CN115430815A - Crystallizer liquid level control method and device, electronic equipment and storage medium - Google Patents
Crystallizer liquid level control method and device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN115430815A CN115430815A CN202210951137.XA CN202210951137A CN115430815A CN 115430815 A CN115430815 A CN 115430815A CN 202210951137 A CN202210951137 A CN 202210951137A CN 115430815 A CN115430815 A CN 115430815A
- Authority
- CN
- China
- Prior art keywords
- liquid level
- crystallizer
- data
- fluctuation
- influence
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 239000007788 liquid Substances 0.000 title claims abstract description 280
- 238000000034 method Methods 0.000 title claims abstract description 57
- 238000003860 storage Methods 0.000 title claims abstract description 8
- 230000009467 reduction Effects 0.000 claims description 82
- 238000012549 training Methods 0.000 claims description 81
- 238000012360 testing method Methods 0.000 claims description 70
- 238000004891 communication Methods 0.000 claims description 15
- 238000012545 processing Methods 0.000 claims description 10
- 238000004590 computer program Methods 0.000 claims description 7
- 229910000831 Steel Inorganic materials 0.000 abstract description 36
- 239000010959 steel Substances 0.000 abstract description 36
- 230000006870 function Effects 0.000 description 36
- 239000000203 mixture Substances 0.000 description 25
- 238000005266 casting Methods 0.000 description 16
- 238000001228 spectrum Methods 0.000 description 12
- 238000010079 rubber tapping Methods 0.000 description 6
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 5
- 238000011156 evaluation Methods 0.000 description 4
- 238000003064 k means clustering Methods 0.000 description 4
- 238000010801 machine learning Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 230000008859 change Effects 0.000 description 3
- 238000010586 diagram Methods 0.000 description 3
- 238000004519 manufacturing process Methods 0.000 description 3
- TWNQGVIAIRXVLR-UHFFFAOYSA-N oxo(oxoalumanyloxy)alumane Chemical compound O=[Al]O[Al]=O TWNQGVIAIRXVLR-UHFFFAOYSA-N 0.000 description 3
- 238000000513 principal component analysis Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 238000007621 cluster analysis Methods 0.000 description 2
- 238000009749 continuous casting Methods 0.000 description 2
- 230000007797 corrosion Effects 0.000 description 2
- 238000005260 corrosion Methods 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 238000009826 distribution Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000016615 flocculation Effects 0.000 description 2
- 238000005189 flocculation Methods 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 238000012544 monitoring process Methods 0.000 description 2
- 230000000737 periodic effect Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000004220 aggregation Methods 0.000 description 1
- 230000002776 aggregation Effects 0.000 description 1
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 description 1
- 229910052782 aluminium Inorganic materials 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- QVGXLLKOCUKJST-UHFFFAOYSA-N atomic oxygen Chemical compound [O] QVGXLLKOCUKJST-UHFFFAOYSA-N 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000013507 mapping Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 229910052760 oxygen Inorganic materials 0.000 description 1
- 239000001301 oxygen Substances 0.000 description 1
- 239000000843 powder Substances 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 238000013441 quality evaluation Methods 0.000 description 1
- 230000033764 rhythmic process Effects 0.000 description 1
- 230000000630 rising effect Effects 0.000 description 1
- 239000002893 slag Substances 0.000 description 1
- 239000002436 steel type Substances 0.000 description 1
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B22—CASTING; POWDER METALLURGY
- B22D—CASTING OF METALS; CASTING OF OTHER SUBSTANCES BY THE SAME PROCESSES OR DEVICES
- B22D11/00—Continuous casting of metals, i.e. casting in indefinite lengths
- B22D11/16—Controlling or regulating processes or operations
- B22D11/18—Controlling or regulating processes or operations for pouring
- B22D11/181—Controlling or regulating processes or operations for pouring responsive to molten metal level or slag level
Abstract
The application relates to a crystallizer liquid level control method, a crystallizer liquid level control device, electronic equipment and a storage medium, wherein the method comprises the following steps: acquiring influence data of the liquid level of the crystallizer, wherein the influence data is determined according to influence factors of the liquid level of the crystallizer; inputting the influence data into a crystallizer liquid level fluctuation factor determination model, and determining the fluctuation state of the liquid level of the crystallizer; and adjusting the influence factors of the crystallizer according to the fluctuation state to control the liquid level of the crystallizer, wherein the influence data is obtained through the determined influence factors, and the fluctuation state of the liquid level of the crystallizer is determined based on the influence data, so that the fluctuation state of the liquid level of the crystallizer is determined by only detecting the liquid level height of the molten steel, and meanwhile, the influence factors are adjusted according to the fluctuation state determined by the influence data, so that the problem of inaccurate control caused by only detecting the liquid level height of the molten steel to control the liquid level of the crystallizer is avoided, and further, the control accuracy is improved.
Description
Technical Field
The present application relates to the field of crystallizer technologies, and in particular, to a method and an apparatus for controlling a liquid level of a crystallizer, an electronic device, and a storage medium.
Background
In the continuous casting production process of steel, the liquid level control of a crystallizer of a continuous casting machine is an important link of production automation, and the control precision directly influences the quality of a casting blank. If the liquid level fluctuation exceeds +/-5 mm, certain influence is caused on the quality of a casting blank, even unmelted casting powder is rolled into molten steel to form casting blank surface defects; if the liquid level fluctuation exceeds +/-10 mm, the automatic casting mode can be interrupted, the normal production rhythm is influenced, and serious accidents such as bonding, steel leakage and the like can even occur in severe cases.
The existing crystallizer liquid level control system mainly depends on a liquid level detection device to detect the liquid level height of molten steel for control, but the method only judges according to the detected liquid level height, and actually, the liquid level height of the molten steel is influenced by various factors, such as the molten steel flow rate, the angle, the height and the like of a steel outlet of a crystallizer can influence the liquid level fluctuation, and only the liquid level height of the molten steel is detected to control the liquid level of the crystallizer to cause inaccurate control, so how to more finely control the liquid level of the crystallizer becomes the problem which needs to be solved urgently.
It is to be noted that the information disclosed in the above background section is only for enhancement of understanding of the background of the present invention and therefore may include information that does not constitute prior art known to a person of ordinary skill in the art.
Disclosure of Invention
The application provides a crystallizer liquid level control method, a crystallizer liquid level control device, electronic equipment and a storage medium, and aims to solve the problem that control is inaccurate due to the fact that only the liquid level height of molten steel is detected to control the liquid level of a crystallizer in the related art.
In a first aspect, the present application provides a crystallizer liquid level control method, including: acquiring influence data of the liquid level of the crystallizer, wherein the influence data is determined according to influence factors of the liquid level of the crystallizer; inputting the influence data into a crystallizer liquid level fluctuation factor determination model, and determining the fluctuation state of the liquid level of the crystallizer; and adjusting the influence factors of the crystallizer according to the fluctuation state so as to control the liquid level of the crystallizer.
In some examples, before inputting the impact data into a crystallizer liquid level fluctuation factor determination model, the method further comprises: acquiring historical influence data, and determining a normal sample and a fluctuation sample according to the historical influence data; determining a training set and a testing set based on the normal sample and the fluctuation sample; and modeling the training set and the test set through a clustering algorithm to obtain a crystallizer liquid level fluctuation factor determination model.
In some examples, the historical impact data includes: actual liquid level and set liquid level; wherein, according to historical influence data, determining a normal sample and a fluctuation sample comprises: determining the difference value between the actual liquid level and the set liquid level at each moment in the historical influence data; when the difference value between the actual liquid level and the set liquid level at any moment in the historical influence data exceeds a first threshold value, taking the data at any moment as the fluctuation sample; and when the difference value between the actual liquid level and the set liquid level at any moment in the historical influence data is lower than a second threshold value, taking the data at any moment as the normal sample, wherein the second threshold value is lower than the first threshold value.
In some examples, determining a training set and a test set based on the normal sample and the fluctuation sample includes: mixing the normal sample and the fluctuation sample to obtain a total training set; and dividing the total training set based on a preset proportion to obtain the training set and the test set.
In some examples, modeling the training set and the test set by a clustering algorithm to obtain the crystallizer liquid level fluctuation factor determination model includes: respectively extracting time domain characteristics and frequency domain characteristics of variable variables in the training set and the test set to obtain data dimensions of the training set and data dimensions of the test set; performing dimensionality reduction processing on the data dimensionality of the training set and the data dimensionality of the test set to obtain training set dimensionality reduction data and test set dimensionality reduction data; and modeling the dimension reduction data of the training set and the dimension reduction data of the testing set through a clustering algorithm to obtain a crystallizer liquid level fluctuation factor determination model.
In some examples, modeling the training set dimension reduction data and the test set dimension reduction data through a clustering algorithm to obtain the crystallizer liquid level fluctuation factor determination model includes: modeling the training set dimensionality reduction data and the testing set dimensionality reduction data through at least two clustering algorithms to obtain at least two initial crystallizer liquid level fluctuation factor determination models; evaluating at least two initial crystallizer liquid level fluctuation factor determination models based on the normal sample to obtain an optimal initial crystallizer liquid level fluctuation factor determination model; and taking the optimal initial crystallizer liquid level fluctuation factor determination model as the crystallizer liquid level fluctuation factor determination model.
In some examples, after the optimal initial crystallizer liquid level fluctuation factor determination model is used as the crystallizer liquid level fluctuation factor determination model, the method further comprises: determining clustering conditions of classifying the training set dimension reduction data and the test set dimension reduction data by a model according to the crystallizer liquid level fluctuation factors to obtain multiple preset fluctuation states; and setting an influence factor adjusting method corresponding to the preset fluctuation state.
In a second aspect, the present application provides a crystallizer liquid level control device, comprising: the acquisition module is used for acquiring influence data of the liquid level of the crystallizer, and the influence data is determined according to influence factors of the liquid level of the crystallizer; the input module is used for inputting the influence data into a crystallizer liquid level fluctuation factor determination model and determining the fluctuation state of the liquid level of the crystallizer; and the control module is used for adjusting the influence factors of the crystallizer according to the fluctuation state so as to control the liquid level of the crystallizer.
In a third aspect, an electronic device is provided, which includes a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory complete communication with each other through the communication bus;
a memory for storing a computer program;
and the processor is used for realizing the steps of the crystallizer liquid level control method in any embodiment of the first aspect when executing the program stored in the memory.
In a fourth aspect, a computer-readable storage medium is provided, on which a computer program is stored, which computer program, when being executed by a processor, is adapted to carry out the steps of the crystallizer liquid level control method according to any of the embodiments of the first aspect.
Compared with the prior art, the technical scheme provided by the embodiment of the application has the following advantages:
the crystallizer liquid level control method provided by the embodiment of the application comprises the following steps: acquiring influence data of the liquid level of the crystallizer, wherein the influence data is determined according to influence factors of the liquid level of the crystallizer; inputting the influence data into a crystallizer liquid level fluctuation factor determination model, and determining the fluctuation state of the liquid level of the crystallizer; the influence factors of the crystallizer are adjusted according to the fluctuation state to control the liquid level of the crystallizer, wherein the influence data are obtained through the determined influence factors, the fluctuation state of the liquid level of the crystallizer is determined based on the influence data, the fluctuation state of the liquid level of the crystallizer is determined by only detecting the liquid level height of the molten steel, meanwhile, the influence factors are adjusted according to the fluctuation state determined by the influence data, the liquid level of the crystallizer is accurately controlled, the problem that the liquid level of the crystallizer is controlled inaccurately by only detecting the liquid level height of the molten steel is avoided, the control accuracy is improved, the quality of a casting blank is improved, major accidents such as adhesion, steel leakage and the like can be avoided, and the safety is improved.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the invention and together with the description, serve to explain the principles of the invention.
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious for those skilled in the art that other drawings can be obtained according to the drawings without inventive exercise.
FIG. 1 is a schematic diagram of a basic flow chart of an alternative crystallizer liquid level control method provided by an embodiment of the present application;
FIG. 2 is a schematic diagram of a basic flow of an alternative crystallizer liquid level control device provided in an embodiment of the present application;
fig. 3 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the embodiments of the present application clearer, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some embodiments of the present application, but not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
In order to solve the problem of inaccurate control caused by only detecting the liquid level height of molten steel to control the crystallizer in the related art, please refer to fig. 1, fig. 1 is a crystallizer liquid level control method provided by this embodiment, as shown in fig. 1, the crystallizer liquid level control method includes:
s101, obtaining influence data of the crystallizer liquid level, wherein the influence data are determined according to influence factors of the crystallizer liquid level;
s102, inputting the influence data into a crystallizer liquid level fluctuation factor determination model, and determining the fluctuation state of the liquid level of the crystallizer;
s103, adjusting influence factors of the crystallizer according to the fluctuation state so as to control the liquid level of the crystallizer.
The main influence factors of the crystallizer liquid level fluctuation can be analyzed in two directions above the crystallizer (including the crystallizer) and below the crystallizer by taking the crystallizer as a dividing point. The direction above the crystallizer (including the crystallizer) can be considered from three aspects of molten steel flow, angle and height of a steel outlet of the crystallizer;
taking the tapping hole flow as an example, the main influencing factors of the tapping hole flow include but are not limited to: (1) the molten steel quality influences the fluidity of the molten steel; (2) The opening degree of the stopper rod influences the outflow speed of molten steel, so that the pulling speed of the withdrawal and straightening machine is influenced; 3) The change of the weight of the tundish influences the change of the molten steel flow; (4) The accretion and corrosion of the stopper rod and the water gap also have certain influence on the molten steel flow at the steel outlet; (5) The clearance and the blockage of the stopper mechanism also influence the liquid level control, thereby influencing the flow of the steel tapping hole.
Taking the angle of the steel tapping hole as an example, when the aluminum content in steel is high, the steel is easy to react with oxygen to generate aluminum oxide, the aluminum oxide is high-melting-point inclusion, and the aluminum oxide is slowly flocculated and collected at the rod head, the water gap, the bowl part, the inner wall and other parts of the stopper rod along with the extension of casting time, so that the angle of the steel tapping hole is influenced; along with the extension of the casting time, the stopper rod and the water gap can generate corrosion phenomena, and the liquid level in the crystallizer can also fluctuate by influencing the angle of the steel tapping hole.
Taking the height of the tap hole as an example, the zero drift phenomenon of the crystallizer liquid level detection system and the thickness of the crystallizer covering slag layer further cause the fluctuation of the liquid level of the crystallizer by influencing the height of the tap hole. It can be understood that the bulging of the cast strand is also liable to cause the fluctuation of the liquid level of the crystallizer. Before the casting blank is not completely solidified, the deviation of the roller surface of one or a plurality of fan-shaped sections is large, so that the liquid core in the casting blank is stressed unevenly to cause the fluctuation of the liquid level. In addition, the withdrawal speed of the withdrawal and straightening machine is unstable, and the liquid level fluctuation is easy to cause;
therefore, as can be seen from the above examples, the stopper, the tundish, the withdrawal and straightening unit, the billet bulging and the like all have influence factors on the liquid level of the crystallizer, and the determination of the influence data influencing the crystallizer based on the influence factors includes but is not limited to: the system comprises an actual liquid level, a set liquid level, an actual pulling speed, a given pulling speed, an actual position of a stopper rod, a given position of the stopper rod, the weight of a tundish, a section number, a bulging compensation amount, an actual liquid level (electromagnetic) electromagnetic instrument N value, an electromagnetic instrument NP value, a sensor temperature, a steel grade number and casting starting time.
The above example is followed, wherein obtaining the influence data of the crystallizer liquid level comprises: and acquiring influence data of the crystallizer liquid level in real time, wherein the influence data of the crystallizer liquid level are acquired in real time by monitoring the influence factors in real time. In some examples of this embodiment, obtaining the influence data of the crystallizer liquid level comprises: and periodically acquiring influence data of the crystallizer liquid level, wherein the influence factors are periodically monitored, and then the influence data of the crystallizer liquid level are acquired in real time.
In some examples of this embodiment, before inputting the influence data into the crystallizer liquid level fluctuation factor determination model, the method further comprises: acquiring historical influence data, and determining a normal sample and a fluctuation sample according to the historical influence data; determining a training set and a testing set based on the normal sample and the fluctuation sample; and modeling the training set and the test set through a clustering algorithm to obtain a crystallizer liquid level fluctuation factor determination model. It should be understood that the historical impact data is data over one or more periods of time, which includes impact data over one or more periods of time. A plurality of normal samples and fluctuation samples can be determined through the historical influence data, then a training set and a testing set can be determined based on the normal samples and the fluctuation samples, and finally modeling is carried out based on the training set and the testing set to obtain a crystallizer liquid level fluctuation factor determination model.
Bearing the above example, the historical impact data includes: actual liquid level and set liquid level; wherein, according to historical influence data, determining a normal sample and a fluctuation sample comprises: determining the difference value between the actual liquid level and the set liquid level at each moment in the historical influence data; when the difference value between the actual liquid level and the set liquid level at any moment in the historical influence data exceeds a first threshold value, taking the data at any moment as the fluctuation sample; and when the difference value between the actual liquid level and the set liquid level at any moment in the historical influence data is lower than a second threshold value, taking the data at any moment as the normal sample, wherein the second threshold value is lower than the first threshold value. The actual liquid level is the actual value of the current liquid level of the crystallizer, the set liquid level is a liquid level value set for the crystallizer, and the difference value between the actual liquid level and the set liquid level can be obtained by calculating the difference value between the actual liquid level and the set liquid level; the first threshold value and the second threshold value are values set by the relevant personnel according to actual requirements, and the second threshold value is lower than the first threshold value, for example, the first threshold value is 5mm, and the second threshold value is 3mm. And if the absolute value of the deviation between the actual liquid level and the set liquid level is not higher than 3mm, the data at the current moment are taken as normal samples.
Taking the above example as an example, in some examples, when the difference between the actual liquid level and the set liquid level at any time in the historical influence data exceeds the first threshold, taking the data at any time as the fluctuation sample, including: when the difference value between the actual liquid level and the set liquid level at any moment in the historical influence data exceeds a first threshold value, taking the time at any moment as a time point, obtaining a sample interval, and taking the data of the sample interval as a fluctuation sample, wherein the sample interval is determined according to set first preset time and second preset time; specifically, taking an example that a first preset time period is five minutes and a second preset time period is 1 minute, when it is determined that a difference value between an actual liquid level at the current time and a set liquid level exceeds a first threshold value, taking an interval between five minutes before the current time and one minute after the current time as a sample interval, and acquiring data of the sample interval as a fluctuation sample; similarly, when the difference between the actual liquid level and the set liquid level at any time in the historical influence data is lower than a second threshold, taking the data at any time as the normal sample, including: when the difference value between the actual liquid level and the set liquid level at any moment in the historical influence data is lower than a second threshold value, taking the time of the any moment as a time point, obtaining a sample interval, and taking the data of the sample interval as a fluctuation sample, wherein the sample interval is determined according to a set third preset time and a set fourth preset time; specifically, the third preset time period is 4 minutes, the fourth preset time period is 4 minutes, and when it is determined that the difference between the actual liquid level at the current time and the set liquid level is lower than the second threshold, an interval between five minutes before the current time and one minute after the current time is used as a sample interval, and data of the sample interval is acquired and used as a normal sample; it should be appreciated that in some examples, the first predetermined period of time to the fourth predetermined period of time may vary; in some examples, the first to fourth preset durations are the same. And by analogy, dividing all historical influence data.
As an example, it can be understood that when the difference between the actual liquid level and the set liquid level at any time in the historical influence data exceeds the first threshold, a sample interval is obtained by using the time at any time as a time point, and when the data in the sample interval is used as a fluctuation sample, if a time b exists between a first preset time and any time except the time a and when the difference between another actual liquid level and the set liquid level exceeds the first threshold in the sample interval, the two times are merged into the same fluctuation sample, a second preset time from a first preset time before the time b to the time a after the time a is used as a sample interval, and the data in the sample interval is used as a fluctuation sample; for example, the first preset time period is five minutes, the second preset time period is one minute, if the absolute value of the liquid level deviation at the current time a is greater than 5mm, and a time b with the absolute value of the liquid level deviation greater than 5mm exists within 5 minutes before the current time a, the time a and the time b are merged into the same sample, the 5 minutes before the time b and 1 minute after the time a are taken as sample intervals, and data in the sample intervals are taken as sample data. And by analogy, dividing all historical influence data.
In some examples of this embodiment, determining a training set and a test set based on the normal sample and the fluctuation sample comprises: mixing the normal sample and the fluctuation sample to obtain a total training set; and dividing the total training set based on a preset proportion to obtain the training set and the test set. The preset ratio is a ratio of the training set to the test set, and the preset ratio is a ratio set by the relevant personnel according to actual requirements, for example, the preset ratio may be 3:1, that is, the training set: test set =3:1. And mixing the fluctuation sample with the normal sample to obtain a total training set, and dividing the total training set into a training set and a testing set according to the preset proportion.
In some examples of this embodiment, modeling the training set and the test set by a clustering algorithm to obtain the crystallizer liquid level fluctuation factor determination model includes: respectively extracting time domain characteristics and frequency domain characteristics of variable variables in the training set and the test set to obtain data dimensions of the training set and data dimensions of the test set; performing dimensionality reduction processing on the data dimensionality of the training set and the data dimensionality of the test set to obtain training set dimensionality reduction data and test set dimensionality reduction data; and modeling the dimension reduction data of the training set and the dimension reduction data of the testing set through a clustering algorithm to obtain a crystallizer liquid level fluctuation factor determination model. Wherein the time domain features include, but are not limited to: mean value, root mean square value, peak-to-peak value, kurtosis, skewness, peak factor, kurtosis factor, margin factor, waveform factor and pulse factor; frequency domain features include, but are not limited to: maximum amplitude frequency, power spectrum sum, power spectrum mean, power spectrum variance, power spectrum skewness, power spectrum kurtosis and power spectrum relative peak. It should be understood that, in some examples, the time-domain feature and the frequency-domain feature of the variable in the training set and the test set are extracted separately to obtain a feature sample, and the missing value is deleted from the obtained feature sample to obtain a training set data dimension and a test set data dimension.
It should be understood that the training set data dimension and the test set data dimension have more feature samples and higher dimensions, so the dimension reduction processing is performed on the training set data dimension and the test set data dimension. The dimension reduction processing mode is not limited, for example, the dimension reduction processing may be performed through a kernel function, where the kernel function includes but is not limited to: one of a Linear kernel function, a Poly kernel function, an RBF kernel function, a Sigmoid kernel function, and a Cosine kernel function, preferably, the kernel function is an RBF kernel function.
Bearing the above example, performing dimensionality reduction processing on the training set data dimensionality and the test set data dimensionality to obtain training set dimensionality reduction data and test set dimensionality reduction data, including but not limited to: and performing dimensionality reduction processing on the data dimensionality of the training set and the data dimensionality of the test set through at least two kernel functions to obtain at least two kinds of training set dimensionality reduction data and at least two kinds of test set dimensionality reduction data. It should be understood that after obtaining at least two training set dimension reduction data and at least two test set dimension reduction data, respectively performing quality evaluation on the various training set dimension reduction data, and taking the optimal training set dimension reduction data and test set dimension reduction data as the finally determined training set dimension reduction data and test set dimension reduction data. How to evaluate the quality of the training set dimension reduction data and the test set dimension reduction data to select the optimal kernel function will be described in detail later, and details are not repeated herein.
Considering that the fluctuation samples are label-free samples, a clustering algorithm in machine learning is selected to model the samples and is used for distinguishing different types of liquid level fluctuation, and the clustering algorithm includes but is not limited to: at least one of a Gaussian mixture model, a variational Bayes Gaussian mixture model, K-Means clustering and hierarchical clustering, preferably, the clustering algorithm adopts the Gaussian mixture model; meanwhile, the specific classification conditions of the training set dimension reduction data and the test set dimension reduction data are not known before modeling, so that the classification conditions can be set by related personnel, preferably, the classification conditions are set to be 6-15 classes, and 10 classes are set;
the previous example is carried out, modeling is carried out on the training set dimension reduction data and the testing set dimension reduction data through a clustering algorithm, and a crystallizer liquid level fluctuation factor determination model is obtained, and the method comprises the following steps: modeling the training set dimensionality reduction data and the testing set dimensionality reduction data through at least two clustering algorithms to obtain at least two initial crystallizer liquid level fluctuation factor determination models; evaluating at least two initial crystallizer liquid level fluctuation factor determination models based on the normal sample to obtain an optimal initial crystallizer liquid level fluctuation factor determination model; and taking the optimal initial crystallizer liquid level fluctuation factor determination model as the crystallizer liquid level fluctuation factor determination model. When at least two initial crystallizer liquid level fluctuation factor determination models are evaluated based on the normal sample, the quality of each initial crystallizer liquid level fluctuation factor determination model can be evaluated through the recall ratio and the precision ratio of the normal sample, wherein the precision ratio is as follows: in the predicted category, the correct proportion is predicted, and the denominator is the number of samples of the predicted category. And (4) recall ratio checking: in the true category, the correct proportion is predicted, and the denominator is the number of samples in the true category. For example, modeling is carried out on dimension reduction data of a training set and dimension reduction data of a test set respectively through a Gaussian mixture model, a variational Bayesian mixture model, K-Means clustering and hierarchical clustering to obtain four initial crystallizer liquid level fluctuation factor determination models, then the determination models of the four initial crystallizer liquid level fluctuation factors are judged through the recall ratio and the precision ratio of a normal sample to determine the advantages and the disadvantages of the determination models of the four initial crystallizer liquid level fluctuation factors, and the optimal initial crystallizer liquid level fluctuation factor determination model is used as the crystallizer liquid level fluctuation factor determination model.
Bearing the above example, when multiple kernel functions exist, and multiple training set dimension reduction data and multiple test set dimension reduction data are obtained according to the multiple kernel functions, modeling is respectively performed based on each training set dimension reduction data, test set dimension reduction data and each clustering algorithm, then the determination models of the four initial crystallizer liquid level fluctuation factors are evaluated through the recall ratio and the precision ratio of the normal sample, so as to determine the advantages and disadvantages of the determination models of the four initial crystallizer liquid level fluctuation factors, the optimal initial crystallizer liquid level fluctuation factor determination model is used as the crystallizer liquid level fluctuation factor determination model, and the kernel function corresponding to the crystallizer liquid level fluctuation factor determination model is used as the optimal kernel function.
In some embodiments, after the optimal initial crystallizer liquid level fluctuation factor determination model is used as the crystallizer liquid level fluctuation factor determination model, the method further comprises: determining clustering conditions of classifying the training set dimension reduction data and the test set dimension reduction data by a model according to the crystallizer liquid level fluctuation factors to obtain multiple preset fluctuation states; and setting an influence factor adjusting method corresponding to the preset fluctuation state. Wherein the preset fluctuation state includes: the method comprises the steps of starting pouring within 5 minutes, wherein liquid level fluctuation, low-frequency liquid level fluctuation, high-frequency liquid level fluctuation, liquid level fluctuation caused by a water changing port, divergent liquid level fluctuation, sudden-rising type liquid level fluctuation, liquid level fluctuation caused by flocculation flow and periodic liquid level fluctuation are positioned, specific influence factors of each liquid level fluctuation category are positioned based on a preset fluctuation state, influence factor adjusting methods corresponding to the fluctuation states are preset, and when a crystallizer liquid level fluctuation factor determining model determines the fluctuation state of the liquid level of a crystallizer, the corresponding influence factors can be adjusted according to the influence factor adjusting methods corresponding to the fluctuation states, so that the liquid level of the crystallizer is accurately controlled.
The crystallizer liquid level control method provided by the embodiment comprises the following steps: acquiring influence data of the liquid level of the crystallizer, wherein the influence data is determined according to influence factors of the liquid level of the crystallizer; inputting the influence data into a crystallizer liquid level fluctuation factor determination model, and determining the fluctuation state of the liquid level of the crystallizer; the influence factors of the crystallizer are adjusted according to the fluctuation state to control the liquid level of the crystallizer, wherein the influence data are obtained through the determined influence factors, the fluctuation state of the liquid level of the crystallizer is determined based on the influence data, the fluctuation state of the liquid level of the crystallizer is determined by only detecting the liquid level height of the molten steel, meanwhile, the influence factors are adjusted according to the fluctuation state determined by the influence data, the liquid level of the crystallizer is accurately controlled, the problem that the liquid level of the crystallizer is controlled inaccurately by only detecting the liquid level height of the molten steel is avoided, the control accuracy is improved, the quality of a casting blank is improved, major accidents such as adhesion, steel leakage and the like can be avoided, and the safety is improved.
In order to better understand the invention, the embodiment provides a more specific example to illustrate the invention;
the embodiment provides a crystallizer liquid level control method, which comprises the following steps:
1) Collecting the data of the influence of the liquid level fluctuation of the crystallizer,
wherein the impact data comprises: the actual liquid level, the set liquid level, the actual pulling speed, the given pulling speed, the actual position of the stopper rod, the given position of the stopper rod, the weight of a tundish, the serial number of a section, the drum compensation amount, the actual liquid level (electromagnetic) electromagnetic instrument N value, the electromagnetic instrument NP value, the temperature of a sensor, the steel grade number and the casting starting time are 14 variables;
specifically, wherein, for example, historical impact data is first collected over a period of time. And (3) dividing variables and irrelevant signal variables which are kept unchanged all the time to obtain 14 variables of an actual liquid level, a set liquid level, an actual pulling speed, a given pulling speed, an actual position of a stopper rod, a given position of the stopper rod, the weight of a tundish, a section serial number, a bulging compensation quantity, an actual liquid level (electromagnetic) electromagnetic instrument N value, an electromagnetic instrument NP value, a sensor temperature and a steel type number. Considering that the frequency of liquid level fluctuation is increased along with the increase of the casting time, a casting starting time label is manually added, and the data is collected to obtain historical influence data.
2) Preprocessing data;
processing historical influence data, and taking data 5 minutes before the current time a and 1 minute after the current time a as fluctuation samples when the absolute value of the deviation between the actual liquid level and the set liquid level is greater than 5 mm; if a moment b with the liquid level deviation absolute value larger than 5mm exists within 5 minutes before the current moment a, merging the moment a and the moment b into the same sample, and taking data of 5 minutes before the moment b and 1 minute after the moment a as fluctuation samples;
3) In the data, the fluctuation of the liquid level in each pouring time is within +/-3 mm, the sample is kept for about 8 minutes and is taken as a normal sample, and the rest samples are fluctuation samples; mixing the fluctuation sample with the normal sample, and according to the training set: test set =3:1, dividing a training set and a test set in proportion, and respectively extracting time domain characteristics of continuously variable of the training set and the test set: mean, root mean square value, peak-to-peak value, kurtosis, skewness, peak factor, kurtosis factor, margin factor, form factor, impulse factor, and frequency domain characteristics: the maximum amplitude frequency, the power spectrum sum, the power spectrum mean, the power spectrum variance, the power spectrum skewness, the power spectrum kurtosis and the power spectrum relative peak value are 17 characteristic factors, and missing values are deleted from the obtained characteristic samples to obtain the data dimension of a training set and the data dimension of the test set; as shown in table 1 below:
TABLE 1 training set data dimension and the test set data dimension
4) The data reduction and the cluster analysis are carried out,
as can be seen from table 1, there are 175 training set data dimensions and 175 test set data dimensions, which are higher, so that the dimensionality reduction is considered, and therefore, the dimensionality reduction needs to be performed;
respectively training set data dimensions and the test set data dimensions, applying different kernel functions to carry out principal component dimensionality reduction, classifying the data sets, selecting a clustering algorithm in machine learning to model the samples, evaluating the quality of a clustering model by using the recall ratio and precision ratio of normal samples, and preferably selecting the optimal dimensionality reduction kernel function;
wherein the principal component analysis is adapted to a linear dimensionality reduction of the data. And the kernel principal component analysis can realize the nonlinear dimensionality reduction of data and is used for processing a linear inseparable data set. The general idea of the kernel principal component analysis is as follows: for the matrix X in the input space, firstly, a nonlinear mapping is used to map all samples in X to a high-dimensional or even infinite-dimensional space, so that the matrix X is linearly separable, and then principal component dimensionality reduction is carried out in the high-dimensional space. Thus, the kernel functions provided by this example include, but are not limited to, a Linear kernel, a Poly kernel, an RBF kernel, a Sigmoid kernel, a Cosine kernel; the number of main components extracted by different kernel functions is shown in table 2:
TABLE 2 number of principal Components extracted by different Kernel functions
It should be understood that clustering algorithms include, but are not limited to: a Gaussian mixture model, a K-Means model, a variational Bayesian mixture model and a hierarchical clustering model;
when the data has a plurality of classes and is expected to be divided into some clusters, samples in different clusters can be assumed to respectively follow different Gaussian distributions, and the obtained clustering algorithm is called a Gaussian mixture model. The core idea of the gaussian mixture model is to assume that the data can be seen as generated from a plurality of gaussian distributions. Under this assumption, each individual partial model is a standard gaussian model, whose mean and variance are the parameters to be estimated. In addition, each partial model has a parameter, which can be understood as a weight or a probability of generating data. The parameters of the gaussian mixture model are typically solved using the EM algorithm.
The variational Bayes Gaussian mixture model is a variant of the Gaussian mixture model with a variational inference algorithm, can avoid the singularity which is usually generated in the expectation maximization EM algorithm, but can bring tiny deviation to the model.
K-Means is a clustering algorithm based on Euclidean distance, which considers that the closer the two targets are, the greater the similarity. The method comprises the steps of dividing data into K groups, randomly selecting K objects as initial clustering centers, calculating the distance between each object and each seed clustering center, and allocating each object to the nearest clustering center. The cluster center is recalculated for each sample assigned based on the existing objects in the cluster. This process is repeated until it is satisfied that none (or a minimum number) of objects are reassigned to a different cluster, or that none (or a minimum number) of cluster centers are changed again, and that the sum of squared errors is locally minimal.
The hierarchical clustering model is as follows: hierarchical clustering assumes a hierarchical structure between categories, clustering samples into hierarchical classes. Hierarchical clustering has two methods, aggregation and splitting. Clustering and clustering divide each sample into one class, calculate the distance between every two samples, combine the two classes with the shortest distance into a new class, at the moment, calculate the distance between the new class and other classes, select the two classes with the shortest distance from all the distances to combine, and repeat the operation until the stop condition is met. Aggregated clustering may also be referred to as bottom-up clustering. Splitting the cluster divides all classes into one class and then divides the two most distant classes into two new classes, and repeating the operation until a stop condition is satisfied. Split clustering is also known as top-down clustering.
Considering that the fluctuation samples are label-free samples, a clustering algorithm in machine learning is selected to model the samples, and the purpose of distinguishing different types of liquid level fluctuation is achieved.
Since the labels of the normal samples are known, the recall ratio and precision ratio of the normal samples can be used to evaluate the quality of the clustering model.
Precision ratio: in the predicted class, the correct proportion is predicted, and the denominator is the number of samples in the predicted class. And (3) recall ratio: in the true category, the correct proportion is predicted, and the denominator is the number of samples in the true category.
And performing clustering analysis based on the main components extracted by the 5 kernel functions respectively, wherein the algorithm of the clustering analysis in the text is a Gaussian mixture model, a variational Bayesian Gaussian mixture model, K-Means clustering and hierarchical clustering. Since the specific classification condition of the data set is not known in advance, and the specific classification condition is not known to be suitable for several classes, the classification into 6-15 classes, which is selected to obtain 10 conditions in total, is shown in the following tables 3, 4, 5 and 6 based on the recall ratio and precision ratio of the normal samples of different clustering models of the training set under the condition of obtaining different kernel function dimension reduction:
TABLE 3 recall and precision for Gaussian mixture models under different clustering conditions
TABLE 4 recall ratio and precision ratio of variational Bayes Gaussian mixture model under different clustering conditions
TABLE 5K-Means model recall and precision for different clustering scenarios
TABLE 6 recall and precision for hierarchical clustering under different clustering conditions
In the selection of the dimension reduction method, the principal component dimension reduction effect based on the RBF kernel function and the Sigmoid kernel function is obviously superior to that of other kernel functions. If the model with the recall ratio and precision ratio of more than 60 percent on the normal sample is defined as a qualified model, the hierarchical clustering performance is worst, and no qualified model exists; the Gaussian mixture model performs better than a variational Bayesian Gaussian mixture model and K-Means clustering.
In a Gaussian mixture model, when principal component dimensionality reduction is carried out based on a Sigmoid kernel function, and training samples are divided into 11 types, the best recall ratio 67.024% and precision ratio 99.365% of comprehensive evaluation can be achieved; when the principal component dimensionality reduction is carried out based on the RBF kernel function, the training samples are divided into 9 classes, and the best recall ratio 67.88% and precision ratio 92.42% of comprehensive evaluation can be achieved.
By calculating the recall ratio and precision ratio of a normal sample, the best comprehensive evaluation is found when a Gaussian mixture model obtained by principal component dimensionality reduction based on a Sigmoid kernel function is also used for classifying test samples into 11 classes, wherein the recall ratio is 41.509%, and the precision ratio is 98.507%; when the principal component dimensionality reduction is carried out based on the RBF kernel function, the training samples are also divided into 9 classes, the comprehensive evaluation is best, the recall ratio is 57.233%, and the precision ratio is 94.792%. As can be seen from the change of the recall ratio from the training set to the test set, when the principal component dimensionality reduction is carried out based on the RBF kernel function, the generalization capability of the model is better than that of a Gaussian mixture model for the principal component dimensionality reduction based on the Sigmoid kernel function
5) Classifying liquid level fluctuation clustering conditions according to a model constructed by an optimal dimensionality reduction method, and setting corresponding solutions under various conditions;
by inquiring an actual data curve, the liquid level fluctuation clustering condition of principal component dimension reduction based on the RBF kernel function in a test set can be roughly divided into the following 8 types: within 5 minutes of pouring, liquid level fluctuation, low-frequency liquid level fluctuation, high-frequency liquid level fluctuation, liquid level fluctuation caused by a water changing port, divergent liquid level fluctuation, sudden rising type liquid level fluctuation, liquid level fluctuation caused by flocculation flow and periodic liquid level fluctuation. After a specific clustering result is obtained, specific influence factors of each liquid level fluctuation category can be positioned according to actual conditions and the experience of high workers, and available treatment measures can be obtained. If a fluctuating sample is classified as a flocculated flow induced level fluctuation, the level fluctuation can be mitigated by adding jitter.
6) And inputting the detected data into the constructed model in real time for monitoring, outputting a corresponding solution when the data is abnormal, and providing the solution for an operator or a control center for controlling.
In the step 1), setting the liquid level, the section serial number and the steel grade number as system set values; setting a pulling speed, a stopper rod setting position, a bulging compensation amount, a casting starting time and an NP value of an electromagnetic instrument as theoretical values calculated by a system; the actual liquid level, the actual pulling speed, the actual position of the stopper rod, the weight of the tundish, the actual N value of the liquid level (electromagnetic) electromagnetic instrument and the sensor temperature are obtained values of the corresponding sensor or equipment.
With the continuous maturity of new generation information technologies represented by big data, artificial intelligence and the like, more and more difficult-to-solve industrial problems find new solutions and methods. The method carries out cluster analysis on the problem of crystallizer liquid level fluctuation by using a machine learning method. The obtained Gaussian mixture model for principal component dimension reduction based on the Sigmoid kernel function has better precision ratio and recall ratio when sample data is divided into 11 classes; the Gaussian mixture model for performing principal component dimensionality reduction based on the RBF kernel function has better precision ratio and recall ratio when sample data is divided into 9 classes; and the generalization capability of the Gaussian mixture model for carrying out principal component dimension reduction based on the RBF kernel function is better than that of the Gaussian mixture model for carrying out principal component dimension reduction based on the Sigmoid kernel function. Through comparing historical curves, the clustering result can reflect the actual situation to a certain extent, a new scheme is provided for positioning the specific influence factors of liquid level fluctuation, the problem of inaccurate control caused by controlling the liquid level of the crystallizer by only detecting the liquid level height of molten steel is avoided, and the accuracy of liquid level control of the crystallizer is further improved.
Based on the same concept, the present embodiment provides a crystallizer liquid level control device, as shown in fig. 2, the crystallizer liquid level control device includes:
the acquisition module 1 is used for acquiring influence data of the liquid level of the crystallizer, wherein the influence data are determined according to influence factors of the liquid level of the crystallizer;
the input module 2 is used for inputting the influence data into a crystallizer liquid level fluctuation factor determination model and determining the fluctuation state of the liquid level of the crystallizer;
and the control module 3 is used for adjusting the influence factors of the crystallizer according to the fluctuation state so as to control the liquid level of the crystallizer.
It should be understood that, each module of the crystallizer liquid level control device provided in this embodiment can be combined to implement each step of the crystallizer liquid level control method, so as to achieve the same technical effect as each step of the crystallizer liquid level control method, and thus, no further description is provided herein.
As shown in fig. 3, an electronic device according to an embodiment of the present application includes a processor 111, a communication interface 112, a memory 113, and a communication bus 114, where the processor 111, the communication interface 112, and the memory 113 complete mutual communication via the communication bus 114,
a memory 113 for storing a computer program;
in an embodiment of the present application, the processor 111, when executing the program stored in the memory 113, is configured to implement the steps of the crystallizer liquid level control method provided in any one of the foregoing method embodiments.
The present application further provides a computer readable storage medium, on which a computer program is stored, which when executed by a processor, implements the steps of the crystallizer liquid level control method provided in any of the foregoing method embodiments.
It is noted that, in this document, relational terms such as "first" and "second," and the like, may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising a … …" does not exclude the presence of another identical element in a process, method, article, or apparatus that comprises the element.
The above description is merely illustrative of particular embodiments of the invention that enable those skilled in the art to understand or practice the invention. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the invention. Thus, the present invention is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A crystallizer liquid level control method is characterized by comprising the following steps:
acquiring influence data of the liquid level of the crystallizer, wherein the influence data is determined according to influence factors of the liquid level of the crystallizer;
inputting the influence data into a crystallizer liquid level fluctuation factor determination model, and determining the fluctuation state of the liquid level of the crystallizer;
and adjusting influence factors of the crystallizer according to the fluctuation state so as to control the liquid level of the crystallizer.
2. The crystallizer liquid level control method of claim 1, wherein before inputting the influence data into a crystallizer liquid level fluctuation factor determination model, the method further comprises:
acquiring historical influence data, and determining a normal sample and a fluctuation sample according to the historical influence data;
determining a training set and a testing set based on the normal sample and the fluctuation sample;
and modeling the training set and the testing set through a clustering algorithm to obtain the crystallizer liquid level fluctuation factor determination model.
3. Crystallizer liquid level control method as claimed in claim 2, characterized in that said historical influence data comprise: actual liquid level and set liquid level; wherein, according to historical influence data, determining a normal sample and a fluctuation sample comprises:
determining the difference value between the actual liquid level and the set liquid level at each moment in the historical influence data;
when the difference value between the actual liquid level and the set liquid level at any moment in the historical influence data exceeds a first threshold value, taking the data at any moment as the fluctuation sample;
and when the difference value between the actual liquid level and the set liquid level at any moment in the historical influence data is lower than a second threshold value, taking the data at any moment as the normal sample, wherein the second threshold value is lower than the first threshold value.
4. Crystallizer liquid level control method according to claim 2, characterized in that determining a training set and a test set based on said normal samples and said undulation samples comprises:
mixing the normal sample and the fluctuation sample to obtain a total training set;
and dividing the total training set based on a preset proportion to obtain the training set and the test set.
5. The crystallizer liquid level control method of claim 4, wherein modeling the training set and the test set by a clustering algorithm to obtain the crystallizer liquid level fluctuation factor determination model comprises:
respectively extracting time domain characteristics and frequency domain characteristics of variable variables in the training set and the test set to obtain data dimensions of the training set and data dimensions of the test set;
performing dimensionality reduction processing on the data dimensionality of the training set and the data dimensionality of the testing set to obtain training set dimensionality reduction data and testing set dimensionality reduction data;
and modeling the dimension reduction data of the training set and the dimension reduction data of the testing set through a clustering algorithm to obtain a crystallizer liquid level fluctuation factor determination model.
6. The crystallizer liquid level control method of claim 5, wherein modeling the training set dimension reduction data and the testing set dimension reduction data through a clustering algorithm to obtain the crystallizer liquid level fluctuation factor determination model comprises:
modeling the training set dimensionality reduction data and the testing set dimensionality reduction data through at least two clustering algorithms to obtain at least two initial crystallizer liquid level fluctuation factor determination models;
evaluating at least two initial crystallizer liquid level fluctuation factor determination models based on the normal sample to obtain an optimal initial crystallizer liquid level fluctuation factor determination model;
and taking the optimal initial crystallizer liquid level fluctuation factor determination model as the crystallizer liquid level fluctuation factor determination model.
7. The crystallizer liquid level control method of claim 5, wherein after the optimal initial crystallizer liquid level fluctuation factor determination model is used as the crystallizer liquid level fluctuation factor determination model, the method further comprises:
determining clustering conditions of classifying the training set dimension reduction data and the test set dimension reduction data by a model according to the crystallizer liquid level fluctuation factors to obtain multiple preset fluctuation states;
and setting an influence factor adjusting method corresponding to the preset fluctuation state.
8. The crystallizer liquid level control device is characterized by comprising:
the acquisition module is used for acquiring influence data of the liquid level of the crystallizer, and the influence data is determined according to influence factors of the liquid level of the crystallizer;
the input module is used for inputting the influence data into a crystallizer liquid level fluctuation factor determination model and determining the fluctuation state of the liquid level of the crystallizer;
and the control module is used for adjusting the influence factors of the crystallizer according to the fluctuation state so as to control the liquid level of the crystallizer.
9. An electronic device is characterized by comprising a processor, a communication interface, a memory and a communication bus, wherein the processor and the communication interface are used for realizing mutual communication by the memory through the communication bus;
a memory for storing a computer program;
a processor for implementing the steps of the crystallizer liquid level control method of any one of claims 1-7 when executing the program stored in the memory.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the crystallizer liquid level control method according to any of claims 1-7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210951137.XA CN115430815A (en) | 2022-08-09 | 2022-08-09 | Crystallizer liquid level control method and device, electronic equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210951137.XA CN115430815A (en) | 2022-08-09 | 2022-08-09 | Crystallizer liquid level control method and device, electronic equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN115430815A true CN115430815A (en) | 2022-12-06 |
Family
ID=84242180
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210951137.XA Pending CN115430815A (en) | 2022-08-09 | 2022-08-09 | Crystallizer liquid level control method and device, electronic equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115430815A (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104226953A (en) * | 2013-06-24 | 2014-12-24 | 上海梅山钢铁股份有限公司 | Method for compensating molten steel liquid level of crystallizer on basis of throwing force |
| CN108494594A (en) * | 2018-03-21 | 2018-09-04 | 中国人民解放军陆军炮兵防空兵学院郑州校区 | A kind of analysis method and system of EIGRP route networks failure |
| CN109676106A (en) * | 2019-02-13 | 2019-04-26 | 衡阳镭目科技有限责任公司 | A kind of method and device controlling continuous cast mold liquid fluctuating |
| CN110703006A (en) * | 2019-09-04 | 2020-01-17 | 国网浙江省电力有限公司金华供电公司 | Three-phase power quality disturbance detection method based on convolutional neural network |
| CN111090050A (en) * | 2020-01-21 | 2020-05-01 | 合肥工业大学 | Lithium battery fault diagnosis method based on support vector machine and K mean value |
| US20210048402A1 (en) * | 2018-12-11 | 2021-02-18 | Dalian University Of Technology | Prediction method for mold breakout based on feature vectors and hierarchical clustering |
| CN112935213A (en) * | 2019-12-11 | 2021-06-11 | 中冶京诚工程技术有限公司 | Crystallizer, molten steel liquid level height measuring method in crystallizer and related device |
| CN113469253A (en) * | 2021-07-02 | 2021-10-01 | 河海大学 | Electricity stealing detection method based on triple twin network |
| CN114091553A (en) * | 2020-08-06 | 2022-02-25 | 长沙理工大学 | Diagnosis method for rolling bearing fault |
-
2022
- 2022-08-09 CN CN202210951137.XA patent/CN115430815A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104226953A (en) * | 2013-06-24 | 2014-12-24 | 上海梅山钢铁股份有限公司 | Method for compensating molten steel liquid level of crystallizer on basis of throwing force |
| CN108494594A (en) * | 2018-03-21 | 2018-09-04 | 中国人民解放军陆军炮兵防空兵学院郑州校区 | A kind of analysis method and system of EIGRP route networks failure |
| US20210048402A1 (en) * | 2018-12-11 | 2021-02-18 | Dalian University Of Technology | Prediction method for mold breakout based on feature vectors and hierarchical clustering |
| CN109676106A (en) * | 2019-02-13 | 2019-04-26 | 衡阳镭目科技有限责任公司 | A kind of method and device controlling continuous cast mold liquid fluctuating |
| CN110703006A (en) * | 2019-09-04 | 2020-01-17 | 国网浙江省电力有限公司金华供电公司 | Three-phase power quality disturbance detection method based on convolutional neural network |
| CN112935213A (en) * | 2019-12-11 | 2021-06-11 | 中冶京诚工程技术有限公司 | Crystallizer, molten steel liquid level height measuring method in crystallizer and related device |
| CN111090050A (en) * | 2020-01-21 | 2020-05-01 | 合肥工业大学 | Lithium battery fault diagnosis method based on support vector machine and K mean value |
| CN114091553A (en) * | 2020-08-06 | 2022-02-25 | 长沙理工大学 | Diagnosis method for rolling bearing fault |
| CN113469253A (en) * | 2021-07-02 | 2021-10-01 | 河海大学 | Electricity stealing detection method based on triple twin network |
Non-Patent Citations (2)
| Title |
|---|
| 徐芊;汪健冬;王梦琳;杨秋莲;刘广臣;孙涛;王柳青;陈文;尹立军;: "基于XGBoost两层算法模型的风机齿轮箱输入轴故障监测与诊断方法", 电力设备管理, no. 03 * |
| 郑德玲, 赵星浩, 曾永生: "连铸机结晶器液面状态识别及遗传模糊控制方法", 北京科技大学学报, no. 06, pages 596 - 598 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112527788B (en) | Method and device for detecting and cleaning abnormal value of transformer monitoring data | |
| CN108038044B (en) | Anomaly detection method for continuous monitored object | |
| US9996444B2 (en) | Automated methods and systems for calculating hard thresholds | |
| Khademi et al. | Fuzzy rules for fuzzy $\overline {X} $ and $ R $ control charts | |
| CN112905412A (en) | Method and device for detecting abnormity of key performance index data | |
| CN111898443A (en) | Flow monitoring method for wire feeding mechanism of FDM type 3D printer | |
| CN111046532A (en) | Coherent cluster recognition method based on elbow criterion | |
| CN102087787A (en) | Prediction device and prediction method for short time traffic conditions | |
| CN116413604A (en) | Battery parameter monitoring method, system, device and storage medium | |
| CN111222968A (en) | Enterprise tax risk management and control method and system | |
| CN115430815A (en) | Crystallizer liquid level control method and device, electronic equipment and storage medium | |
| CN107212882A (en) | The real-time detection method and system of a kind of EEG signals state change | |
| CN114155717A (en) | Traffic flow data screening method, device, equipment and readable storage medium | |
| CN111624978B (en) | Method and device for evaluating performance of water supply flow controller | |
| CN111062133A (en) | Wind turbine generator performance analysis method and system | |
| CN114819628B (en) | Visual engineering supervision system | |
| CN111224830A (en) | Data monitoring method and device, Internet of things network element and computer readable storage medium | |
| CN102142192A (en) | Method and system for forewarning pedestrian traffic congestion based on combined multi-step prediction | |
| CN110020000B (en) | Method and device for judging abnormal wind speed data | |
| CN110517731A (en) | Genetic test quality monitoring data processing method and system | |
| CN109840309A (en) | A kind of calculation method of iron melt desulfurizing agent dosage | |
| CN105095649B (en) | The data analysing method and system of building inclination angle | |
| CN102831105B (en) | A kind of method of EXCEL and MINITAB15 software programming *-R control chart coefficient table | |
| CN112149833B (en) | Prediction method, device, equipment and storage medium based on machine learning | |
| CN102156822A (en) | Pedestrian traffic data assembly multi-step forecasting method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |