Intelligent control method and control system for chemical production equipment

Abstract

The invention relates to the technical field of equipment control, in particular to an intelligent control method and system for chemical production equipment. The method comprises the following steps: constructing a forward analysis data segment of a newest data point in a historical rotating speed data time sequence of the newest data point of the stirrer; calculating a primary change factor of the latest data point; obtaining a motor power data segment corresponding to the forward analysis data segment of the latest data point, and obtaining the motor power data segment according to the primary change factor of the latest data point and the difference between the cumulative sum of the collection moments of the extreme points in the forward analysis data segment of the latest data point and the cumulative sum of the collection moments of the extreme points in the corresponding motor power data segment. Calculating a target change factor of the latest data point; and the preset maximum restart frequency in the random restart hill climbing algorithm is adjusted based on the target change factor of the latest data point, so that intelligent control of the stirrer is realized, and the accuracy of intelligent control of the stirrer is effectively improved.

Description

Technical Field

[0001] The present invention relates to the technical field of equipment control, and particularly to an intelligent control method and control system for chemical production equipment. Background Art

[0002] In the process of chemical production, the stirrer is one of the core equipment and can be applied to various reactors, mixing equipment, etc. Accurately controlling the rotation speed of the stirrer can ensure the uniformity of material mixing, enabling each component to fully contact and react. If the rotation speed is unstable, it may lead to uneven material mixing and fluctuations in product quality. For example, the uneven distribution of active ingredients in drugs affects the efficacy; the taste and quality of food are inconsistent, reducing the market competitiveness of the product, etc.

[0003] Traditional stirrer control methods usually rely on fixed rotation speeds or simple manual adjustments. The disadvantage of this method is that it cannot respond in a timely manner to complex and changing production conditions and is also difficult to achieve precise control. With the development of chemical production towards automation and refinement, the deficiencies of existing control methods in the precise regulation of process parameters are becoming increasingly prominent. There is an urgent need to introduce new control technologies with the capabilities of condition perception and intelligent decision-making to achieve efficient and stable operation of the stirring process.

[0004] In chemical production, complex conditions such as changes in material viscosity and temperature fluctuations caused by reaction heat effects will make the optimal solution space for stirrer rotation speed control exhibit multi-peak characteristics, and traditional control algorithms are prone to falling into local optima. The random restart hill climbing algorithm can explore in different solution space regions through multiple random initializations of the search, breaking through the bondage of local optimal solutions and better fitting complex and changing production conditions. However, the random restart hill climbing algorithm uses a fixed maximum number of restarts. If the maximum number of restarts is set too large, it may cause the algorithm to end prematurely and fail to accurately find the optimal solution; if the maximum number of restarts is set too small, it may lead to waste of computing resources and increased time costs.

[0005] Therefore, there is an urgent need to solve the problem of how to dynamically adjust the maximum number of restarts in the random restart hill climbing algorithm according to the actual production conditions of the stirrer, so as to accurately achieve the rotation speed regulation of the stirrer. Summary of the Invention

[0006] In order to solve the technical problem of how to dynamically adjust the maximum number of restarts in the random restart hill climbing algorithm according to the actual production conditions of the stirrer, so as to accurately achieve the rotation speed regulation of the stirrer, the present invention provides an intelligent control method and control system for chemical production equipment.

[0007] In a first aspect, the present invention provides an intelligent control method for chemical production equipment, adopting the following technical solution: An intelligent control method for chemical production equipment includes the steps: Construct a forward analysis data segment of the latest data point in the historical rotational speed data time series of the latest data point of the agitator; calculate the primary change factor of the latest data point : ; 、 are respectively the number of extreme points and the number of adjacent extreme point groups in the forward analysis data segment of the latest data point, and is the average acquisition interval of adjacent extreme points in the forward analysis data segment of the latest data point, and is the acquisition interval of the th group of adjacent extreme points in the forward analysis data segment of the latest data point, where is a linear normalization function; obtain the motor power data segment corresponding to the forward analysis data segment of the latest data point, and calculate the target change factor of the latest data point according to the difference between the primary change factor of the latest data point, the cumulative sum of the acquisition times of extreme points in the forward analysis data segment of the latest data point, and the cumulative sum of the acquisition times of extreme points in the corresponding motor power data segment; adjust the preset maximum restart times in the random restart hill climbing algorithm based on the target change factor of the latest data point to achieve intelligent control of the agitator.

[0008] According to the intelligent control method for chemical production equipment provided by the present invention, the agitator can accurately achieve intelligent control by obtaining the rotational speed value required at each moment through the random restart hill climbing algorithm. In the process of obtaining the rotational speed value required at each moment of the agitator, the present invention takes into account that the forward analysis data segments of data points with different rotational speed changes have different requirements for the maximum restart times, and random setting may lead to failure to accurately find the global optimal solution or waste of resources; therefore, the present invention determines its primary change factor by analyzing the change intensity of rotational speed data in the forward analysis data segment of each data point, so as to evaluate the degree of demand of the data point for the maximum restart times, effectively improving the accuracy of the obtained rotational speed value required by the agitator. In addition, the present invention also takes into account that interference factors of rotational speed data will be generated during the operation of the agitator, affecting the accuracy of rotational speed data; therefore, the present invention can effectively improve the accuracy of the maximum restart times of the obtained data points by introducing the motor power data of the agitator to correct the primary change factor, thereby effectively improving the accuracy of the rotational speed control of the agitator.

[0009] According to the intelligent control method for chemical production equipment provided by the present invention, before constructing the forward analysis data segment of the latest data point in the historical rotational speed data time series of the latest data point of the agitator, it further includes: obtaining the rotational speed value of the agitator at each acquisition moment as a data point, and obtaining a rotational speed data time series after preprocessing.

[0010] In view of the fact that the acquisition of agitator speed data is prone to noise or data loss due to interference from the surrounding environment during the process, a series of preprocessing measures are adopted to improve such situations, laying a foundation for accurately realizing the control of the agitator speed.

[0011] According to an intelligent control method for chemical production equipment provided by the present invention, constructing a forward analysis data segment of the latest data point in the historical speed data time series of the agitator includes: sequentially obtaining data points in the historical speed data time series of the latest data point of the agitator based on a preset length of the forward analysis data segment, so as to obtain the forward analysis data segment of the latest data point.

[0012] According to an intelligent control method for chemical production equipment provided by the present invention, obtaining the motor power data segment corresponding to the forward analysis data segment of the latest data point includes: taking each acquisition moment and the corresponding motor power value as a power data point to obtain a power data time series; determining the motor power data segment corresponding to the forward analysis data segment of the latest data point according to the delay time between the speed data and the motor power; wherein the lengths of the forward analysis data segment of the latest data point and the corresponding motor power data segment are the same.

[0013] The present invention verifies the authenticity of the primary change factor of the agitator speed data by introducing the motor power. In this process, considering that there will be a delay in the response of the motor power to the change of the speed data, the present invention determines the motor power data segment corresponding to the forward analysis data segment of the latest data point according to the delay time between the speed data and the motor power, which can effectively improve the accuracy of the correlation between the obtained speed data and the motor power data, so as to accurately evaluate the authenticity of the primary change factor of the agitator speed data.

[0014] According to an intelligent control method for chemical production equipment provided by the present invention, calculating the target change factor of the latest data point includes: calculating the target change factor of the latest data point : ; is the primary change factor of the latest data point, is the acquisition moment of the nth extreme point in the forward analysis data segment of the latest data point, are the numbers of extreme points in the forward analysis data segment of the latest data point and the corresponding motor power data segment respectively, 、 are respectively the latest data point forward analysis data segment and the corresponding motor power data segment, is the The acquisition time of the th extreme point in the motor power data segment corresponding to the forward analysis data segment of the latest data point, is a hyperparameter, is a linear normalization function, is the absolute value symbol.

[0015] The present invention provides an accurate calculation formula for the target change factor of the latest data point. By analyzing the difference in the acquisition time of the extreme points in the forward analysis data segment of the latest data point and the corresponding motor power data segment, the correlation between the two can be evaluated, and the primary change factor of the latest data point can be accurately adjusted, so as to accurately obtain the target change factor of the latest data point.

[0016] According to an intelligent control method for chemical production equipment provided by the present invention, adjusting the preset maximum restart times in the random restart hill climbing algorithm based on the target change factor of the latest data point includes: ; is the maximum restart times of the latest data point, is the target change factor of the latest data point, is a preset adjustment coefficient, is the preset maximum restart times, is the ceiling symbol.

[0017] According to an intelligent control method for chemical production equipment provided by the present invention, adjusting the preset maximum restart times in the random restart hill climbing algorithm based on the target change factor of the latest data point to achieve intelligent control of the stirrer includes: obtaining the local optimal solution and the corresponding objective function value of the latest data point in each restart, and obtaining the local optimal solution corresponding to the minimum objective function value in the maximum restart times as the global optimal solution of the latest data point; realizing intelligent control of the stirrer according to the global optimal solution.

[0018] The present invention can obtain the required rotational speed value of the latest data point through the random restart hill climbing algorithm, can adapt to the complex and changeable production conditions of the stirrer, and can also jump out of the local optimal solution to accurately obtain the global optimal solution of the latest data point when the material to be stirred changes, so as to accurately realize the intelligent control of the rotational speed of the stirrer.

[0019] According to an intelligent control method for chemical production equipment provided by the present invention, a method for obtaining the objective function value includes: constructing an objective function in the historical rotational speed data time series of the latest data point: ; is the objective function value of the solution, The motor power value required for this solution, The remaining stirring time required for this solution when stirring the product, Is a linear normalization function.

[0020] According to an intelligent control method for chemical production equipment provided by the present invention, the intelligent control of the stirrer is realized according to the global optimal solution, including: adjusting the rotation speed value of the stirrer to the rotation speed value corresponding to the global optimal solution based on PID.

[0021] In a second aspect, the present invention provides an intelligent control system for chemical production equipment, adopting the following technical solution: An intelligent control system for chemical production equipment, including: a processor and a memory, the memory stores computer program instructions, and when the computer program instructions are executed by the processor, the above-mentioned intelligent control method for chemical production equipment is realized.

[0022] By adopting the above technical solution, the above-mentioned intelligent control method for chemical production equipment is generated into a computer program and stored in the memory, so as to be loaded and executed by the processor, and thus a terminal device is manufactured according to the memory and the processor, which is convenient to use.

[0023] The present invention has the following technical effects: Based on the above technical solution, for an intelligent control method and control system for chemical production equipment provided by the present invention, when realizing the intelligent control of the stirrer rotation speed, the rotation speed value required by the stirrer at each moment is obtained through the random restart hill climbing algorithm, and the intelligent control of the stirrer can be accurately realized. In the process of obtaining the rotation speed value required by the stirrer at each moment, the present invention considers that the forward analysis data segments of different rotation speed change data points have different requirements for the maximum restart times, and random setting may lead to the inability to accurately find the global optimal solution or resource waste; therefore, the present invention determines its primary change factor by analyzing the rotation speed data change intensity in the forward analysis data segment of each data point, so as to evaluate the degree of demand of the data point for the maximum restart times, effectively improving the accuracy of the obtained rotation speed value required by the stirrer. In addition, the present invention also considers that interference factors of rotation speed data will be generated during the operation of the stirrer, affecting the accuracy of the rotation speed data; therefore, the present invention corrects the primary change factor by introducing the stirrer motor power data, which can effectively improve the accuracy of the maximum restart times of the obtained data points, and thus effectively improve the accuracy of the stirrer rotation speed control. Description of the Drawings

[0024] Figure 1 Is a flow chart in an intelligent control method for chemical production equipment provided by an embodiment of the present invention. Detailed Embodiment

[0025] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the accompanying drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, rather than all the embodiments.

[0026] The embodiments of the present invention disclose an intelligent control method for chemical production equipment. This method evaluates the required maximum restart times by analyzing the severity of the change in the agitator speed, so that the required speed value of the agitator at the latest moment can be accurately obtained based on this, and the accuracy of the agitator speed regulation can be effectively improved.

[0027] Specifically, please refer to Figure 1 as shown in Figure 1 which is a schematic flowchart of an intelligent control method for chemical production equipment provided by the embodiments of the present invention. This method specifically includes the following steps: S1: Obtain each data point in the time series of agitator speed data.

[0028] Exemplarily, in the embodiments of the present invention, obtaining each data point in the time series of agitator speed data includes: obtaining the speed value of the agitator at each acquisition moment as a data point, and obtaining the time series of speed data after preprocessing.

[0029] Among them, the preprocessing method can be data denoising, missing data interpolation, analog-to-digital conversion, etc., which can be specifically set according to actual needs.

[0030] Specifically, when collecting the time series of agitator speed data, the current acquisition moment can be used as the latest acquisition moment, and the current acquisition moment and the corresponding speed value can be used as the latest data point. When obtaining the required speed value of the agitator at the latest data point, the time series of speed data from the start of the agitator to the latest acquisition moment can be obtained before the latest data point to obtain the historical time series of speed data of the latest data point.

[0031] Among them, the acquisition frequency can be set to 5 times per second; the acquisition frequency can be specifically set according to actual needs, and the embodiments of the present invention do not limit this too much here.

[0032] It should be noted that when using the random restart hill climbing algorithm to intelligently control the speed data of the agitator, the maximum restart times in the random restart hill climbing algorithm need to be adjusted according to the speed data. In the time series of agitator speed data, if the change in the speed data of the data point is more severe, it means that the agitator needs more maximum restart times to explore the solution space to seek the optimal solution at the acquisition moment corresponding to this data point; on the contrary, if the speed data changes relatively regularly and smoothly, it does not require too many maximum restart times to determine the optimal solution.

[0033] Therefore, the embodiments of the present invention can evaluate the degree of demand for the maximum restart times at each acquisition moment by analyzing the severity of the change in the rotational speed data in the time series of the rotational speed data of the stirrer, and dynamically adjust the preset maximum restart times according to the degree of demand for the maximum restart times at the acquisition moment, so as to accurately obtain the rotational speed value required by the stirrer at the acquisition moment and achieve intelligent control, that is, perform the following steps.

[0034] S2: Construct a forward analysis data segment of the latest data point in the historical rotational speed data time series of the latest data point of the stirrer, and calculate the primary change factor of the latest data point according to the number of extreme points in the forward analysis data segment of the latest data point and the corresponding acquisition moment.

[0035] It should be noted that the operating conditions of the stirrer may change over time, such as equipment wear. Therefore, when analyzing the degree of demand for the maximum restart times of the latest data point, it is necessary to construct a forward analysis data segment of the latest data point, pay attention to the current and recent working conditions of the stirrer, analyze its performance in the recent period of time, and avoid the influence of premature rotational speed data on the rotational speed value required by the latest data point.

[0036] Exemplarily, in the embodiments of the present invention, constructing a forward analysis data segment of the latest data point in the historical rotational speed data time series of the latest data point of the stirrer includes: sequentially obtaining data points in the historical rotational speed data time series of the latest data point of the stirrer based on the preset length of the forward analysis data segment to obtain the forward analysis data segment of the latest data point.

[0037] Among them, the length of the forward analysis data segment can be set to 300; the length of the forward analysis data segment is the number of data points included in the forward analysis data segment; the length can be specifically set according to actual needs, and the embodiments of the present invention will not elaborate here.

[0038] Illustrate the construction process of the forward analysis data segment of the latest data point: The time series of the rotational speed data of the stirrer is distributed from left to right according to the acquisition moment, and the left adjacent data point of the latest data point is the data point closest to the latest data point. Take the left adjacent data point of the latest data point as the rightmost data point in the forward analysis data segment of the latest data point, and finally obtain the forward analysis data segment of the latest data point in the historical rotational speed data time series of the latest data point. If the acquisition frequency is 5 times per second and the length of the forward analysis data segment is 300, the acquisition duration of the forward analysis data segment of the obtained latest data point is 60 seconds.

[0039] It can be understood that there will be a buffer waiting time in the process from the start of the stirrer to the start of stirring, and the length of this buffer waiting time is at least greater than the length of the forward analysis data segment of the latest data point, so that the forward analysis data segment of each latest data point can be obtained completely.

[0040] It should be noted that the number of extreme points in the forward analysis data segment of the latest data point can reflect the severity of the speed change in the forward analysis data segment. The more extreme points there are in the forward analysis data segment, the more severe the speed change in the forward analysis data segment is, and the more maximum restarts are required. However, the conventional and stable speed change in the forward analysis data segment may also contain multiple extreme points. For example, factors such as the shape, size, and installation angle of the agitator blades will affect the flow pattern of the fluid. When the blades rotate, the flow rate and pressure of the fluid will undergo complex changes at the edges and corners of the blades, forming local eddies and shear forces, which will cause small fluctuations in the speed, which appear as extreme points in the speed data. However, this fluctuation is usually within a certain range and is a normal operating phenomenon.

[0041] Based on this, when the embodiment of the present invention evaluates the degree of need for the maximum number of restarts based on the number of extreme points in the forward analysis data segment of the latest data points of the agitator, the distribution characteristics of the extreme points can also be obtained. The worse the distribution regularity of the extreme points in the forward analysis data segment of the latest data points of the agitator, the higher the possibility that rare speed data changes exist in the forward analysis data segment of the latest data points. By combining the number of extreme points and the distribution regularity of extreme points, it is possible to accurately distinguish between conventional and stable speed data and rare speed data with drastic changes.

[0042] For example, in the embodiment of the present invention, the primary change factor of the latest data point is calculated, and the specific relationship can be seen from the following equation: ; For the The primary change factor for the latest data point, For the The number of extreme points in the forward analysis data segment of the latest data point, For the The number of adjacent extreme value point groups in the forward analysis data segment of the latest data point, For the The mean value of the intervals between adjacent extreme value points in the forward analysis data segment of the latest data point, For the Latest data point forward analysis data segment The interval between adjacent extreme value points in a group. is a linear normalization function.

[0043] It is understandable that Two adjacent extreme value points in the forward analysis data segment of the latest data point are an adjacent extreme value point group. The collection interval between adjacent extreme value points is the absolute value of the difference in collection time. The final obtained The number of adjacent extreme value point groups in the forward analysis data segment of the latest data point .

[0044] In the above formula, The larger the The more drastic the change in the speed data in the forward analysis data segment of the latest data point is, the more maximum restart times are required to explore the solution space, and the larger the corresponding primary change factor is.

[0045] Indicates The variance of the adjacent extreme point collection intervals in the forward analysis data segment of the latest data point quantifies the The regularity of the distribution of extreme points in the forward analysis data segment of the latest data point. The larger the value, the more The worse the regularity of the distribution of extreme points in the forward analysis data segment of the latest data point, the higher the possibility of rare speed data, the greater the corresponding maximum number of restarts required, and the greater the primary change factor.

[0046] In summary, The more extreme points there are in the forward analysis data segment of the latest data point and the higher the regularity of the distribution of the extreme points, the larger the primary change factor of the latest data point is, and the higher the demand for the maximum number of restarts for the corresponding latest data point is.

[0047] The primary change factor of the latest data point characterizes the data change intensity in its forward analysis data segment. However, if there is error data in the forward analysis data segment of the latest data point, the speed data change intensity will also be high, and the accuracy of the primary change factor of the latest data point will be low.

[0048] It should be noted that there is a certain correlation between the speed of the agitator and the motor power. During the operation of the agitator, when the speed increases, the motor needs to provide more power to overcome the resistance during the stirring process, and the power value will also increase. That is, there will be a good correlation between the speed data of the agitator and the motor power data over a period of time.

[0049] Based on this, the embodiment of the present invention can obtain the speed data and motor power data of the agitator, and verify the authenticity of the primary change factor of the latest data point through the correlation between the two, so as to accurately obtain the degree of demand for the maximum number of restarts of the agitator at the latest data point, that is, continue to execute the following steps.

[0050] S3: Obtain the motor power data segment corresponding to the forward analysis data segment of the latest data point, and calculate the target change factor of the latest data point based on the primary change factor of the latest data point, the difference between the cumulative sum of the collection time of the extreme value points in the forward analysis data segment of the latest data point and the cumulative sum of the collection time of the extreme value points in the corresponding motor power data segment.

[0051] It should be noted that if the forward analysis data segment of the latest data point has a high correlation with the corresponding motor power data segment, it indicates that the primary change factor of the latest data point obtained based on the above steps is highly authentic, and the possibility of error data is smaller. On the contrary, if the forward analysis data segment of the latest data point has a low correlation with the corresponding motor power data segment, it indicates that due to the influence of error data, the change intensity reflected by the forward analysis data segment of the latest data point may be higher than its actual change intensity, resulting in an overestimation, and the primary change factor of the latest data point needs to be reduced.

[0052] However, there are differences in the dynamic responses between the mechanical system of the agitator and the electrical system of the motor, resulting in a delay in the power change of the motor. Therefore, when obtaining the motor power data segment corresponding to the forward analysis data segment of the latest data point, it is necessary to eliminate the time difference between the two first.

[0053] Exemplarily, in the embodiment of the present invention, obtaining the motor power data segment corresponding to the forward analysis data segment of the latest data point includes: taking each acquisition time and the corresponding motor power value as a power data point to obtain a power data time series; determining the motor power data segment corresponding to the forward analysis data segment of the latest data point according to the delay time between the rotational speed data and the motor power; wherein the lengths of the forward analysis data segment of the latest data point and the corresponding motor power data segment are the same.

[0054] Wherein the lengths of the forward analysis data segment of the latest data point and the corresponding motor power data segment are the same, and the data points in the forward analysis data segment of the latest data point correspond one by one to the power data points in the motor power data segment.

[0055] Specifically, when obtaining the delay time between the rotational speed data and the motor power, a rotational speed data reference segment and a power data reference segment can be obtained within a preset time period of historical data, and the Pearson correlation coefficient between the rotational speed data reference segment and the power data reference segment is determined; then, the start acquisition time of the power data reference segment is moved forward by one acquisition time to obtain a new power data reference segment, and the Pearson correlation coefficient between the rotational speed data reference segment and the new power data reference segment is calculated continuously in turn until the end acquisition time of the new power data reference segment is the same as the start acquisition time of the rotational speed data reference segment, and the loop is stopped. The absolute value of the difference between the acquisition times of the central data points of the rotational speed data reference segment and the new power data reference segment corresponding to the maximum Pearson correlation coefficient is taken as the delay time between the rotational speed data and the motor power.

[0056] Among them, the preset time period can be set to 50 seconds; the preset time period can be specifically set according to actual needs, and the embodiment of the present invention does not limit it too much here.

[0057] For example, when determining the motor power data segment corresponding to the forward analysis data segment of the latest data point according to the delay time between the rotational speed data and the motor power, the sum of the starting acquisition time of the forward analysis data segment of the latest data point and the delay time can be used as the starting acquisition time of the motor power data segment corresponding to the forward analysis data segment of the latest data point, and finally the motor power data segment corresponding to the forward analysis data segment of the latest data point is obtained.

[0058] After obtaining the motor power data segment corresponding to the forward analysis data segment of the latest data point based on the above steps, the difference between the acquisition time of the extreme value point of the rotational speed data and the acquisition time of the extreme value point of the motor power can reflect the synchronization of the two in time. The smaller the difference, the higher the correlation between the two.

[0059] For example, in the embodiment of the present invention, to calculate the target change factor of the latest data point, the following relational expression can be specifically referred to: ; is the target change factor of the latest data point, is the primary change factor of the latest data point, is the acquisition time of the th extreme value point in the forward analysis data segment of the latest data point, is the number of extreme value points in the forward analysis data segment of the latest data point, is the number of extreme value points in the motor power data segment corresponding to the forward analysis data segment of the latest data point, is the acquisition time of the th extreme value point in the motor power data segment corresponding to the forward analysis data segment of the latest data point, is a hyperparameter, is a linear normalization function, is an absolute value symbol.

[0060] Among them, the hyperparameter is to avoid being 0, which may cause the formula to be meaningless, and it can be set to 0.1. Specifically, it can be set according to actual needs.

[0061] In the above formula, quantifies the correlation between the forward analysis data segment of the latest data point and the corresponding motor power data segment. The larger this value is, the worse the correlation, indicating that the forward analysis data segment of the latest data point is more likely to contain error data. Based on the above steps, the The rotation speed change severity reflected by the primary change factor of the latest data point is greater than the actual one. The real The change degree of the forward analysis data segment of the latest data point is smaller, and the corresponding final target change factor is also smaller.

[0062] Based on the above steps, the target change factor of the latest data point can be obtained. The target change factor of the latest data point is used to characterize the degree of demand of the latest data point for the maximum restart times. Based on this, the preset maximum restart times of the latest data point can be accurately adjusted.

[0063] S4: Adjust the preset maximum restart times in the random restart hill climbing algorithm based on the target change factor of the latest data point to achieve intelligent control of the stirrer.

[0064] It can be understood that the larger the target change factor of the latest data point, the higher the credibility that there are rare rotation speed changes in the forward analysis data segment of the latest data point, and the higher the required maximum restart times to ensure accurate search for the optimal solution. Based on this, the maximum restart times of the latest data point can be accurately obtained.

[0065] Exemplarily, in the embodiment of the present invention, adjusting the preset maximum restart times in the random restart hill climbing algorithm based on the target change factor of the latest data point, specifically, the following relational expression can be referred to: ; is the maximum restart times of the latest data point, is the target change factor of the latest data point, is the preset adjustment coefficient, is the preset maximum restart times, is the ceiling symbol.

[0066] Among them, the preset adjustment coefficient can be set to 0.5; the preset maximum restart times can be set to 5. The values of the preset adjustment coefficient and the preset maximum restart times can be specifically set according to actual needs.

[0067] After obtaining the maximum restart times of the latest data point based on the above steps, the maximum restart times of the latest data point can be used in the random restart hill climbing algorithm, so as to accurately obtain the rotation speed value required by the stirrer at the acquisition moment corresponding to the latest data point.

[0068] Exemplarily, in the embodiment of the present invention, constructing an objective function in the time series of the historical rotation speed data of the latest data point, specifically, the following relational expression can be referred to: ; is the objective function value of the solution, The motor power value required for this solution, The remaining stirring time required when using this solution to stir the product, is a linear normalization function.

[0069] It can be understood that the solution is the possible rotational speed value. The motor power value required for each solution and the remaining stirring time required when the solution is used to stir the product can be obtained from the historical rotational speed data according to the prior art, and this is not elaborated in the embodiments of the present invention.

[0070] In the above formula, the larger the motor power value required for the solution, the greater the energy consumption corresponding to this solution, the lower its preference level, and the larger the objective function value. The larger the remaining stirring time required when the solution is used to stir the product, the lower the working efficiency when this solution is applied to the current chemical product, and the lower its preference level. Therefore, the minimum objective function value corresponding to the local optimal solution obtained within the maximum number of restarts is the solution with the highest preference level, and this local optimal solution is also the global optimal solution.

[0071] Exemplarily, in the embodiments of the present invention, the intelligent control of the stirrer is realized based on the global optimal solution of the latest data point, including: obtaining the local optimal solution of the latest data point and the corresponding objective function value in each restart, and obtaining the local optimal solution corresponding to the minimum objective function value within the maximum number of restarts as the global optimal solution of the latest data point; realizing the intelligent control of the stirrer according to the global optimal solution.

[0072] Specifically, before obtaining the optimal solution, the range of the solution randomly generated by the random restart hill climbing algorithm can be set to 100 rpm to 500 rpm; the random perturbation range in each iteration during the optimal solution output process is preset to the current solution value range, that is, a new solution is generated within the current solution value range; the number of iterations during the optimal solution output is 20 times.

[0073] After obtaining the global optimal solution of the latest data point based on the above steps, the rotational speed value of the latest data point can be adjusted to the global optimal solution.

[0074] Exemplarily, in the embodiments of the present invention, the intelligent control of the stirrer is realized according to the global optimal solution, including: adjusting the rotational speed value of the stirrer to the rotational speed value corresponding to the global optimal solution based on PID.

[0075] Specifically, the PID controller obtains the difference between the global optimal solution of the latest data point and the rotational speed value of this latest data point and outputs a control signal; the control signal output by the PID controller acts on the power supply frequency of the drive motor of the stirrer, thereby realizing the intelligent regulation of the rotational speed of the stirrer.

[0076] It can be seen that in the embodiments of the present invention, when implementing intelligent control of a device, a forward analysis data segment of the latest data point can be constructed in the historical rotational speed data time series of the latest data point of the stirrer; calculate the primary change factor of the latest data point : ; and are respectively the number of extreme points and the number of adjacent extreme point groups in the forward analysis data segment of the latest data point, is the average acquisition interval of adjacent extreme points in the forward analysis data segment of the latest data point, is the acquisition interval of the th group of adjacent extreme points in the forward analysis data segment of the latest data point, is a linear normalization function; obtain the motor power data segment corresponding to the forward analysis data segment of the latest data point, and calculate the target change factor of the latest data point according to the difference between the primary change factor of the latest data point, the cumulative sum of the acquisition times of the extreme points in the forward analysis data segment of the latest data point, and the cumulative sum of the acquisition times of the extreme points in the corresponding motor power data segment; adjust the preset maximum restart times in the random restart hill climbing algorithm based on the target change factor of the latest data point to achieve intelligent control of the stirrer, effectively improving the accuracy of intelligent regulation of the stirrer speed.

[0077] The embodiments of the present invention also disclose an intelligent control system for chemical production equipment, including a processor and a memory. The memory stores computer program instructions, and when the computer program instructions are executed by the processor, an intelligent control method for chemical production equipment provided by the present invention is implemented.

[0078] The above control system also includes other components well-known to those skilled in the art such as a communication bus and a communication interface, and their settings and functions are known in the art, so they will not be described in detail here.

[0079] The above are all the preferred embodiments of the present invention. The protection scope of the present invention is not limited by this. Therefore, all equivalent changes made according to the structure, shape, and principle of the present invention should be covered within the protection scope of the present invention.​

Claims

1. An intelligent control method for chemical production equipment, characterized in that, Including: Construct a forward analysis data segment of the latest data point in the historical rotational speed data time series of the agitator; Calculate the primary change factor of the latest data point : ； , are respectively the number of extreme points and the number of adjacent extreme point groups in the forward analysis data segment of the latest data point, is the average acquisition interval of adjacent extreme points in the forward analysis data segment of the latest data point, is the acquisition interval of the th group of adjacent extreme points in the forward analysis data segment of the latest data point, is a linear normalization function;​ Obtain the motor power data segment corresponding to the forward analysis data segment of the latest data point, and calculate the target change factor of the latest data point according to the primary change factor of the latest data point and the difference between the cumulative sum of the acquisition times of the extreme points in the forward analysis data segment of the latest data point and the cumulative sum of the acquisition times of the extreme points in the corresponding motor power data segment; Adjust the preset maximum restart times in the random restart hill climbing algorithm based on the target change factor of the latest data point to achieve intelligent control of the agitator.

2. The intelligent control method for a chemical production device according to claim 1, characterized in that Before constructing the forward analysis data segment of the latest data point in the historical rotational speed data time series of the agitator, it further includes: Obtain the rotational speed value of the agitator at each acquisition time as a data point, and obtain the rotational speed data time series after preprocessing.

3. An intelligent control method for a chemical production device according to claim 1, characterized in that, Constructing the forward analysis data segment of the latest data point in the historical rotational speed data time series of the agitator includes: Successively obtain data points in the historical rotational speed data time series of the latest data point of the agitator based on the preset length of the forward analysis data segment to obtain the forward analysis data segment of the latest data point.

4. A method for intelligent control of a chemical production device according to claim 1, characterized in that, Obtaining the motor power data segment corresponding to the forward analysis data segment of the latest data point includes: Take each acquisition time and the corresponding motor power value as a power data point to obtain the power data time series; determine the motor power data segment corresponding to the forward analysis data segment of the latest data point according to the delay time between the rotational speed data and the motor power; where the lengths of the forward analysis data segment of the latest data point and the corresponding motor power data segment are the same.

5. A method for intelligent control of a chemical production device according to claim 1, characterized in that, Calculating the target change factor of the latest data point includes: Calculate the target change factor for the latest data point : ； is the primary change factor of the latest data point, is the acquisition time of the th extreme point in the forward analysis data segment of the latest data point, and are respectively the number of extreme points in the forward analysis data segment of the latest data point and the corresponding motor power data segment, is the acquisition time of the th extreme point in the motor power data segment corresponding to the forward analysis data segment of the latest data point, is a hyperparameter, is a linear normalization function, is the absolute value symbol.

6. A method for intelligent control of a chemical production device according to claim 1, characterized in that, Adjusting the preset maximum restart times in the random restart hill climbing algorithm based on the target change factor of the latest data point includes: ； is the maximum number of restarts for the latest data point, is the target change factor for the latest data point, is the preset adjustment coefficient, is the preset maximum number of restarts, is the ceiling symbol.

7. An intelligent control method for a chemical production device according to claim 1, characterized in that Adjusting the preset maximum restart times in the random restart hill climbing algorithm based on the target change factor of the latest data point to achieve intelligent control of the agitator includes: Obtain the local optimal solution of the latest data point and the corresponding objective function value in each restart, and obtain the local optimal solution corresponding to the minimum objective function value in the maximum restart times as the global optimal solution of the latest data point; achieve intelligent control of the agitator according to the global optimal solution.

8. An intelligent control method for a chemical production device according to claim 7, characterized in that, The method for obtaining the objective function value includes: Construct an objective function in the historical rotational speed data time series of the latest data point; ； is the objective function value of the solution, is the motor power value required for this solution, is the remaining stirring time required when this solution is used to stir the product, is the linear normalization function.

9. An intelligent control method for a chemical production device according to claim 7, characterized in that, Achieving intelligent control of the agitator according to the global optimal solution includes: Adjust the rotational speed value of the agitator to the rotational speed value corresponding to the global optimal solution based on PID.

10. An intelligent control system for chemical production equipment, characterized in that, Including: A processor and a memory, the memory stores computer program instructions, and when the computer program instructions are executed by the processor, it implements an intelligent control method for a chemical production equipment according to any one of claims 1-9.

Images