Dynamic data intelligent control system and method

Abstract

The invention belongs to the technical field of intelligent equipment control, and discloses a dynamic data intelligent control system and method. Comprising the steps of collecting multi-dimensional operation state data of a moving object in real time; historical multi-dimensional operation state data are collected, and the entropy value of a moving object is calculated; preliminarily evaluating whether the moving object has resistance abnormity or not according to the entropy value of the moving object; if the initial resistance abnormity exists, extracting a dynamic change rate index of the moving object; based on the multi-dimensional operation state data of the moving object, the dynamic change rate index and a preset resistance evaluation model, an evaluation result is output, and the evaluation result comprises normal resistance or abnormal resistance; if the resistance is determined to be abnormal, generating a resistance regulation and control strategy through a swarm intelligence optimization algorithm according to the multi-dimensional operation state data of the moving object, and executing and feeding back the data; quantitative and accurate friction reduction of the moving object can be achieved, and the fault risk of the moving object is reduced.

Description

Technical Field

[0001] The present invention relates to the technical field of intelligent device control, and more specifically, to a dynamic data intelligent control system and method. Background Art

[0002] Moving objects, as key basic equipment in the industrial field, have extensive and important applications in industries such as mining, building materials, and chemical engineering. Their working efficiency and running stability are directly related to the quality and output of the production process. However, during long-term operation, moving objects are affected by factors such as vibration and friction, and the effectiveness of the anti-friction system becomes a key link to ensure the normal operation of the equipment. Traditional anti-friction methods mostly rely on manual experience, and many drawbacks are exposed in practical applications: First, there is an easy occurrence of lagging dripping. Whether the supply amount of the anti-friction medium oil is too much or too little, it will have an adverse impact on the service life and working efficiency of the equipment; Second, it is impossible to monitor the anti-friction quality and component wear status in real time, resulting in difficulty in quickly troubleshooting potential problems; Third, there is a lack of the ability to optimize the resistance-related parameters, making it difficult to improve the equipment efficiency.

[0003] Although the prior art can achieve the resistance regulation of moving objects, it does not elaborate on how to perform resistance regulation based on the collected multi-parameters. There is a lack of targeted resistance regulation methods, and it is difficult to effectively associate multi-parameters such as vibration and temperature with resistance-related parameters, resulting in the inability to effectively integrate and utilize the collected data, thereby affecting the real-time adjustment and accuracy of the resistance regulation strategy; In addition, the above-mentioned technology only briefly describes injecting a certain amount of grease into the bearing, but does not clarify the specific control logic and implementation method, lacking in-depth analysis and description of the anti-friction process. For example, how to determine the resistance-related parameters such as the amount of medium oil and the temperature of the medium, resulting in improper control of the resistance-related parameters, further increasing the wear and failure risks of moving objects, shortening the service life of moving objects, and ultimately affecting production efficiency and safety.

[0004] In view of this, the present invention proposes a dynamic data intelligent control system and method to solve the above problems. Summary of the Invention

[0005] To overcome the above-mentioned defects of the prior art and achieve the above object, the present invention provides the following technical solution: A dynamic data intelligent control method, including:

[0006] Real-time collecting multi-dimensional operating state data of a moving object;

[0007] Collecting historical multi-dimensional operating state data and calculating the entropy value of the moving object;

[0008] Based on the entropy value of the moving object, preliminarily evaluating whether there is abnormal resistance in the moving object;

[0009] If there is an initial resistance anomaly, extract the dynamic change rate index of the moving object;

[0010] Based on the multi-dimensional operating state data of the moving object, the dynamic change rate index, and the preset lubrication evaluation model resistance evaluation model, output the evaluation result, and the evaluation result includes normal resistance or abnormal resistance;

[0011] If it is determined that the resistance is abnormal, then according to the multi-dimensional operating state data of the moving object, generate a resistance regulation strategy through a swarm intelligence optimization algorithm, execute and feedback the data;

[0012] Dynamically adjust the resistance-related parameters according to the feedback data, and re-formulate the resistance regulation strategy. The feedback data is the multi-dimensional operating state data feedback after the resistance regulation is completed.

[0013] Furthermore, the method for generating a resistance regulation strategy through a swarm intelligence optimization algorithm includes:

[0014] Step 1: Construct n sets of resistance-related parameter sets, where n is an integer greater than 1. Each set contains a random combination of the medium temperature of the moving parts, the conveying pressure of the working medium, and the flow rate of the working medium. Set different parameter labels for each set of resistance-related parameter sets;

[0015] Step 2: Initialize the population, define that the individual positions in the one-dimensional search space correspond to the parameter labels, and set the initial iteration number to 0;

[0016] Step 3: Calculate the fitness function, predict the screening quality and screening efficiency through a deep belief network model, and determine the fitness function based on the weighted sum;

[0017] Step 4: Iteratively update the population, including:

[0018] Update the exploration individual position and the following individual position to generate an updated population;

[0019] Apply Gaussian perturbation to the individuals in the updated population to generate a perturbed population;

[0020] Calculate the dynamic boundary, reverse the solutions of the individuals in the updated population to generate a reverse solution population;

[0021] After merging the populations, screen the optimal set of resistance-related parameters based on the selection probability until the iteration threshold is reached, and use the set of resistance-related parameters corresponding to the individual with the maximum fitness as the resistance regulation strategy.

[0022] Further, the methods for obtaining the screening quality and screening efficiency are as follows: Use the resistance correlation parameter set corresponding to the test data and the individual position as the analysis data. The test data includes multi-dimensional operation state data and the dynamic change rate index of the moving object. Based on the analysis data and the preset quality prediction model and efficiency prediction model, predict the corresponding screening quality and screening efficiency. Both the quality prediction model and the efficiency prediction model are deep belief network models.

[0023] Further, the first individual in the population is the exploration individual, and the remaining individuals are the following individuals. The range of the parameter label and the range of the one-dimensional search space are both [1, n].

[0024] Further, the methods for setting the population size include:

[0025] Divide the population size range based on historical test data; Dynamically adjust the population size interval through the segmentation ratio and calculate the fitness comparison result;

[0026] Shrink the size interval according to the fitness comparison result until the interval width is less than the threshold, and take the mean of the maximum value and the minimum value in the updated population size range as the population size.

[0027] Further, the methods for updating the position of the exploration individual include:

[0028] Sort by fitness to determine the exploration individual, the second position, and the last position;

[0029] The position update of the exploration individual is based on the random weighted offset between the position of the individual with the highest fitness and the second position or the last position in the population. Determine the offset direction through a random threshold and combine with the random perturbation amplitude of the iteration times to dynamically adjust the offset and update the position of the exploration individual;

[0030] The position update of the following individual is generated by taking the average of its own current position and the position of the previous individual.

[0031] Further, the method for calculating the entropy value of the moving object includes:

[0032] Add the same data in the historical multi-dimensional operation state data and the multi-dimensional operation state data into an analysis set, that is, the data in the multi-dimensional operation state data in the analysis set corresponds one by one. Use the clustering algorithm to cluster the data in each analysis set to obtain each cluster corresponding to each analysis set. Obtain the number of data in each cluster, add up the number of data corresponding to each analysis set in turn to obtain the total number of data in each analysis set. Divide the number of data in each cluster by the total number of data in the corresponding analysis set to obtain the interval probability of each cluster. Calculate the set entropy value of each analysis set according to the interval probability of each cluster in each analysis set. Take the set entropy value of each analysis set as the entropy value of the moving object.

[0033] Further, the method for preliminarily evaluating whether there is an abnormal resistance of a moving object includes:

[0034] Obtain the historical entropy value, where the historical entropy value is the entropy value of the moving object calculated at a historical moment; mark the entropy value of the moving object calculated in real time as the real-time entropy value; add each set entropy value in the real-time entropy value to the corresponding set entropy value in the historical entropy value respectively for addition to obtain the total entropy value of the moving object corresponding to each data in the multi-dimensional operation state data; count the number of entropy values of the moving object in the historical entropy value and add one as the number of entropy values; divide each total vibration entropy value by the number of entropy values to obtain the entropy value mean corresponding to each data in the multi-dimensional operation state data; subtract the entropy value mean corresponding to each from the real-time entropy value and each set entropy value in the historical entropy value and then square to obtain the entropy value difference; add up the entropy value differences corresponding to each data in the multi-dimensional operation state data in sequence and then divide by the number of entropy values to obtain the variance corresponding to each data in the multi-dimensional operation state data; take the square root of each variance to obtain the entropy value standard deviation corresponding to each data in the multi-dimensional operation state data; add three times the corresponding entropy value standard deviation to the entropy value mean of each data in the multi-dimensional operation state data as the entropy value threshold.

[0035] Compare each set entropy value in the real-time entropy value with the corresponding entropy value threshold respectively.

[0036] If each set entropy value is less than or equal to the corresponding entropy value threshold, it is preliminarily evaluated that the moving object has no abnormal resistance.

[0037] If there is a set entropy value greater than the corresponding entropy value threshold, it is preliminarily evaluated that the moving object has abnormal resistance.

[0038] Further, the method for re-formulating the resistance control strategy includes:

[0039] Mark the multi-dimensional operation state data collected in real time as real-time data.

[0040] Calculate the entropy value of the moving object based on the feedback data, and evaluate whether there is abnormal resistance for the moving object; if there is no abnormal resistance, do not re-formulate the resistance control strategy; if there is still abnormal resistance, input the resistance-related parameter set and real-time data in the resistance control strategy into the trained parameter prediction model to predict the corresponding prediction data; the prediction data includes vibration frequency, vibration amplitude, bearing temperature, and screen temperature; among them, the frequency prediction model is used to predict the vibration frequency, the amplitude prediction model is used to predict the vibration amplitude, the shaft temperature prediction model is used to predict the bearing temperature, and the screen temperature prediction model is used to predict the screen temperature; the parameter prediction model includes a frequency prediction model, an amplitude prediction model, a shaft temperature prediction model, and a screen temperature prediction model; each model in the parameter prediction model is a deep belief network model; subtract each data in the prediction data from the corresponding data in the feedback data to obtain the unadjusted amount of each data; subtract each data in the prediction data from the corresponding data in the real-time data to obtain the amount to be adjusted of each data; divide the unadjusted amount corresponding to each data in the prediction data by the corresponding amount to be adjusted to calculate the unadjusted rate corresponding to each data in the prediction data; add 1 to each unadjusted rate as the adjustment rate corresponding to each data in the prediction data.

[0041] Multiply the vibration frequency, vibration amplitude, bearing temperature, and screen temperature in the real-time data by the corresponding adjustment rate to obtain the actual data; add the screen load in the feedback data to the actual data, extract the dynamic change rate index from the actual data, and mark it as the actual characteristic data; re-formulate the resistance control strategy according to the actual data and the actual characteristic data.

[0042] A dynamic data intelligent control system, implementing the described dynamic data intelligent control method, includes:

[0043] A data acquisition module, used to collect multi-dimensional operating state data of the moving object in real time;

[0044] A data analysis module, used to collect historical multi-dimensional operating state data and calculate the entropy value of the moving object;

[0045] A preliminary evaluation module, based on the entropy value of the moving object, preliminarily evaluate whether there is abnormal resistance for the moving object;

[0046] A feature extraction module, if there is preliminary abnormal resistance, extract the dynamic change rate index of the moving object;

[0047] An abnormal evaluation module, based on the multi-dimensional operating state data of the moving object, the dynamic change rate index, and a preset resistance evaluation model, outputs an evaluation result, and the evaluation result includes normal resistance or abnormal resistance;

[0048] The resistance regulation module, if it determines that the resistance is abnormal, generates a resistance regulation strategy according to the multi-dimensional operation state data of the moving object through a swarm intelligence optimization algorithm, executes it and feeds back the data;

[0049] The feedback optimization module dynamically adjusts the resistance-related parameters according to the feedback data, formulates a new resistance regulation strategy, and the feedback data is the multi-dimensional operation state data fed back after the resistance regulation is completed.

[0050] The technical effects and advantages of a dynamic data intelligent control system and method of the present invention:

[0051] By collecting multi-parameters to comprehensively monitor the operation status of the moving object, and using entropy value analysis to preliminarily evaluate the resistance state; extracting characteristic data and combining deep learning technology to further evaluate the resistance state, so as to timely grasp the abnormal resistance situation of the moving object; using a swarm intelligence optimization algorithm to formulate an accurate resistance regulation strategy and perform dynamic feedback adjustment; it can effectively integrate and utilize the collected multiple data, and perform accurate calculation of resistance-related parameters, realize quantitative and accurate friction reduction of the moving object, improve the operation efficiency and service life of the moving object, and reduce the failure risk of the moving object. Description of the Drawings

[0052] Figure 1 It is a schematic diagram of a dynamic data intelligent control system according to Embodiment 1 of the present invention;

[0053] Figure 2 It is a flowchart of the steps for formulating a resistance regulation strategy according to Embodiment 1 of the present invention;

[0054] Figure 3 It is a flowchart of the steps for iteratively updating the population according to Embodiment 1 of the present invention;

[0055] Figure 4 It is a flowchart of a dynamic data intelligent control method according to Embodiment 2 of the present invention. Detailed Embodiments

[0056] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0057] Embodiment 1

[0058] Please refer to Figure 1As shown in the figure, a dynamic data intelligent control system described in this embodiment includes a data acquisition module, a data analysis module, a preliminary evaluation module, a feature extraction module, an anomaly evaluation module, a resistance regulation module, and a feedback optimization module; each module is connected by wired and / or wireless means to achieve data transmission between modules;

[0059] The data acquisition module is used to collect multi-dimensional operating state data of a moving object in real time; the moving object includes, but is not limited to, vibrating screens, motors, compressors, etc., and the influencing factors of its resistance include, but are not limited to, the lubrication effect of medium oil (such as lubricating oil) during movement.

[0060] The multi-dimensional operating state data includes vibration frequency, vibration amplitude, bearing temperature, screen temperature, and screen load; among them, the vibration frequency is the number of vibrations of the screen in the moving object per unit time, which affects the screening efficiency and material handling capacity, and the vibration frequency is obtained by a vibration sensor installed on the screen; the vibration amplitude is the displacement range of the screen in the moving object during vibration, which affects the screening effect and material flow, and the vibration amplitude is obtained by a displacement sensor installed on the screen; the bearing temperature is the working temperature of the bearing of the moving object, and too high a temperature may cause the medium oil to fail and the bearing to be damaged, and the bearing temperature is obtained by a temperature sensor installed at the bearing housing; the screen temperature is the working temperature of the screen in the moving object, which affects the fluidity of the material and the screening effect, and the screen temperature is obtained by a temperature sensor installed on the screen; the screen load is the weight of the material applied to the screen, which affects the screening efficiency and the service life of the screen, and the screen load is obtained by a pressure sensor installed on the support structure in the moving object; the multi-dimensional operating state data is collected according to a preset sampling frequency, and the sampling frequency is preset by those skilled in the art according to the actual situation;

[0061] It should be noted that the reason for collecting multi-dimensional operating state data is that monitoring the vibration frequency can evaluate the operating state and workload of a moving object. An abnormal change in the vibration frequency leads to an increase in friction, so it is necessary to adjust the resistance control strategy; the vibration amplitude reflects the vibration intensity and stability of the moving object. A large vibration amplitude will cause uneven flow of the dielectric oil, so it is necessary to adjust the resistance control strategy; the bearing temperature is an important indicator for judging the resistance state. An excessively high bearing temperature usually indicates insufficient lubrication or a decline in the quality of the dielectric oil, so it is necessary to adjust the resistance control strategy; the screen temperature directly affects the fluidity and screening effect of the material. An excessively high screen temperature will cause material adhesion and blockage. As the screen temperature increases, the performance of the dielectric oil decreases, so it is necessary to adjust the resistance control strategy; the screen load directly affects the screening efficiency and the stability of the moving object. When the screen load increases, it is necessary to increase the supply of the dielectric oil to cope with greater frictional force and wear risk; by collecting vibration characteristic data, it helps to optimize the resistance control strategy subsequently, adjust the resistance-related parameters, achieve a more refined and intelligent resistance control, and improve the operating efficiency and reliability of the moving object;

[0062] A data analysis module for collecting historical multi-dimensional operating state data and calculating the entropy value of the moving object;

[0063] The historical multi-dimensional operating state data is the multi-dimensional operating state data collected at historical moments;

[0064] The methods for calculating the entropy value of the moving object include:

[0065] Adding the same data in the historical multi-dimensional operating state data and the multi-dimensional operating state data to an analysis set, that is, the data in the analysis set corresponds one by one to the data in the multi-dimensional operating state data; using a clustering algorithm (such as K-Means clustering, hierarchical clustering, DBSCAN, etc.) to cluster the data in each analysis set to obtain each cluster corresponding to each analysis set; obtaining the number of data in each cluster, adding up the number of data in each cluster corresponding to each analysis set in turn to obtain the total number of data in each analysis set; dividing the number of data in each cluster by the total number of data in the corresponding analysis set to obtain the interval probability of each cluster; calculating the set entropy value of each analysis set according to the interval probability of each cluster in each analysis set; taking the set entropy value of each analysis set as the entropy value of the moving object;

[0066] The expression of the set entropy value is In the formula, H is the set entropy value, p j is the interval probability of the j-th cluster, j ∈ [1, J], and J is the total number of clusters;

[0067] A preliminary evaluation module for preliminarily evaluating whether there is an abnormal resistance in the moving object according to the entropy value of the moving object;

[0068] Methods for initially assessing whether a moving object has resistance anomalies include:

[0069] Obtain historical entropy values, which are entropy values of moving objects calculated at historical moments; mark the entropy values of moving objects calculated in real time as real-time entropy values; add each set entropy value in the real-time entropy value to the corresponding set entropy value in the historical entropy value, and obtain the total entropy value of the moving object corresponding to each data in the multidimensional running state data; count the number of entropy values of moving objects in the historical entropy value and add one as the number of entropy values; divide each total entropy value of vibration by the number of entropy values, and obtain the mean entropy value corresponding to each data in the multidimensional running state data; subtract the corresponding mean entropy value from the real-time entropy value and each set entropy value in the historical entropy value, and then square them to obtain the entropy value difference; add the entropy value difference corresponding to each data in the multidimensional running state data in sequence, and then divide it by the number of entropy values to obtain the variance corresponding to each data in the multidimensional running state data; square root each variance to obtain the entropy value standard deviation corresponding to each data in the multidimensional running state data; add the entropy value mean of each data in the multidimensional running state data plus three times the corresponding entropy value standard deviation as the entropy value threshold;

[0070] Compare each set entropy value in the real-time entropy value with the corresponding entropy threshold;

[0071] If each set entropy value is less than or equal to the corresponding entropy threshold, it is preliminarily assessed that the moving object does not have resistance anomaly;

[0072] If there is a set entropy value greater than the corresponding entropy threshold, it is preliminarily assessed that the moving object has resistance anomaly;

[0073] The feature extraction module extracts the dynamic change rate index of the moving object if there is a preliminary resistance anomaly;

[0074] Methods for extracting the dynamic change rate index of a moving object include:

[0075] Obtain historical multidimensional operating status data, which is the multidimensional operating status data of the moving object collected last time; subtract the vibration frequency, vibration amplitude, bearing temperature and screen temperature in the historical multidimensional operating status data from the vibration frequency, vibration amplitude, bearing temperature and screen temperature in the multidimensional operating status data of the moving object to obtain the vibration frequency change, vibration amplitude change, bearing temperature change and screen temperature change; divide the vibration frequency change, vibration amplitude change, bearing temperature change and screen temperature change by the sampling frequency to obtain the vibration frequency change rate, vibration amplitude change rate, bearing temperature change rate and screen temperature change rate, and use them as the dynamic change rate indicators of the moving object.

[0076] Anomaly assessment module, based on multi-dimensional operating state data of a moving object, dynamic change rate indicators, and a preset resistance assessment model, outputs an assessment result, where the assessment result includes normal resistance or abnormal resistance;

[0077] The method for assessing whether there is abnormal resistance in a moving object includes:

[0078] Taking the multi-dimensional operating state data and dynamic change rate indicators as test data, inputting the test data into a trained resistance assessment model, outputting an assessment label, obtaining the corresponding assessment result according to the assessment label, and assessing whether there is abnormal resistance in the moving object according to the assessment result; the assessment label is the digital label corresponding to the assessment result, the assessment results include normal and abnormal, and the digital labels corresponding to different assessment results are different. Exemplarily, the digital label for normal is set to 1, and the digital label for abnormal is set to 2.

[0079] The specific training process of the resistance assessment model includes:

[0080] Pre-set corresponding judgment results for a group of test data, where a is an integer greater than 1. The judgment results corresponding to the test data are collected by those skilled in the art during the historical resistance regulation process of moving objects. Those skilled in the art analyze a group of different test data in sequence according to actual experience, judge whether the corresponding moving object has abnormal resistance, and set corresponding judgment results for a group of different test data in sequence.

[0081] Convert the test data and the corresponding assessment label into a corresponding group of feature vectors; use each group of feature vectors as the input of the resistance assessment model. The resistance assessment model outputs a group of predicted assessment labels corresponding to each group of test data, and uses the actual assessment label corresponding to each group of test data as the prediction target. The actual assessment label is the assessment label pre-collected corresponding to the test data; use the mean absolute percentage error MAPE to evaluate the model accuracy for the prediction result. When the calculated MAPE is less than the preset MAPE, the resistance assessment model training is completed; where the calculation formula of MAPE is where η b is the actual assessment label corresponding to the b-th group of test data, is the predicted assessment label corresponding to the b-th group of test data, b is the group number of the feature vector corresponding to the test data, b ∈ [1, a]; generate a resistance assessment model for predicting the assessment label according to the test data; where the resistance assessment model is a deep belief network model, and the preset MAPE is pre-set by those skilled in the art according to the required accuracy of the resistance assessment model.

[0082] It should be noted that the reason for first calculating the entropy value and entropy threshold of the moving object to conduct a preliminary assessment of the abnormal resistance of the moving object, and then calculating the dynamic change rate index and combining deep learning technology for the final assessment of the abnormal resistance of the moving object is that the entropy value can quickly reflect the complexity and uncertainty of the data, helping to preliminarily judge whether there is abnormal resistance in the moving object. If the entropy value increases, it may indicate a poor resistance state; the change rate can capture the trend of data change over time, helping to identify whether the anti-friction supply is appropriate; and the change rate reflects real-time and subtle changes, being more sensitive to the judgment of abnormalities; deep learning technology can process high-dimensional data, automatically extract features, reducing the need for manual feature engineering; and it is good at capturing non-linear relationships in the data, being able to better identify the complex connections between the resistance state and the characteristics of the moving object; in addition, the deep learning model can be adaptively adjusted through training, and with the input of more data, the accuracy of the model can be continuously improved to adapt to new working conditions or environmental changes.

[0083] The resistance regulation module, if it determines that there is abnormal resistance, generates a resistance regulation strategy according to the multi-dimensional operating state data of the moving object through a swarm intelligence optimization algorithm, executes and feeds back the data;

[0084] Such as Figure 2 shown, the method for generating a resistance regulation strategy through a swarm intelligence optimization algorithm includes:

[0085] Step 1: Construct n sets of resistance correlation parameters, where n is an integer greater than 1, and each set contains a random combination of the medium temperature of the moving parts, the conveying pressure of the working medium, and the flow rate of the working medium. Set different parameter labels for each set of resistance correlation parameters;

[0086] Step 2: Initialize the population, define the correspondence between the individual positions in the one-dimensional search space and the parameter labels, and set the initial iteration number to 0;

[0087] Step 3: Calculate the fitness function, predict the screening quality and screening efficiency through a deep belief network model, and determine the fitness function based on the weighted sum;

[0088] Refer to Figure 3 , Step 4: Iteratively update the population, including:

[0089] Update the exploration individual position and the following individual position to generate an updated population;

[0090] Apply Gaussian perturbation to the individuals in the updated population to generate a perturbed population;

[0091] Calculate the dynamic boundary, generate the reverse solution population from the reverse solutions of the individuals in the updated population;

[0092] After merging the populations, the optimal set of drag correlation parameters is selected based on the selection probability until the iteration threshold is reached. The set of drag correlation parameters corresponding to the individual with the maximum fitness is used as the drag control strategy.

[0093] A set of drag correlation parameters corresponds to a group of drag correlation parameters. The medium temperature of the moving part is the temperature of the medium oil (such as medium oil) added to the moving object. The conveying pressure of the working medium is the pressure for conveying the medium oil by the friction reduction system in the moving object. The flow rate of the working medium is the amount of medium oil passing through the friction reduction system in the moving object per unit time. The method for constructing the set of drag correlation parameters is as follows: Obtain the range of drag correlation parameters, which includes the range of medium temperature of the moving part, the range of conveying pressure of the working medium, and the range of flow rate of the working medium. The range of drag correlation parameters is obtained by those skilled in the art according to the technical parameters of the moving object. Randomly select a value from each range in the range of drag correlation parameters to construct a set of drag correlation parameters. A total of n sets of drag correlation parameters are constructed, and the n sets of drag correlation parameters are all different.

[0094] The method for setting the population size includes:

[0095] Divide the population size range based on historical test data; Dynamically adjust the population size interval through the segmentation ratio and calculate the fitness comparison result; Shrink the size interval according to the fitness comparison result until the interval width is less than the threshold, and take the mean of the maximum and minimum values in the updated population size range as the population size. Specifically as follows:

[0096] Step 201: Collect c sets of historical test data, where the historical test data is the test data obtained at historical moments, and c is an integer greater than 1.

[0097] Step 202: Set different population sizes for the c sets of historical test data, and the range of the population size is [1, c]; Set the segmentation ratio L and the number of iterations L'. In this embodiment, it is preferably L = 0.618, and the number of iterations L' is set by those skilled in the art according to the actual situation.

[0098] Step 203: According to the segmentation ratio h, divide the population size into two new points ζ1 and ζ2, where ζ1 = 1 + (1 - L)(c - 1), and ζ2 = 1 + L(c - 1).

[0099] Step 204: Update the population size according to ζ1 and ζ2 respectively, perform iteration according to the updated population size, and obtain the fitness corresponding to the position of the individual with the highest fitness; Mark the fitness obtained according to ζ1 as the first fitness, and mark the fitness obtained according to ζ2 as the second fitness.

[0100] Step 205: If the first fitness is greater than or equal to the second fitness, the minimum value of the updated population size range is the minimum value of the population size range before update, and the maximum value is the point corresponding to the second fitness; if the first fitness is less than the second fitness, the minimum value of the updated population size range is the point corresponding to the first fitness, and the maximum value is the maximum value of the population size range before update corresponding to the second fitness; Exemplarily, after the population size range [1, c] is updated by points δ1 and δ2, if the first fitness is greater than or equal to the second fitness, the updated population size range is [1, δ2], and if the first fitness is less than the second fitness, the updated population size range is [δ1, c];

[0101] Step 206: Preset a width threshold, which is preset by those skilled in the art according to the actual situation. Subtract the minimum value from the maximum value in the updated population size range to obtain the interval width; compare the interval width with the width threshold; if the interval width is greater than or equal to the width threshold, re-divide the updated population size range into two new points δ1 and δ2 according to the segmentation ratio h, and return to Step 204; if the interval width is less than the width threshold, take the average value of the maximum value and the minimum value in the updated population size range as the population size Z.

[0102] The first individual in the population is the exploration individual, and the remaining individuals are the following individuals; the parameter label range and the range of the one-dimensional search space are both [1, n].

[0103] The expression of the fitness function is: f = ω1×sz + ω2×sx; where f is the fitness, sz is the screening quality, sx is the screening efficiency, and ω1 and ω2 are both preset weight coefficients; the screening quality is the material quality control ability in the classification and screening process of moving objects, and the screening efficiency is the amount of material passing through the sieve per unit time in the classification and screening process of moving objects; the specific values of the weight coefficients in the formula can be set according to the actual situation. The weight coefficients reflect the influence degrees of the screening quality and the screening efficiency. Those skilled in the art can preset the corresponding weight coefficients according to the actual influence degrees of the screening quality and the screening efficiency to accurately evaluate the comprehensive screening performance of moving objects.

[0104] The methods for obtaining the screening quality and the screening efficiency are: taking the set of resistance-related parameters corresponding to the test data and the parameter labels of the individual positions as the analysis data, and inputting the analysis data into the trained quality prediction model and efficiency prediction model respectively to predict the corresponding screening quality and screening efficiency; the specific training processes of the quality prediction model and the efficiency prediction model are the same as the specific training process of the resistance evaluation model, and both are deep belief network models;

[0105] The method for updating the position of the exploration individual includes:

[0106] Determine the exploration individual, the second position, and the last position according to the fitness sorting;

[0107] The position update of the exploration individual is based on the random weighted offset between the position of the individual with the highest fitness and the second position or the last position in the population. The offset direction is determined by a random threshold, and combined with the random perturbation amplitude of the iteration number, the offset is dynamically adjusted to update the position of the exploration individual. Specifically as follows:

[0108] Sort the fitness of all individuals in the population from largest to smallest. The position corresponding to the fitness after the fitness corresponding to the position of the individual with the highest fitness is used as the second position, and the position corresponding to the last fitness is used as the last position;

[0109]

[0110] In the formula, X new is the position of the exploration individual, F is the position of the individual with the highest fitness, g(G) is a random value in the g-distribution with the iteration number G, W' is the second position, W” is the last position, is a random number in [﹣0.5,0.5], is a random number in [0,1].

[0111] By randomly selecting the second or the last position, the algorithm is prevented from falling into local optimum; combining the randomness of the g(G) distribution and the iteration number correlation to balance the breadth and depth of the search; the random coefficient limits the offset amplitude ([﹣0.5,0.5]) to ensure that the position update is within a reasonable range and prevent excessive deviation from the current optimal region.

[0112] The method for updating the position of the follower individual includes:

[0113] The new position of the follower individual is generated by taking the average of its own current position and the position of the previous individual, realizing the gradual convergence of the positions among individuals in the population. The individual differences are balanced through the local cooperation mechanism to avoid the population converging to a single region prematurely, while maintaining the population stability to support continuous optimization.

[0114] The method for generating the perturbed population includes:

[0115] Set Gaussian noise and add Gaussian noise to each individual in the updated population to generate new individuals. All new individuals form the perturbed population; the expression of the new individual is: In the formula, is the position of the i-th new individual, X iis the position of the i-th individual, N(0,σ) is Gaussian noise with a mean of 0 and a standard deviation of σ, and i ∈ [1, Z]; the standard deviation of the Gaussian noise is set by those skilled in the art according to the required degree of perturbation; controllable randomness is introduced to enhance the diversity of the population, help the algorithm jump out of the local optimal trap, and at the same time ensure the search efficiency through the standard deviation constraint of the perturbation amplitude (such as small-range perturbation).

[0116] The method for generating the reverse solution population includes:

[0117] The dynamic boundary is [θ1, θ2], where θ1 is the parameter label corresponding to the smallest value of the individual position in the updated population, and θ2 is the parameter label corresponding to the largest value of the individual position in the updated population, 1 ≤ θ1 < θ2 ≤ n; the dynamic boundary can make up for the defect that the fixed boundary cannot preserve the search experience, which is beneficial to shortening the search time.

[0118] The method for generating the reverse solution includes:

[0119]

[0120] In the formula, is the reverse solution of the i-th individual, is a random number in [0, 1].

[0121] If the generated reverse solution is greater than or less than the dynamic boundary, it is marked as an out-of-bounds solution, and a value within the dynamic boundary is randomly generated using a random number function, and the randomly generated value is assigned to the out-of-bounds solution;

[0122] All reverse solutions form a reverse solution population. The dynamic boundary adaptation adjusts the solution range in real time according to the current population state to avoid the limitations of the fixed boundary; uses the reverse operation to generate symmetric solutions to explore potential optimal regions and break through local optima; forcibly resets out-of-bounds solutions to ensure the rationality of the solutions while maintaining the search efficiency.

[0123] It should be noted that the reason for formulating the drag regulation strategy using the above method is as follows:

[0124] 1. Systematics and flexibility: By constructing a set of drag-related parameters and dynamically updating, different parameter combinations can be systematically explored, improving the flexibility of the strategy;

[0125] 2. Adaptive optimization: The introduction of the fitness function ensures that the algorithm can adjust the strategy according to real-time feedback and optimize the performance;

[0126] 3. Diversity maintenance: Using methods such as Gaussian perturbation and reverse solutions can maintain the diversity of the population during the optimization process and avoid falling into local optimal solutions;

[0127] 4. Iterative improvement: By setting the number of iterations and dynamic boundaries, ensure that the strategy can be continuously improved to adapt to changing operating conditions;

[0128] 5. Data-based decision-making: Combine real-time data feedback and fitness evaluation to make the finally formulated resistance regulation strategy more scientific and effective.

[0129] The feedback optimization module dynamically adjusts the resistance correlation parameters according to the feedback data and re-formulates the resistance regulation strategy. The feedback data is the multi-dimensional operating state data feedback after the resistance regulation is completed.

[0130] The method for re-formulating the resistance regulation strategy includes:

[0131] Mark the multi-dimensional operating state data of the moving object feedback after the resistance regulation is completed as feedback data, and mark the multi-dimensional operating state data of the moving object collected in real time as real-time data; calculate the entropy value of the moving object according to the feedback data and evaluate whether there is abnormal resistance in the moving object; if there is no abnormal resistance, do not re-formulate the resistance regulation strategy; if there is still abnormal resistance, input the resistance correlation parameter set in the resistance regulation strategy and the real-time data into the trained parameter prediction model to predict the corresponding prediction data; the parameter prediction model includes a frequency prediction model, an amplitude prediction model, a shaft temperature prediction model, and a screen temperature prediction model; the prediction data includes vibration frequency, vibration amplitude, bearing temperature, and screen temperature; among them, the frequency prediction model is used to predict the vibration frequency, the amplitude prediction model is used to predict the vibration amplitude, the shaft temperature prediction model is used to predict the bearing temperature, and the screen temperature prediction model is used to predict the screen temperature; the specific training process of each model in the parameter prediction model is the same as the specific training process of the resistance evaluation model, and both are deep belief network models;

[0132] Subtract each data in the prediction data from the corresponding data in the feedback data to obtain the unadjusted amount of each data; subtract each data in the prediction data from the corresponding data in the real-time data to obtain the amount to be adjusted of each data; divide the unadjusted amount corresponding to each data in the prediction data by the corresponding amount to be adjusted to calculate the unadjusted rate corresponding to each data in the prediction data; add 1 to each unadjusted rate as the adjustment rate corresponding to each data in the prediction data;

[0133] Multiply the vibration frequency, vibration amplitude, bearing temperature, and screen temperature in the real-time data by the corresponding adjustment rate to obtain the actual data; add the screen load in the feedback data to the actual data, extract the dynamic change rate index of the moving object from the actual data, and mark it as the actual characteristic data; re-formulate the resistance regulation strategy according to the actual data and the actual characteristic data;

[0134] It should be noted that the purpose of re - formulating the resistance regulation strategy is to accurately identify the control elements that need to be optimized. For example, if the adjustment effect of some parameters is not good, they need to be emphasized and optimized in the new strategy; to enhance the pertinence of the control strategy, and customize a more efficient new strategy according to the difference between the existing control effect and the current demand; to achieve the dynamic optimization of the control strategy, and through closed - loop iteration, make the resistance regulation strategy consistent with the dynamic performance of the equipment, thereby improving the quality of resistance regulation.

[0135] In this embodiment, the operating conditions of the moving object are comprehensively monitored through multi - parameter acquisition, and the resistance state is initially evaluated by entropy value analysis; characteristic data is extracted and combined with deep learning technology to further evaluate the resistance state in order to timely grasp the abnormal resistance situation of the moving object; a precise resistance regulation strategy is formulated using a swarm intelligence optimization algorithm and dynamic feedback adjustment is performed; it can effectively integrate and utilize the multiple collected data, and perform accurate calculation of resistance - related parameters, achieve quantitative and precise friction reduction of the moving object, improve the operating efficiency and service life of the moving object, reduce the failure risk of the moving object, and provide technical support for improving production capacity and ensuring safety.

[0136] Embodiment 2

[0137] Please refer to Figure 4 As shown, for the parts not described in detail in this embodiment, refer to the description content of Embodiment 1. A dynamic data intelligent control method is provided, and the method includes:

[0138] Real - time collect multi - dimensional operating state data of the moving object;

[0139] Collect historical multi - dimensional operating state data and calculate the entropy value of the moving object;

[0140] Based on the entropy value of the moving object, preliminarily evaluate whether there is abnormal resistance in the moving object;

[0141] If there is preliminary abnormal resistance, extract the dynamic change rate index of the moving object;

[0142] Based on the multi - dimensional operating state data, dynamic change rate index of the moving object and a preset resistance evaluation model, output an evaluation result, and the evaluation result includes normal resistance or abnormal resistance;

[0143] If it is determined that the resistance is abnormal, then generate a resistance regulation strategy according to the multi - dimensional operating state data of the moving object through a swarm intelligence optimization algorithm, execute and feedback data;

[0144] Dynamically adjust the resistance - related parameters according to the feedback data, re - formulate the resistance regulation strategy, and the feedback data is the multi - dimensional operating state data feedback after the resistance regulation is completed.

[0145] The parameters in the above formula are all dimensionless parameters.

[0146] As described above, it is only the specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present invention can easily think of changes or substitutions, which should all be covered within the protection scope of the present invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims described above.

[0147] Finally: The above description is only the preferred embodiment of the present invention and is not used to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention shall all be included within the protection scope of the present invention.

Claims

1. A dynamic data intelligent control method, characterized in that Including: Collecting multi-dimensional operation state data of a moving object in real time; Collecting historical multi-dimensional operation state data and calculating the entropy value of the moving object; Based on the entropy value of the moving object, preliminarily evaluating whether there is abnormal resistance in the moving object; If there is preliminary abnormal resistance, extracting the dynamic change rate index of the moving object; Based on the multi-dimensional operation state data, dynamic change rate index of the moving object and a preset resistance evaluation model, outputting an evaluation result, where the evaluation result includes normal resistance or abnormal resistance; If it is determined that the resistance is abnormal, then according to the multi-dimensional operation state data of the moving object, generating a resistance regulation strategy through a swarm intelligence optimization algorithm, executing and feeding back data; Dynamically adjusting the resistance-related parameters according to the feedback data, re-formulating the resistance regulation strategy, and the feedback data is the multi-dimensional operation state data fed back after the resistance regulation is completed.

2. The dynamic data intelligent control method according to claim 1, wherein The method for generating a resistance regulation strategy through a swarm intelligence optimization algorithm includes: Step 1: Constructing n sets of resistance-related parameter sets, where n is an integer greater than 1, and each set contains a random combination of the medium temperature of the moving part, the conveying pressure of the working medium, and the flow rate of the working medium, and setting different parameter labels for each set of resistance-related parameter sets; Step 2: Initializing the population, defining the correspondence between the individual positions in the one-dimensional search space and the parameter labels, and setting the initial number of iterations to 0; Step 3: Calculating the fitness function, predicting the screening quality and screening efficiency through a deep belief network model, and determining the fitness function based on the weighted sum; Step 4: Iteratively updating the population, including: Updating the exploration individual position and the following individual position to generate an updated population; Applying Gaussian perturbation to the individuals in the updated population to generate a perturbed population; Calculating the dynamic boundary, generating the reverse solution population from the reverse solutions of the individuals in the updated population; After merging the populations, screening the optimal resistance-related parameter set based on the selection probability until the iteration threshold is reached, and taking the resistance-related parameter set corresponding to the individual with the maximum fitness as the resistance regulation strategy.

3. A dynamic data intelligent control method according to claim 2, characterized in that, The method for obtaining the screening quality and screening efficiency is: using the test data and the resistance-related parameter set corresponding to the individual position as the analysis data, where the test data includes multi-dimensional operation state data and the dynamic change rate index of the moving object; predicting the corresponding screening quality and screening efficiency based on the analysis data and the preset quality prediction model and efficiency prediction model; both the quality prediction model and the efficiency prediction model are deep belief network models.

4. A dynamic data intelligent control method according to claim 2, characterized in that, The first individual in the population is the exploration individual, and the remaining individuals are the following individuals; the range of the parameter labels and the range of the one-dimensional search space are both [1, n].

5. A dynamic data intelligent control method according to claim 2, characterized in that, The method for setting the population size includes: Dividing the population size range based on historical test data; dynamically adjusting the population size interval through the segmentation ratio and calculating the fitness comparison result; Shrinking the size interval according to the fitness comparison result until the interval width is less than the threshold, and taking the mean of the maximum value and the minimum value in the updated population size range as the population size.

6. The dynamic data intelligent control method according to claim 2, characterized in that The method for updating the exploration individual position includes: Sorting by fitness to determine the exploration individual, the second position, and the last position; The exploration individual's position update is based on the position of the individual with the highest fitness and the random weighted offset between the second position or the last position in the population. The offset direction is determined by a random threshold, and combined with the random perturbation amplitude of the number of iterations, the offset is dynamically adjusted to update the exploration individual's position; The position update of the follower individual is generated by taking the average of its own current position and the position of the previous individual.

7. A dynamic data intelligent control method according to claim 6, characterized in that, The method for calculating the entropy value of the moving object includes: Adding the historical multi-dimensional operating state data and the same data in the multi-dimensional operating state data to an analysis set, that is, the data in the multi-dimensional operating state data of the analysis set corresponds one by one; using a clustering algorithm to cluster the data in each analysis set to obtain each cluster corresponding to each analysis set; obtaining the number of data in each cluster, adding the number of data corresponding to each analysis set in sequence to obtain the total number of data in each analysis set; dividing the number of data in each cluster by the total number of data in the corresponding analysis set to obtain the interval probability of each cluster; calculating the set entropy value of each analysis set according to the interval probability of each cluster in each analysis set; taking the set entropy value of each analysis set as the entropy value of the moving object.

8. A dynamic data intelligent control method according to claim 7, characterized in that The method for preliminarily evaluating whether there is abnormal resistance in the moving object includes: Obtaining the historical entropy value, which is the entropy value of the moving object calculated at the historical moment; marking the entropy value of the moving object calculated in real time as the real-time entropy value; adding each set entropy value in the real-time entropy value to the corresponding set entropy value in the historical entropy value respectively to obtain the total entropy value of the moving object corresponding to each data in the multi-dimensional operating state data; counting the number of entropy values of the moving object in the historical entropy value and adding one as the number of entropy values; dividing each total vibration entropy value by the number of entropy values to obtain the entropy value mean corresponding to each data in the multi-dimensional operating state data; subtracting the entropy value mean from each set entropy value in the real-time entropy value and the historical entropy value respectively and then squaring to obtain the entropy value difference; adding the entropy value differences corresponding to each data in the multi-dimensional operating state data in sequence and then dividing by the number of entropy values to obtain the variance corresponding to each data in the multi-dimensional operating state data; taking the square root of each variance to obtain the entropy value standard deviation corresponding to each data in the multi-dimensional operating state data; adding three times the corresponding entropy value standard deviation to the entropy value mean of each data in the multi-dimensional operating state data as the entropy value threshold; Comparing each set entropy value in the real-time entropy value with the corresponding entropy value threshold respectively; If each set entropy value is less than or equal to the corresponding entropy value threshold, it is preliminarily evaluated that there is no abnormal resistance in the moving object; If there is a set entropy value greater than the corresponding entropy value threshold, it is preliminarily evaluated that there is abnormal resistance in the moving object.

9. A dynamic data intelligent control method according to claim 8, characterized in that, The method for re-formulating the resistance regulation strategy includes: Marking the multi-dimensional operating state data collected in real time as real-time data; Calculate the entropy value of the moving object based on the feedback data, and evaluate whether there is an abnormal resistance for the moving object; if there is no abnormal resistance, do not re-formulate the resistance control strategy; if there is still an abnormal resistance, input the resistance-related parameter set and real-time data in the resistance control strategy into the trained parameter prediction model to predict the corresponding prediction data; the prediction data includes vibration frequency, vibration amplitude, bearing temperature, and screen temperature; among them, the frequency prediction model is used to predict the vibration frequency, the amplitude prediction model is used to predict the vibration amplitude, the shaft temperature prediction model is used to predict the bearing temperature, and the screen temperature prediction model is used to predict the screen temperature; the parameter prediction model includes a frequency prediction model, an amplitude prediction model, a shaft temperature prediction model, and a screen temperature prediction model; each model in the parameter prediction model is a deep belief network model; subtract each data in the prediction data from the corresponding data in the feedback data to obtain the unadjusted amount of each data; subtract each data in the prediction data from the corresponding data in the real-time data to obtain the amount to be adjusted for each data; divide the unadjusted amount corresponding to each data in the prediction data by the corresponding amount to be adjusted to calculate the unadjusted rate corresponding to each data in the prediction data; add 1 to each unadjusted rate as the adjustment rate corresponding to each data in the prediction data. Multiply the vibration frequency, vibration amplitude, bearing temperature, and screen temperature in the real-time data by the corresponding adjustment rate to obtain the actual data; add the screen load in the feedback data to the actual data, extract the dynamic change rate index from the actual data, and label it as the actual feature data; re-formulate the resistance control strategy according to the actual data and the actual feature data.

10. A dynamic data intelligent control system that implements the dynamic data intelligent control method described in any one of claims 1-9, characterized in that, Including: A data acquisition module for real-time collecting multi-dimensional operation state data of the moving object; A data analysis module for collecting historical multi-dimensional operation state data and calculating the entropy value of the moving object; A preliminary evaluation module for preliminarily evaluating whether there is an abnormal resistance for the moving object according to the entropy value of the moving object; A feature extraction module for extracting the dynamic change rate index of the moving object if there is a preliminary abnormal resistance; An abnormal evaluation module for outputting an evaluation result based on the multi-dimensional operation state data, the dynamic change rate index of the moving object, and a preset resistance evaluation model, and the evaluation result includes normal resistance or abnormal resistance; A resistance control module for generating a resistance control strategy through a swarm intelligence optimization algorithm according to the multi-dimensional operation state data of the moving object if it is determined that there is an abnormal resistance, executing and feeding back data; A feedback optimization module for dynamically adjusting the resistance-related parameters according to the feedback data, re-formulating the resistance control strategy, and the feedback data is the multi-dimensional operation state data fed back after the resistance control is completed.

Images