Method, system and equipment for regulating and controlling dispersion effect of anti-foaming agent in lubricating oil rehabilitation process

Abstract

The invention discloses a method, a system and equipment for regulating and controlling the dispersion effect of an anti-foaming agent in a lubricating oil rehabilitation process, and relates to the technical field of lubricating oil repair, and the method comprises the following steps: inputting current homogeneous dispersion parameters adopted when the anti-foaming agent is dispersed in lubricating oil into a dispersion effect prediction model, the predicted homogeneous dispersion efficiency and the predicted weighted particle number are obtained; wherein the current homogenizing and dispersing parameters comprise the rotor speed, the homogenizing time and the additive amount; the dispersion effect prediction model is trained and determined based on a machine learning algorithm; and in combination with a multi-target genetic algorithm and a balance strategy, according to the predicted homogeneous dispersion efficiency and the predicted weighted particle number, reversely regulating and optimizing the current homogeneous dispersion parameter so as to determine the optimal homogeneous dispersion parameter. According to the application, the dispersion effect of the anti-foaming agent in the lubricating oil can be intelligently, accurately and quantitatively regulated and controlled.

Description

A method, system, and equipment for regulating the dispersion effect of antifoaming agents during lubricant recovery process. Technical Field

[0001] This application relates to the field of lubricating oil repair technology, and in particular to a method, system and equipment for regulating the dispersion effect of antifoaming agents during the lubricating oil recovery process. Background Technology

[0002] Adding restorative additives to lubricating oil is a common method to repair the performance indicators of lubricating oil, such as foam characteristics, demulsification properties, and antioxidant properties. There are currently three ways in the industry to achieve the restoration of lubricating oil by additives: (1) Adding antifoaming agents, demulsifiers and other additives directly to the mixing tank and restoring the performance of lubricating oil by stirring. This method has low investment, quick results, simple operation and no danger. It has no adverse reaction to the oil originally used in the tank and will delay and stabilize the service life of the original additives. This method may have the dual risks of uneven dispersion and secondary pollution, and requires the configuration of large mixing equipment in the factory, with high initial construction and later operation costs. In particular, additives such as silicone antifoaming agents have extremely poor oil solubility. If the stirring intensity is insufficient or the manual feeding accuracy is not high, it is very easy for them to fail to disperse fully in the oil. This leads to two problems: first, the additives fail to perform their function, and the key properties of the lubricating oil (such as anti-foaming properties) remain low; second, the insufficiently dispersed additive droplets will settle after standing, becoming a new source of oil contamination, resulting in a decrease in system cleanliness and excessive contamination.

[0003] (2) Additives are atomized into lubricating oil using atomizing nozzles and sprayed into the oil in a counter-current manner for pipeline blending. This method is widely used in the preparation and blending processes of lubricating oil. However, this method suffers from high investment costs and process limitations: its implementation relies on high-precision flow meters, proportional valves, and highly automated control systems, resulting in significant initial investment and maintenance costs. Furthermore, this method places high demands on the compatibility of the additives themselves. For components with poor oil solubility, such as silicone antifoaming agents, it is often difficult to achieve rapid and uniform fusion with the base oil under limited residence time and mixing intensity in the pipeline, which can easily lead to uneven blending and affect the final performance.

[0004] (3) The first two methods are for new lubricating oil, while the third method is for treating oil in use. First, the additive package is mixed well and diluted with lubricating oil from a thinner station to form a recovery additive package mother liquor. The recovery additive package mother liquor is then directly added to the lubricating oil in use. However, it has inherent defects such as uneven mixing and precipitation. The mixing uniformity is extremely poor. If the unevenly dispersed additive concentrate is directly added to the defective oil, it is easy to cause local over-concentration or under-concentration, which will produce precipitation or flocculents in the lubricating oil. Summary of the Invention

[0005] The purpose of this application is to provide a method, system, and device for regulating the dispersion effect of antifoaming agents during the lubricating oil recovery process, which can intelligently and precisely quantify the dispersion effect of antifoaming agents in lubricating oil.

[0006] To achieve the above objectives, this application provides the following solution: Firstly, this application provides a method for regulating the dispersion effect of an antifoaming agent during the lubricating oil recovery process, comprising: obtaining the current homogeneous dispersion parameters used when the antifoaming agent is dispersed in the lubricating oil; the current homogeneous dispersion parameters include rotor speed, homogenization time, and dosage; inputting the current homogeneous dispersion parameters into a dispersion effect prediction model to obtain a predicted homogeneous dispersion efficiency and a predicted weighted particle number; wherein the dispersion effect prediction model is determined based on training using a machine learning algorithm; combining a multi-objective genetic algorithm and a balancing strategy, based on the predicted homogeneous dispersion efficiency and the predicted weighted particle number, the current homogeneous dispersion parameters are reverse-regulated and optimized to determine the optimal homogeneous dispersion parameters.

[0007] Secondly, this application provides a system for regulating the dispersion effect of an antifoaming agent during the recovery process of lubricating oil, comprising: a parameter acquisition module for acquiring the current homogeneous dispersion parameters used when the antifoaming agent is dispersed in the lubricating oil; the current homogeneous dispersion parameters include rotor speed, homogenization time, and dosage; a dispersion effect prediction module for inputting the current homogeneous dispersion parameters into a dispersion effect prediction model to obtain a predicted homogeneous dispersion efficiency and a predicted weighted particle number; wherein the dispersion effect prediction model is determined based on training using a machine learning algorithm; and a homogeneous dispersion optimization module for combining a multi-objective genetic algorithm and a balancing strategy to reverse regulate and optimize the current homogeneous dispersion parameters according to the predicted homogeneous dispersion efficiency and the predicted weighted particle number, so as to determine the optimal homogeneous dispersion parameters.

[0008] Thirdly, this application provides a computer device, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to achieve a method for regulating the dispersion effect of an antifoaming agent during the lubricating oil recovery process.

[0009] According to the specific embodiments provided in this application, the following technical effects are disclosed: This application transforms the macroscopic particle size distribution into a quantifiable evaluation value of dispersion effect through two predictive indicators: homogeneous dispersion efficiency and weighted particle number, thus clarifying the superiority or inferiority of dispersion effect corresponding to different parameter combinations. By combining this evaluation system with machine learning algorithms, multi-objective genetic algorithms, and balancing strategies, a precise mapping relationship between homogeneous dispersion parameters and dispersion effect is established, thereby automatically finding and outputting the optimal combination of homogeneous dispersion parameters. The processing method adopted in this application can automatically regulate the dispersion effect of antifoaming agents during the lubricating oil recovery process. It is low-cost and has few process limitations, achieving a leap from experience-based regulation to intelligent optimization, and from fuzzy judgment to precise quantification, providing reliable technical support for improving the quality of lubricating oil products after recovery. Attached Figure Description

[0010] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0011] Figure 1 is an application environment diagram of a method for regulating the dispersion effect of antifoaming agents during the recovery process of lubricating oil according to an embodiment of this application.

[0012] Figure 2 is a flowchart illustrating a method for regulating the dispersion effect of an antifoaming agent during the recovery process of lubricating oil according to an embodiment of this application.

[0013] Figure 3 is a schematic diagram of the calculation results of homogeneous dispersion efficiency and dispersion ratio.

[0014] Figure 4 is a heatmap of feature correlation analysis.

[0015] Figure 5 is a schematic diagram of the mean square error curve during model training.

[0016] Figure 6 shows the regression diagram of the model training effect.

[0017] Figure 7 is a schematic diagram of the prediction results of homogeneous dispersion efficiency and weighted particle number.

[0018] Figure 8 shows the distribution of the first front-end individuals.

[0019] Figure 9 is a schematic diagram of the balanced score results of each group in the optimization solution.

[0020] Figure 10 is a schematic diagram of the homogeneous dispersion parameter control and optimization process in another embodiment.

[0021] Figure 11 is a schematic diagram of the structure of a computer device provided in an embodiment of this application. Detailed Implementation

[0022] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0023] While industry-standard contamination level tests can reflect the particle size distribution in oil, they cannot directly and quantitatively characterize the dispersion effect of antifoaming agents in lubricating oil. This application represents a leap from experience-based control to intelligent optimization, and from fuzzy judgment to precise quantification, providing reliable technical support for improving the quality of lubricating oil products after recovery.

[0024] To make the above-mentioned objectives, features and advantages of this application more apparent and understandable, the application will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0025] The method for regulating the dispersion effect of antifoaming agents during the lubricating oil recovery process provided in this application embodiment can be applied to the application environment shown in Figure 1. Terminal 101 communicates with server 102 via a network. A data storage system can store the data that server 102 needs to process. The data storage system can be set up separately, integrated into server 102, or placed in the cloud or on other servers. Terminal 101 can send the current homogeneous dispersion parameters used when the antifoaming agent is dispersed in the lubricating oil to server 102. After receiving the parameters, server 102 inputs them into the dispersion effect prediction model to obtain the predicted homogeneous dispersion efficiency and the predicted weighted particle number. Then, combining a multi-objective genetic algorithm and a balancing strategy, based on the predicted homogeneous dispersion efficiency and the predicted weighted particle number, the current homogeneous dispersion parameters are reversely regulated and optimized to determine the optimal homogeneous dispersion parameters. Server 102 can feed back the obtained optimal homogeneous dispersion parameters to terminal 101. Furthermore, in some embodiments, the method for regulating the dispersion effect of antifoaming agents during the lubricating oil recovery process can also be implemented separately by server 102 or terminal 101.

[0026] The terminal 101 can be, but is not limited to, various desktop computers, laptops, smartphones, tablets, IoT devices, and portable wearable devices. The server 102 can be implemented using a standalone server or a server cluster consisting of multiple servers, or it can be a cloud server.

[0027] Among various lubricating oil additives, antifoaming agents are the most difficult to disperse. Achieving uniform dispersion proves that the dispersion process is sufficient to ensure the uniform mixing of other additives. Therefore, addressing the current technical problems in the industry, this application proposes a method for cyclic online rehabilitation of lubricating oils with defective antifoaming properties, thereby repairing the antifoaming characteristics of the lubricating oil.

[0028] In an exemplary embodiment, as shown in FIG2, a method for regulating the dispersion effect of antifoaming agents during lubricant recovery is provided. The method is executed by a computer device, specifically by a computer device such as a terminal or a server alone, or by a terminal and a server together. In this embodiment, the method is described using the server 102 in FIG1 as an example, including the following steps 201 to 203.

[0029] Step 201: Obtain the current homogenization and dispersion parameters used when the antifoaming agent is dispersed in the lubricating oil; the current homogenization and dispersion parameters include rotor speed, homogenization time and dosage.

[0030] After obtaining the data, before proceeding to the next step, it is necessary to determine the dispersion effect prediction model. The determination process includes: (1) performing a homogeneous dispersion orthogonal experiment of the antifoaming agent to determine multiple sets of homogeneous dispersion test results; each set of homogeneous dispersion test results includes homogeneous dispersion parameters and the corresponding particle size distribution range in the lubricating oil.

[0031] In a specific application, an orthogonal experiment on the homogeneous dispersion of the antifoaming agent is performed to determine multiple sets of homogeneous dispersion test results, including: 11) designing nine sets of orthogonal experiments according to a three-factor design and setting three sets of blank control experiments with different additive amounts to obtain an orthogonal experimental table; wherein, the following parameters are set for each set of experiments: rotor speed, homogenization time and dosage.

[0032] To analyze the relationship between variables and the homogeneous dispersion effect of particles, it is necessary to control variables through testing. Based on orthogonality, representative points were selected from the comprehensive experiment for further testing. These representative points exhibited characteristics of uniform dispersion and comparable performance. Therefore, nine orthogonal experimental tables (numbers 1-9) were designed according to a three-factor model, and three blank control groups with different additive amounts were set up (numbers 10-12).

[0033] 12) According to the orthogonal experimental table, adjust the homogenizing pump or high-speed shearing machine to shear the antifoaming agent so that the antifoaming agent is dispersed in the lubricating oil.

[0034] In a specific experiment, an FA25 high-shear dispersing emulsifier was used. Since the recovery process for lubricating oil in factories is mostly a continuous process, homogenizing pumps are used to shear additives (such as antifoaming agents) to disperse them in the lubricating oil and restore its antifoaming properties. Therefore, to achieve an equivalent industrial homogenizing pump speed, the control parameters of the high-speed shearing machine are equivalently replaced by the relationship between angular velocity, linear velocity, and rotor diameter. The specific conversion relationship is shown in the following formula: .

[0035] Among them, V t ω is the shear linear velocity, m / s; w is the angular velocity, rpm; d is the rotor diameter, m.

[0036] Experience and relevant literature indicate that the dispersion effect of additives is closely related to the equipment control parameters and the dosage, including rotor speed and homogenization time. Therefore, experimental research was conducted on the effects of rotor speed, dosage, and homogenization time on the dispersion effect of additives in lubricating oil.

[0037] Taking Mobil LITO H46 hydraulic oil as the lubricant and silicone-based antifoaming agent as the additive, the experimental instruments and materials included: an FA25 high-shear dispersing emulsifier, several 400ml beakers, several 500ml sample bottles, droppers, pipettes, a percentile balance, petroleum ether, new Mobil LITO H46 hydraulic oil, and silicone-based antifoaming agent. The experimental steps are as follows: Step 1: Prepare 12 empty 400ml beakers, numbered 1 to 12. Place each empty beaker sequentially on the percentile balance, and pour 180g to 220g of new Mobil LITO H46 hydraulic oil into each beaker.

[0038] Step 2: Using a pipette, adjust the scale and inject 0.002g, 0.004g and 0.008g of silicone antifoaming agent into beakers numbered 10, 11 and 12 respectively. Then transfer the lubricating oil and silicone antifoaming agent mixture solution in beakers numbered 10, 11 and 12 into 500ml sample vials.

[0039] Step 3: Power on the FA25 high-shear dispersion emulsifier and install the shear rotor. In order to effectively simulate the shear conditions of a homogenizing pump in actual industry, a shear rotor with an outer diameter of 25mm is used.

[0040] Step 4: Prepare an empty beaker containing petroleum ether. Before testing, clean and air-dry the rotor.

[0041] Step 5: After cleaning, manually lower the shear rotor into the beaker containing lubricating oil, so that the rotor is completely immersed in the lubricating oil and is positioned in the center of the beaker without contacting the bottom of the beaker.

[0042] Step 6: Turn on the power to the FA25 high shear dispersion emulsifier and run it at the lowest speed. Then, use a pipette to draw the corresponding mass of silicone antifoaming agent and inject it into a beaker.

[0043] Step 7: Adjust the speed of the emulsifier to reach the preset speed for the experiment, and start timing.

[0044] Step 8: After shearing for a certain period of time, turn off the power and transfer the lubricating oil into the sample vial.

[0045] Step 9: Samples are sent for testing to determine the contamination level of the lubricating oil, and the results are awaited.

[0046] Based on the conversion formula above, by controlling the rotational speed (i.e., angular velocity) within the range of 10000rpm to 16000rpm, and combining this with the rotor diameter of the corresponding model of the FA25 high-shear dispersion emulsifier, the calculated shear linear velocity range is 13.09m / s to 20.94m / s.

[0047] 13) For each group of experiments, the particle size distribution of antifoaming agent particles in the lubricating oil was tested using the ISO 4406 method for detecting lubricating oil contamination, in order to obtain the particle size distribution range in the lubricating oil. The particle size distribution test in the lubricating oil was conducted using specialized equipment for testing lubricating oil contamination and the ISO 4406 method for detecting lubricating oil contamination. The particle number was divided into three ranges according to particle size: ≥4 μm / 100 ml, ≥6 μm / 100 ml, and ≥14 μm / 100 ml. The test results are shown in Table 1.

[0048] Table 1

[0049] (2) For any set of homogenization dispersion test results, calculate the homogenization dispersion efficiency, weighted particle number and homogenization dispersion ratio, and determine a sample; the sample includes rotor speed, homogenization time, dosage, homogenization dispersion efficiency, weighted particle number and homogenization dispersion ratio.

[0050] The ISO 4406 method for detecting contamination in lubricating oils provides the size distribution range of particles, but it cannot directly reflect the actual dispersion effect of additives in different experimental groups under specific homogeneous dispersion parameters (such as rotational speed, dosage, and homogenization time). To effectively quantify the dispersion effect of additives in lubricating oils and clarify the relative effectiveness of different parameter combinations, the ISO 4406 contamination test results are further transformed into analytical indicators for dispersion efficiency, weighted particle number, and homogeneous dispersion ratio. Let the homogeneous dispersion efficiency be η, the weighted particle number be β, and the homogeneous dispersion ratio be D.

[0051] The formula for calculating the weighted particle number is: .

[0052] Where, α j Let x be the weighting coefficient corresponding to the particle size distribution range of the j-th group. Corresponding to the previous step, the particle count was divided into three ranges: ≥4um / 100ml, ≥6um / 100ml, and ≥14um / 100ml. According to industry rules for classifying the importance of particle ranges, smaller particles are more important than larger particles; therefore, the corresponding weighting coefficients are 0.5, 0.3, and 0.2, respectively. j Let be the number of particles corresponding to the j-th particle size distribution range. u is the particle size distribution range group, which can take the values ​​1, 2, and 3 to correspond to the three ranges mentioned above.

[0053] The formula for calculating the homogeneous dispersion ratio D is: .

[0054] The formula for calculating the homogeneous dispersion efficiency η is: .

[0055] Where, β n β represents the weighted particle count of the lubricating oil in the nth group of experiments with added antifoaming agent and a homogenization time of zero; i The weighted number of lubricating oil particles in the i-th group of experiments with added antifoaming agent and a non-zero homogenization time is given by i. n, β n With β i The corresponding dosages of the antifoaming agent are 0.002g, 0.004g, and 0.008g, respectively. According to Table 1 above, n can take values ​​from 10 to 12, and i can take values ​​from 1 to 9.

[0056] Based on the orthogonal experiment for additive particle size distribution testing, the calculated results of homogeneous dispersion efficiency and dispersion ratio are shown in Figure 3. As shown in Figure 3, the dispersion efficiency ranges from 33.12% to 89.91%, and the dispersion ratio ranges from 1.50 to 9.91.

[0057] (3) Perform linear interpolation of adjacent non-outlier values ​​on multiple samples to obtain an expanded sample set. Specifically, perform linear interpolation of adjacent non-outlier values ​​to interpolate between each sample point, resulting in 50 interpolated samples. The purpose of this step is to expand the dataset so that the model built subsequently can effectively capture the relationship between the input parameters and the output parameters.

[0058] (4) Based on the Pearson correlation coefficient, feature correlation analysis is performed on the expanded sample set to determine that the input data includes rotor speed, homogenization time and dosage, and the label data includes homogenization dispersion efficiency and weighted particle number.

[0059] The Pearson correlation coefficient method used in this application is simple and computationally efficient, and can clearly measure the linear relationship between two features. That is, it assumes that the data follows a normal distribution and that the two variables are linearly correlated. The formula for calculating the Pearson correlation coefficient is: .

[0060] Among them, X i Y i For each pair of data points; , denoted as the mean of the two variables; r is the Pearson correlation coefficient, ranging from [-1, 1], where r=1 indicates a perfect positive correlation, r=-1 indicates a perfect negative correlation, and r=0 indicates no linear correlation.

[0061] The correlation strength can be classified according to the absolute value of the calculated correlation coefficient r, as shown in Table 2. The correlation analysis heatmap is shown in Figure 4, illustrating the correspondence between the three homogenization control parameters (rotor speed, dosage, and homogenization time) and dispersion efficiency, dispersion ratio, and weighted particle number, respectively. It is evident that the three control parameters are uncorrelated, indicating they are independent variables, providing input features for the subsequent establishment of a homogenization dispersion parameter control model. Notably, since the correlation coefficient between dispersion ratio and dispersion efficiency is 0.83, indicating a very strong correlation, only one parameter needs to be selected as the output. This application selects dispersion efficiency and weighted particle number as the outputs. The correlation between the output parameters and the input parameters is greater than 0.35, indicating a moderate to high correlation; therefore, this control parameter can be used for modeling.

[0062] Table 2

[0063] (5) Construct a BP neural network prediction model with a dual-output structure. Considering that the BP neural network has a strong nonlinear fitting ability and a backpropagation mechanism, it can continuously iterate and optimize its error function during model training, and finally effectively capture the relationship between input and output. Therefore, a BP neural network prediction model with a dual-output structure is established. The model structure consists of an input layer, a hidden layer and an output layer.

[0064] (6) Normalize the input data of the expanded sample set and then input it into the BP neural network prediction model. Combine it with the corresponding label data for training to obtain the dispersion effect prediction model.

[0065] Normalization of feature data aims to map different types of feature data to a specific range, thereby eliminating the influence of units on the data without affecting its characteristics, thus accelerating the processing of machine learning models and shortening training time. Normalization only applies to input features (rotation speed, dosage, homogenization time) and does not include label data. This application specifically employs the max-min normalization method, mapping the feature data to [0,1]. The normalization formula is as follows: .

[0066] in, x represents the normalized data; x represents the original data; x max x represents the maximum value of the data. min This represents the minimum value of the data.

[0067] To train a homogeneous dispersion efficiency and weighted particle number prediction model that meets practical needs, the expanded sample set needs to be divided into a training set and a test set. To utilize the data for model training and evaluation, based on the trade-off between model learning and prediction capabilities, the expanded sample set is divided into training and test sets in an 8:2 ratio. Next, the model hyperparameters are set, including: number of hidden layers, number of hidden layer neurons, number of linear layers, learning rate, backpropagation optimization function, and batch size, as shown in Table 3.

[0068] Table 3

[0069] After setting the model hyperparameters, training the model begins. Training stops when the training loss function reaches its minimum value. The number of training epochs is set to 17. The mean squared error reaches its minimum of 0.00048 in the 11th epoch, as shown in Figure 5. The regression graph of its training effect is shown in Figure 6. The training fit effect of the training, validation, and testing parts are all above 0.94, proving that the model is sufficiently trained and effectively learns the correlation between input and output. Therefore, it can be further used to perform the prediction task of dispersion efficiency and weighted particle number.

[0070] Step 202: Input the current homogeneous dispersion parameters into the dispersion effect prediction model to obtain the predicted homogeneous dispersion efficiency and the predicted weighted particle number; wherein, the dispersion effect prediction model is determined based on machine learning algorithm training.

[0071] The dispersion efficiency and weighted particle number of antifoaming agents in lubricating oil were predicted using a trained BP neural network model. Figures 7(a) and 7(b) show the prediction results for homogeneous dispersion efficiency and weighted particle number, respectively. The goodness of fit (R²) for dispersion efficiency prediction was 0.90, the mean absolute error (MAE) was 3.39%, and the root mean square error (RMSE) was 4.93%. The goodness of fit (R²) for weighted particle number prediction was [not specified in the original text].2 The mean absolute error (MAE) was 0.97, the root mean square error (RMSE) was 12992.87, and the root mean square error (RMSE) was 18499.69. All prediction results fell within the 95% confidence interval, indicating high reliability. The results demonstrate excellent predictive performance, with particularly good prediction of the weighted particle number.

[0072] Step 203: Combining a multi-objective genetic algorithm and a balancing strategy, the current homogeneous dispersion parameters are reverse-regulated and optimized based on the predicted homogeneous dispersion efficiency and the predicted weighted particle number, in order to determine the optimal homogeneous dispersion parameters.

[0073] After using a machine learning-based prediction model for homogenization control parameters, this application employs a multi-objective genetic optimization algorithm for control parameters to find the optimal control parameters, thereby maximizing the dispersion efficiency and minimizing the weighted particle number. Step 203 includes: (31) constructing an objective function with the goal of maximizing homogenization dispersion efficiency and minimizing the weighted particle number, and determining the corresponding parameter constraints; wherein the parameter constraints include: rotor speed limit range, dosage limit range, homogenization time limit range, homogenization dispersion efficiency limit range, and weighted particle number limit range. For example, the rotor speed limit range is 10000rpm~20000rpm; the dosage limit range is 0.002g~0.008g; the homogenization time limit range is 0.5min~3min; the homogenization dispersion efficiency limit range is 0~1; and the weighted particle number limit range is 500~300000.

[0074] (32) Using a multi-objective genetic optimization algorithm, the objective function is solved by combining the parameter constraints, the predicted homogeneous dispersion efficiency and the predicted weighted particle number, so as to obtain the Pareto optimal solution set.

[0075] Considering that multi-objective genetic optimization algorithms can solve optimization problems by simulating the evolutionary process of organisms in nature, and that in multi-objective optimization problems, genetic algorithms can effectively search for optimal solutions to multiple objective functions, the goal is to find a set of solutions. These solutions are non-dominated across all objectives, meaning that no other solution is better than these solutions across all objectives.

[0076] In this application, the parameters of the multi-objective genetic algorithm are set as follows: population size is 100; optimal front-end individual coefficient is 0.4; maximum number of evolutionary iterations is 200; and fitness function value deviation is 1×10⁻⁶. -6 .

[0077] A Pareto optimal solution search was performed. A Pareto optimal solution is a solution that cannot further improve any other objective without worsening at least one objective. The set of Pareto optimal solutions constitutes the Pareto front. The results are shown in Table 4. Forty Pareto optimal solutions were found, with optimized dispersion efficiency ranging from 87.44% to 99.84% and weighted particle number ranging from 87,137 to 98,435. During the execution of the multi-objective optimization algorithm based on genetic algorithm, the distribution of individuals in the first front was automatically plotted, and the distribution was updated once with each generation of algorithm evolution. When the iteration stopped, the distribution map of the first front individuals was obtained as shown in Figure 8. Figure 8 shows that the Pareto optimal solutions in the first front are evenly distributed. The number of returned Pareto optimal solutions was 40, and the population size was 100, indicating that the optimal front individual coefficient of 0.3 was effective.

[0078] Table 4

[0079] (33) A balancing strategy is adopted to find the optimal balance point between homogeneous dispersion efficiency and weighted particle number in the Pareto optimal solution set, so as to obtain the optimal homogeneous dispersion parameters. This application adopts a balancing strategy to find the best trade-off point between conflicting objectives, that is, to find the balance point between homogeneous dispersion efficiency and weighted particle number in the optimal solution set obtained in the previous step. Its purpose is to reduce the influence of subjective preferences based on quantitative indicators. The specific steps include: 1) Calculating the balance score B_Score of each solution in the Pareto optimal solution set using the following formula: .

[0080] Where η is the homogeneous dispersion efficiency, %; β Zero The number of weighted particles after normalization is [number].

[0081] 2) Combining all the balance scores, the solution with the highest score is taken as the optimal homogeneous dispersion parameter.

[0082] The optimized equilibrium scores for each group are shown in Figure 9. Based on the equilibrium score calculations, group 21 achieved the highest score. Therefore, the recommended solution according to the equilibrium strategy is: rotation speed 12247.5 rpm; dosage 0.0077 g; homogenization time 2.97 min. The model predicts the following outputs: dispersion efficiency 99.04%; weighted particle count 94726. Through multi-objective genetic algorithm optimization, compared to the homogenized dispersion orthogonal experiment results, the dispersion efficiency increased by 9.13%, and the weighted particle count decreased by 6.38%.

[0083] In a practical application, after obtaining the optimal homogeneous dispersion parameters, the method further includes: using response surface methodology to analyze the sensitivity between the optimal homogeneous dispersion parameters and the predicted homogeneous dispersion efficiency and the predicted weighted particle number, in order to quantify the degree of parameter influence. By constructing a response surface, this application reveals a complex nonlinear relationship between input parameters and output performance, specifically manifested as a significant synergistic effect among the variables, identifying a high-performance robust operating region with a rotational speed range of 12205.10 rpm to 12448.57 rpm, a dosage range of 0.0056 g to 0.008 g, and a homogenization time of 2.43 min to 2.99 min, thereby providing a basis for precise homogeneous dispersion control.

[0084] Furthermore, to verify the effectiveness of the optimized homogeneous dispersion parameters in practical applications, this application also conducted durability tests on the recovery additive, aiming to verify its dispersion effect and durability. Using the homogeneous dispersion efficiency and weighted particle number calculation method described in this application, the optimal homogeneous dispersion parameter combination range obtained above was applied in practice. Comparing three sets of data, the calculation results are shown in Table 5, with combination 2 showing the best effect. The results of the oil anti-foaming property recovery durability test are shown in Table 6. Foaming characteristics were tested at 1, 4, 7, and 14 days after recovery, and the foaming characteristics remained at a very low level. The results indicate that by using the optimal homogeneous dispersion parameters, the anti-foaming additive can effectively repair the foaming characteristics of the oil, and the additive can effectively disperse, making its effect in lubricating oil more durable. Therefore, the recommended homogeneous dispersion parameter combination for the anti-foaming agent is: rotation speed 12310 rpm, dosage 0.007 g, and homogenization time 2.6 min.

[0085] Table 5

[0086] Table 6

[0087] In summary, as shown in Figure 10, this application establishes the rotational speed relationship between a high-shear dispersion emulsification experimental machine and an industrial homogenizing pump to conduct orthogonal experiments on the homogenization and dispersion of antifoaming agents; constructs a calculation model for homogenization and dispersion efficiency and weighted particle number; constructs a machine learning model for homogenization and dispersion efficiency and weighted particle number; performs data preparation, correlation analysis, feature normalization processing, dataset partitioning, model training and prediction; uses a multi-objective genetic algorithm to optimize homogenization control parameters (rotational speed, dosage, homogenization time), establishes a balance strategy to determine the optimal parameters; uses response surface methodology to analyze the sensitivity between the optimal solution parameters and dispersion efficiency and weighted particle number, thus providing a basis for precise homogenization distribution control; and verifies the model through durability testing. Thus, this application accurately predicts the dispersion effect of antifoaming agents in lubricating oil through a machine learning model, uses a multi-objective genetic algorithm to reverse-regulate and optimize homogenization and dispersion parameters, and also establishes a method for evaluating dispersion effect to determine the optimal homogenization and dispersion control parameters for antifoaming agents in lubricating oil.

[0088] Based on the same inventive concept, this application also provides a system. The solution provided by this system is similar to the solution described in the above method. Therefore, the specific limitations of one or more system embodiments provided below can be found in the limitations of the method above, and will not be repeated here.

[0089] In one exemplary embodiment, a system for regulating the dispersion effect of an antifoaming agent during lubricant recovery is provided, comprising: a parameter acquisition module for acquiring the current homogenization dispersion parameters used when the antifoaming agent is dispersed in the lubricating oil; the current homogenization dispersion parameters include rotor speed, homogenization time, and dosage.

[0090] The dispersion effect prediction module is used to input the current homogeneous dispersion parameters into the dispersion effect prediction model to obtain the predicted homogeneous dispersion efficiency and the predicted weighted particle number; wherein, the dispersion effect prediction model is determined based on machine learning algorithm training.

[0091] The homogeneous dispersion optimization module is used to combine a multi-objective genetic algorithm and a balancing strategy to reverse-regulate and optimize the current homogeneous dispersion parameters based on the predicted homogeneous dispersion efficiency and the predicted weighted particle number, so as to determine the optimal homogeneous dispersion parameters.

[0092] This application proposes a method for regulating and optimizing the dispersion effect of antifoaming agents. First, oil is sequentially fed from the main oil tank into a filter, homogenizing pump or high-speed shear mill, and static mixer. The antifoaming agent is then added to the homogenizing pump or high-speed shear mill, where it is dispersed in the lubricating oil under high-speed shearing. The recovered oil then returns to the main oil tank via a return pipe. The core step is controlling the homogenization and dispersion parameters (including rotor speed, dosage, and homogenization time) of the homogenizing pump or high-speed shear mill to subject the antifoaming agent to shearing action, thus achieving uniform dispersion in the lubricating oil. Next, orthogonal experimental data on the antifoaming agent are preprocessed to establish and train a machine learning prediction model with a dual-output structure, accurately predicting the dispersion effect of the antifoaming agent in the lubricating oil (dispersion efficiency, weighted particle number). A multi-objective genetic algorithm is used to regulate and optimize the dispersion parameters (rotation speed, dosage, and homogenization time). Finally, a balancing strategy is used to determine the optimal homogenization and dispersion parameters. Then, the ISO 4406 contamination test results of the lubricating oil are converted into dispersion efficiency and weighted particle number for analysis.

[0093] Thus, this application accurately predicts the dispersion effect of antifoaming agents in lubricating oil through machine learning models, uses multi-objective genetic algorithms to reverse regulate and optimize homogeneous dispersion parameters, and establishes a method for evaluating dispersion effect to determine the optimal homogeneous dispersion regulation parameters of antifoaming agents in lubricating oil.

[0094] In an exemplary embodiment, a computer device is provided, which may be a server or a terminal, and its internal structure diagram is shown in Figure 11. The computer device includes a processor, memory, input / output interfaces (I / O), and a communication interface. The processor, memory, and I / O interfaces are connected via a system bus, and the communication interface is connected to the system bus via the I / O interfaces. The processor of the computer device provides computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and internal memory. The non-volatile storage medium stores an operating system, computer programs, and a database. The internal memory provides an environment for the operation of the operating system and computer programs in the non-volatile storage medium. The I / O interfaces of the computer device are used for exchanging information between the processor and external devices. The communication interface of the computer device is used for communication with external terminals via a network connection. When the computer program is executed by the processor, it implements a method for regulating the dispersion effect of antifoaming agents during lubricant recovery.

[0095] Those skilled in the art will understand that the structure shown in Figure 11 is merely a block diagram of a portion of the structure related to the present application and does not constitute a limitation on the computer device to which the present application is applied. Specific computer devices may include more or fewer components than those shown in the figure, or combine certain components, or have different component arrangements.

[0096] In one exemplary embodiment, a computer device is also provided, including: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement the steps in the above-described method embodiments.

[0097] In one exemplary embodiment, a computer-readable storage medium is provided storing a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0098] In one exemplary embodiment, a computer program product is provided, including a computer program that, when executed by a processor, implements the steps in the above-described method embodiments.

[0099] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties, and the collection, use and processing of the relevant data must comply with relevant regulations.

[0100] Those skilled in the art will understand that all or part of the processes in the above embodiments can be implemented by a computer program instructing related hardware. The computer program can be stored in a non-volatile computer-readable storage medium. When executed, the computer program can include the processes of the embodiments described above. Any references to memory, databases, or other media used in the embodiments provided in this application can include at least one of non-volatile and volatile memory. Non-volatile memory can include read-only memory (ROM), magnetic tape, floppy disk, flash memory, optical memory, high-density embedded non-volatile memory, resistive random access memory (ReRAM), magnetic random access memory (MRAM), ferroelectric random access memory (FRAM), phase change memory (PCM), graphene memory, etc. Volatile memory can include random access memory (RAM) or external cache memory, etc. By way of illustration and not limitation, RAM can take many forms, such as Static Random Access Memory (SRAM) or Dynamic Random Access Memory (DRAM).

[0101] The databases involved in the embodiments provided in this application may include at least one type of relational database and non-relational database. Non-relational databases may include, but are not limited to, blockchain-based distributed databases. The processors involved in the embodiments provided in this application may be general-purpose processors, central processing units, graphics processing units, digital signal processors, programmable logic devices, quantum computing-based data processing logic devices, etc., and are not limited to these.

[0102] The technical features of the above embodiments can be combined in any way. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0103] This document uses specific examples to illustrate the principles and implementation methods of this application. The descriptions of the above embodiments are only for the purpose of helping to understand the methods and core ideas of this application. Furthermore, those skilled in the art will recognize that, based on the ideas of this application, there will be changes in the specific implementation methods and application scope. Therefore, the content of this specification should not be construed as a limitation of this application.

Claims

1. A method for regulating the dispersion effect of antifoaming agents during the recovery process of lubricating oil, characterized in that, The method includes: obtaining the current homogeneous dispersion parameters used when the antifoaming agent is dispersed in lubricating oil; the current homogeneous dispersion parameters include rotor speed, homogenization time, and dosage; inputting the current homogeneous dispersion parameters into a dispersion effect prediction model to obtain a predicted homogeneous dispersion efficiency and a predicted weighted particle number; wherein, the dispersion effect prediction model is determined based on training using a machine learning algorithm; and combining a multi-objective genetic algorithm and a balancing strategy, the current homogeneous dispersion parameters are reverse-regulated and optimized based on the predicted homogeneous dispersion efficiency and the predicted weighted particle number to determine the optimal homogeneous dispersion parameters.

2. The method for controlling the dispersion effect of antifoaming agents during the lubricating oil recovery process according to claim 1, characterized in that, The process of determining the dispersion effect prediction model includes: performing an orthogonal experiment on the homogeneous dispersion of the antifoaming agent to determine multiple sets of homogeneous dispersion test results; each set of homogeneous dispersion test results includes homogeneous dispersion parameters and the corresponding particle size distribution range in the lubricating oil; for any set of homogeneous dispersion test results, calculating the homogeneous dispersion efficiency, weighted particle number, and homogeneous dispersion ratio, and determining a sample; the sample includes rotor speed, homogenization time, dosage, homogeneous dispersion efficiency, weighted particle number, and homogeneous dispersion ratio; performing linear interpolation of adjacent non-outlier values ​​on multiple samples to obtain an expanded sample set; performing feature correlation analysis on the expanded sample set based on the Pearson correlation coefficient to determine that the input data includes rotor speed, homogenization time, and dosage, and the label data includes homogeneous dispersion efficiency and weighted particle number; constructing a BP neural network prediction model with a dual-output structure; normalizing the input data of the expanded sample set, then inputting it into the BP neural network prediction model, and training it in conjunction with the corresponding label data to obtain the dispersion effect prediction model.

3. The method for controlling the dispersion effect of antifoaming agents during the lubricating oil recovery process according to claim 2, characterized in that, The formula for calculating the weighted particle number is: ; where α j x is the weighting coefficient corresponding to the particle size distribution range of the j-th group; j Let be the number of particles corresponding to the j-th particle size distribution range, and u be the number of particle size distribution range groups; the formula for calculating the homogeneous dispersion ratio D is: The formula for calculating the homogeneous dispersion efficiency η is as follows: ; where β n β represents the weighted particle count of the lubricating oil in the nth group of experiments with added antifoaming agent and a homogenization time of zero; i The weighted number of lubricating oil particles in the i-th group of experiments with added antifoaming agent and a non-zero homogenization time is given by i. n。 4. The method for controlling the dispersion effect of antifoaming agents during the lubricating oil recovery process according to claim 1, characterized in that, Combining a multi-objective genetic algorithm and a balancing strategy, the current homogeneous dispersion parameters are reverse-regulated and optimized based on the predicted homogeneous dispersion efficiency and the predicted weighted particle number to determine the optimal homogeneous dispersion parameters. This includes: constructing an objective function with the goal of maximizing homogeneous dispersion efficiency and minimizing the weighted particle number, and determining the corresponding parameter constraints; using a multi-objective genetic optimization algorithm, combining the parameter constraints, the predicted homogeneous dispersion efficiency, and the predicted weighted particle number, to solve the objective function and obtain the Pareto optimal solution set; and using a balancing strategy to find the optimal balance point between homogeneous dispersion efficiency and weighted particle number in the Pareto optimal solution set to obtain the optimal homogeneous dispersion parameters.

5. The method for controlling the dispersion effect of antifoaming agents during the lubricating oil recovery process according to claim 4, characterized in that, The parameter constraints include: rotor speed limit range, dosage limit range, homogenization time limit range, homogenization dispersion efficiency limit range, and weighted particle number limit range.

6. The method for controlling the dispersion effect of antifoaming agents during the lubricating oil recovery process according to claim 4, characterized in that, A balancing strategy is employed to find the optimal balance point between homogeneous dispersion efficiency and weighted particle number in the Pareto optimal solution set, thereby obtaining the optimal homogeneous dispersion parameters. This includes calculating the balance score B_Score for each solution in the Pareto optimal solution set using the following formula: Where η is the homogeneous dispersion efficiency; β Zero The normalized weighted number of particles; combining all the balance scores, the solution with the highest score is taken as the optimal homogeneous dispersion parameter.

7. The method for controlling the dispersion effect of antifoaming agents during the lubricating oil recovery process according to claim 4, characterized in that, After obtaining the optimal homogeneous dispersion parameters, the method further includes: using response surface methodology to analyze the sensitivity between the optimal homogeneous dispersion parameters and the predicted homogeneous dispersion efficiency and the predicted weighted particle number, in order to quantify the degree of influence of the parameters.

8. The method for controlling the dispersion effect of antifoaming agents during the lubricating oil recovery process according to claim 2, characterized in that, An orthogonal experiment was conducted to homogenize and disperse the antifoaming agent to determine multiple sets of homogenization and dispersion test results. This included: designing nine orthogonal experiments according to a three-factor model and setting three blank control experiments with different additive amounts to obtain an orthogonal experimental table; each experiment set the following parameters: rotor speed, homogenization time, and dosage; according to the orthogonal experimental table, a homogenizing pump or high-speed shearing machine was used to shear the antifoaming agent to disperse it in the lubricating oil; for each experiment, the particle size distribution of the antifoaming agent particles in the lubricating oil was tested using the ISO 4406 lubricating oil contamination detection method to obtain the particle size distribution range in the lubricating oil.

9. A system for regulating the dispersion effect of antifoaming agents during lubricant recovery, characterized in that, The system includes: a parameter acquisition module for acquiring the current homogeneous dispersion parameters used when the antifoaming agent is dispersed in lubricating oil; the current homogeneous dispersion parameters include rotor speed, homogenization time, and dosage; a dispersion effect prediction module for inputting the current homogeneous dispersion parameters into a dispersion effect prediction model to obtain a predicted homogeneous dispersion efficiency and a predicted weighted particle number; wherein the dispersion effect prediction model is determined based on training using a machine learning algorithm; and a homogeneous dispersion optimization module for combining a multi-objective genetic algorithm and a balancing strategy to reverse-regulate and optimize the current homogeneous dispersion parameters based on the predicted homogeneous dispersion efficiency and the predicted weighted particle number, in order to determine the optimal homogeneous dispersion parameters.

10. A computer device, comprising: A memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that the processor executes the computer program to implement a method for regulating the dispersion effect of an antifoaming agent during the lubricating oil recovery process according to any one of claims 1-8.

Images