast prediction advantage of the target screening performance prediction model with the real-time adjustment advantage of PID control. It not only realizes the accurate prediction of the screening efficiency of the air flow microporous sieve by the target screening performance prediction model, but also uses PID control to adjust parameters in real time according to the prediction results, thus providing a screening performance control method with high accuracy and high efficiency. The optimization of the PID control parameters is realized based on the improved grey wolf algorithm, which not only improves the accuracy of obtaining the PID control parameters, but also optimizes the mechanism of the PID controller for self-adaptive adjustment of parameters in the face of complex scenario changes, reducing costs and energy consumption and improving the usage efficiency. Further, by optimizing the convergence factor and coefficient vector in the grey wolf algorithm, the grey wolf algorithm can explore the solution space more widely in the initial stage, avoid falling into the local optimum prematurely, and increase the chance of finding the global optimum. It can approach the optimum solution more efficiently during the search process, reduce unnecessary calculations, thus accelerating the convergence speed. In addition, the optimized parameters can make the algorithm perform more stably when facing different problems, reduce the dependence on the initial conditions or problem characteristics, and improve the robustness of the algorithm, so that the grey wolf algorithm can find the optimum solution faster, more accurately and more stably. After calculating the screening efficiency based on the current initial design parameters, if the screening efficiency requirement is not met, the genetic algorithm is used to obtain the initial design parameters that meet the screening efficiency. Therefore, the present invention reduces the difficulty of adjusting the initial design parameters. It does not require relevant professionals to adjust according to professional knowledge, and only relies on the provided genetic algorithm to quickly obtain the initial design parameters that match the current air flow microporous sieve, and this method is applicable to any complex scenario. Further, by combining neighborhood solution optimization in the mutation operation of the genetic algorithm, high-quality solutions can be quickly found through efficient local search, which can not only improve the search efficiency and solution quality of the algorithm, but also enhance the robustness and global search ability of the algorithm, thus showing better performance when solving complex optimization problems.

[0121] Those skilled in the art can understand that all or part of the processes of implementing the methods in the above embodiments can be completed by instructing relevant hardware through a computer program, and the program can be stored in a computer-readable storage medium. Among them, the computer-readable storage medium is a magnetic disk, an optical disk, a read-only memory or a random access memory, etc.

[0122] The above is only a preferred specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any changes or substitutions that can be easily thought of by those skilled in the art within the technical scope disclosed by the present invention should be covered by the protection scope of the present invention.

Claims

1. A method for controlling the screening performance of an air-flow microporous sieve, characterized in that Including: Obtain the initial design parameters of the airflow microporous sieve to be measured, and input the initial design parameters into the target screening performance prediction model to obtain the first predicted screening efficiency corresponding to the initial design parameters; If the first predicted screening efficiency meets the expected target, use the initial design parameters as the operating parameters of the airflow microporous sieve to be measured. If it does not meet the expected target, use the genetic algorithm to optimize the initial design parameters, and use the optimized design parameters as the operating parameters of the airflow microporous sieve to be measured; During the operation of the airflow microporous sieve to be measured, for each moment, perform the following operations: Obtain the feeding speed and vibration frequency in the operating parameters at the current moment, input the feeding speed, the vibration frequency, and other parameters in the operating parameters into the target screening performance prediction model to obtain the second predicted screening efficiency corresponding to this moment, and calculate the actual screening efficiency at this moment based on the current actual screening result. If the difference between the second predicted screening efficiency and the actual screening efficiency is greater than the preset threshold, optimize the control parameters of the feeding speed controller and the vibration frequency controller based on the second predicted screening efficiency and the actual screening efficiency, so that the airflow microporous sieve to be measured can reach the second predicted screening efficiency.

2. The screening performance control method of an air flow microporous sieve according to claim 1, characterized in that, The target screening performance prediction model is trained based on the historical operating parameter sample data set. Each piece of data in the historical operating parameter sample data set includes the particle size distribution ratio, water content, sieve hole size, sieve hole shape, sieve surface inclination angle, air flow velocity, vibration frequency and amplitude, feeding speed, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, tail blade length, and screening efficiency.

3. The screening performance control method of an air flow microporous sieve according to claim 1, characterized in that, Optimizing the control parameters of the feeding speed controller and the vibration frequency controller based on the second predicted screening efficiency and the actual screening efficiency includes: If the difference is greater than the preset critical value, adjust the vibration frequency based on the improved grey wolf algorithm, and keep the feeding speed unchanged; If the difference is less than the preset critical value, adjust the feeding speed based on the improved grey wolf algorithm, and keep the vibration frequency unchanged; where the preset critical value is greater than the preset threshold.

4. A screening performance control method for an air flow microporous sieve according to claim 3, characterized in that, The improved grey wolf algorithm includes: A1: Initialize the wolf pack, set the population size and the first preset number of iterations. Each individual {K p , K i , K d} in the population represents a possible solution. Set the initial values of the convergence factor a, the coefficient vector A, and the coefficient vector C. Among them, K p , K i , K d are the control parameters of the feed rate controller or the control parameters of the vibration frequency controller; A2: Calculate the fitness function value of each wolf, and the fitness function is: Fitness=(STEF(t)-SSEF(t)) 2 , where STEF(t) is the third predicted screening efficiency corresponding to the t-th moment, SSEF(t) is the second predicted screening efficiency corresponding to the t-th moment, and t is the operating moment of the airflow microporous sieve to be measured; The third predicted screening efficiency refers to inputting the control parameters corresponding to this wolf and the difference between the second predicted screening efficiency and the actual screening efficiency at the t-th moment into the feeding speed controller to obtain the feeding rate or inputting it into the vibration frequency controller to obtain the vibration frequency, and then inputting the feeding rate or the vibration frequency and other parameters in the operating parameters at the t-th moment into the target screening performance prediction model to obtain; A3: Sort the wolf pack according to the fitness value, and select the top three as the leading wolves; A4: In each iteration, update the positions of the non-leading wolves through the positions of the leading wolves, the convergence factor, the coefficient vectors A and C; where the convergence factor a is obtained in the following way: where a1 and a2 are constants, σ is a regulation factor, max(Fitness) is the maximum value of the fitness function, and std(Fitness) is the standard deviation of the fitness function values of all wolves in this iteration; the coefficient vector A is obtained in the following way: A = 2a * r1 - a + γ * (max(Fitness) - avg(Fitness)), where r1 is a vector randomly generated within the range of [0, 1], γ is the learning rate, and avg(Fitness) is the average value of the fitness function; the coefficient vector C is obtained in the following way: where d is the current iteration number, D is the first preset iteration number, and β is a constant; A5: Calculate the fitness function value of the new position and compare it with the original position. If the new position is better than the original position, update the position of this non-leader wolf; otherwise, retain the position of this non-leader wolf. A6: Repeat A2 - A5 until the first preset iteration number is reached, and set the α wolf at this time as the PID control parameter of this airflow microporous sieve to be measured.

5. The screening performance control method of an air flow microporous sieve according to claim 1, characterized in that After obtaining the historical operation parameter sample data set, it is also necessary to perform data cleaning operations on the historical operation parameter sample data set. The data cleaning operations include: removing outliers using the dynamic threshold method, supplementing missing values using a small BPNN auxiliary model, or normalizing data using a sliding window.

6. A screening performance control method for an air flow microporous sieve according to claim 1, characterized in that, The initial design parameters include: particle size distribution ratio, water content, sieve hole size, sieve hole shape, sieve surface inclination angle, air flow velocity, vibration frequency and amplitude, feeding speed, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, and tail blade length; Optimizing the initial design parameters using the genetic algorithm includes: B1: Initialize the population, set the population size N, and each individual in the population includes: particle size distribution ratio, water content, sieve hole size, sieve hole shape, sieve surface inclination angle, air flow velocity, vibration frequency and amplitude, feeding speed, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, and tail blade length; Perform binary encoding on the particle size distribution ratio, water content, sieve hole size, sieve hole shape, sieve surface inclination angle, air flow velocity, tooth-shaped blade spacing, tooth-shaped blade length, tail blade spacing, and tail blade length, and perform Gray code encoding on the vibration frequency and amplitude, and feeding speed; B2: Calculate the fitness value of each individual according to the fitness function, where the fitness function is: y z The third screening efficiency obtained by inputting the target screening performance prediction model for the z-th individual is the screening efficiency corresponding to the initial design parameters, where z = 1, 2, 3,..., N; B3: Selection operation, select individuals with higher fitness from the current population as parents based on the tournament algorithm for generating the next generation; B4: Crossover operation, for the parameter group with binary encoding, use the single-point crossover method, and for the parameter group with Gray code encoding, use the multi-point crossover method; B5: Mutation operation, select 10% of the individuals with the highest fitness in the population for local search. For each selected individual, generate several neighborhood solutions near it, and find a solution better than the current solution in the neighborhood solutions. If a better solution is found, replace the current solution, and put the individuals after local search back into the population; B6: Generate a new population, combine the parent and offspring individuals, select a preset number of individuals as the new generation population according to the fitness value, and determine whether the iteration stop condition is reached. If so, use the optimal individual in the current population as the optimized operating parameter.

7. A method for controlling the screening performance of an air flow microporous sieve according to claim 6, characterized in that, The neighborhood solution is generated in the following way: where x new_p is the newly generated neighborhood solution, and x old is the current individual, denotes rounding up, q is the current iteration number, and Q is the second preset iteration number.

8. A method for controlling the screening performance of an air flow microporous sieve according to claim 1, characterized in that The target screening performance prediction model is a BP neural network model, and a multi-stage training strategy is used to train the BP neural network model; the activation function of the BP neural network model is: f(x) = max(0, x) + αmin(0, x), where α is an adjustable parameter, 0 < α < 0.1, and the loss function is: L = λ * MSE + (1 - λ) * HuberLoss, where MSE is the mean square error, HuberLoss is the Huber loss, and 0 < λ < 1.

9. A screening performance control method for an air flow microporous sieve according to claim 8, characterized in that, In the warm-up stage of the multi-stage training strategy, the learning rate of each iteration increases by 0.0001 compared with the previous iteration.

10. A method for controlling the screening performance of an air flow microporous sieve according to claim 1, characterized in that, It includes: Collect the intermediate data during the operation of the microporous sieve of the airflow to be measured. After the screening process ends, retrain the target screening performance prediction model based on the intermediate data to improve the stability of the target screening performance prediction model.

Images