An intelligent control method for key variables in aluminum electrolysis production process
By constructing an intelligent control method based on data-driven and experiential knowledge, the instability of electrolytic cells caused by reliance on human experience in the aluminum electrolysis industry has been solved. This has improved the accuracy of alumina concentration and aluminum fluoride addition, thereby enhancing the stability of the electrolytic cells and product quality.
Patent Information
- Application Number
- CN202210829186.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- Filing Date
- 2022-06-27
- Publication Date
- 2024-11-22
- Estimated Expiration
- 2042-06-27
AI Technical Summary
In the aluminum electrolysis industry, existing control methods rely on the experience of operators, resulting in poor stability of electrolytic cells, unstable product quality and production capacity, and difficulty in achieving automatic adjustment.
By employing data mining, pattern recognition, and machine learning techniques, an intelligent control method for alumina concentration, aluminum fluoride addition, and anode rise and fall is constructed. This method combines fuzzy decision trees, fuzzy cognitive graphs, and unknown dynamic estimators to achieve automatic adjustment.
This improved the accuracy of alumina concentration and aluminum fluoride addition, enhanced the stability of the electrolytic cell and product quality, and reduced the uncertainty of human experience.
Smart Images

Figure CN115202301B_ABST
Abstract
Description
Technical Field
[0001] The invention belongs to the technical field of aluminum electrolysis and has the characteristic of designing an intelligent control method for key variables in an aluminum electrolysis process. Background Art
[0002] Aluminum is produced industrially by electrolysis. The Hall-Heroult cryolite-alumina molten salt electrolysis method is used in aluminum electrolysis industry, that is, fluoride salts mainly composed of cryolite are used as flux, and alumina is used as melt to form a multiphase electrolyte system. Among them, the Na2AlF6-Al2O3 binary system and the Na3AlF6-AlF3-Al2O3 ternary system are the basis of industrial electrolytes.
[0003] The electrolytic aluminum industry has a great impact on the environment and is a high-energy-consuming and high-polluting industry. The waste gas discharged from the electrolytic aluminum production is mainly CO2, and gas-solid fluoride mainly composed of HF gas. CO2 is a greenhouse gas and the main cause of global warming. The greenhouse effect of CF4 and C2F6 in fluoride is 6500-10000 times that of carbon dioxide, and will have different degrees of impact on the ozone layer. HF is a highly toxic gas that enters the human body through the skin or respiratory tract, and only 1.5g can be lethal.
[0004] At present, the control of material feeding and pole distance based on experience knowledge has been realized in the production management process of the aluminum electrolysis industry. However, it is difficult to implement automatic adjustment according to the cell conditions due to the problem of measuring important state parameters of the aluminum electrolysis cell. Most of the existing controls still rely on the experience of operators for adjustment, which is highly subjective, resulting in poor stability of the electrolysis cell, inconsistent product quality and production capacity. Summary of the invention
[0005] In order to solve the above technical problems, the present invention provides an intelligent control method for key variables in the aluminum electrolysis process, which utilizes new artificial intelligence methods such as data mining, pattern recognition, and machine learning, and integrates data-driven and empirical knowledge to construct intelligent control of three key process variables: alumina concentration, aluminum fluoride addition, and anode lifting and lowering.
[0006] In order to achieve the technical effect of solving the above technical problems, the present invention is implemented by the following technical solutions: an intelligent control method for key variables in aluminum electrolysis process, characterized by including three variable controls: aluminum oxide concentration, aluminum fluoride addition and anode lifting and lowering;
[0007] A method for controlling the variable concentration of aluminum oxide comprises the following steps:
[0008] S1. First, extract knowledge from production data to form a knowledge representation framework;
[0009] S2. Based on the idea in S1 above, a fuzzy decision tree is proposed to extract relationship rules from the historical production data of aluminum electrolysis;
[0010] S3, using fuzzy cognitive graph as a model for knowledge representation and reasoning, and obtaining the fuzzy cognitive graph model through optimization methods;
[0011] S4, inputting the production process information into the above-mentioned fuzzy cognitive graph model, and after iterative calculation and denormalization, obtaining the amount of aluminum chloride added;
[0012] S5, using the amount of aluminum chloride added in S4 in the production process or process equipment to obtain feedback information, and updating the relationship rules in S2 based on the information;
[0013] S6. Combined with the specific measures for the standard feeding of the electrolytic cell, a feeding plan for the electrolytic cell with different alumina concentration intervals is given to control the alumina concentration to reach the desired interval value;
[0014] The variable control method of aluminum fluoride addition includes the following steps:
[0015] S1. Select model feature variables based on expert experience and knowledge, and perform fuzzy division;
[0016] S2, then obtain production rules based on the experience knowledge of aluminum electrolysis process personnel and express them as IT2FPNs;
[0017] S3, integrating expert cognition into the production rules in S2, introducing type II fuzzy membership function, using the entropy weight approximation ideal solution sorting method to fuse multiple expert cognitions for each fuzzy partition, and constructing a knowledge decision model based on IT2FPNs and approximation ideal solution sorting method;
[0018] S4, input the production information into the knowledge decision model of S3, and obtain the amount of aluminum fluoride added based on IT2FPNs;
[0019] Anode lifting variable control method, including building an unknown dynamic estimator, controlling the anode lifting system by means of PID feedback plus feedforward compensation;
[0020] The beneficial effects of the present invention are:
[0021] 1. The present invention provides a variable control method for aluminum oxide concentration, which constructs a decision model based on production data and cognitive knowledge, and uses actual production information as the basis to bring the model into the calculation of the amount of aluminum chloride added. Compared with the traditional method of adding aluminum chloride based on human experience, the amount of aluminum chloride added is more accurate, which is beneficial to improving the stability of the electrolytic cell and the product quality.
[0022] 2. The variable control method for adding aluminum fluoride provided by the present invention is also based on expert experience and knowledge to construct a decision-making model, thereby improving the accuracy of the amount of aluminum fluoride added and eliminating the uncertainty of manual addition; BRIEF DESCRIPTION OF THE DRAWINGS
[0023] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the accompanying drawings required for describing the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other accompanying drawings can be obtained based on these accompanying drawings without paying creative work.
[0024] Figure 1 It is a guideline for constructing a data-driven intelligent decision-making model in the variable control method of aluminum oxide concentration of the present invention;
[0025] Figure 2 It is a guideline for constructing an intelligent decision-making model based on empirical knowledge in the aluminum fluoride addition variable control method of the present invention;
[0026] Figure 3 It is a design idea map of the anode lifting device controller in the anode lifting variable control method of the present invention. DETAILED DESCRIPTION
[0027] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.
[0028] Example 1
[0029] Intelligent control of alumina concentration based on data drive
[0030] First, knowledge is extracted from production data to form a knowledge representation framework, and then a decision model is obtained. Based on the above idea, a fuzzy decision tree is proposed to extract relationship rules from the historical production data of aluminum electrolysis. Fuzzy cognitive graphs are used as models for knowledge representation and reasoning, and fuzzy cognitive graph models are obtained through optimization methods to form a knowledge model with structured knowledge representation models and simple and efficient reasoning calculations, such as Figure 1 When there is production process information input, the designed inference algorithm obtains the output result and acts on the production process or process equipment, obtains feedback information, updates the rules, and then combines the specific measures of the electrolytic cell benchmark unloading to give the unloading plan of the alumina concentration in the electrolytic cell at different intervals, so as to control the alumina concentration to reach the desired interval value.
[0031] Example 2
[0032] Aluminum fluoride addition control based on experience knowledge decision-making
[0033] The addition of aluminum fluoride is an important operation in the aluminum electrolysis process, and the decision on the amount of addition currently relies more on manual experience decision-making methods. However, experts in aluminum electrolysis processes have different experiences and cognitions. This paper studies the knowledge representation and reasoning problems based on interval type-2 fuzzy Petri nets (IT2FPNs) to achieve the fusion of different expert cognitions and use a unified model to make decisions. On this basis, it studies how to use parallel knowledge reasoning to obtain the amount of aluminum fluoride addition. First, the model feature variables are selected according to the expert experience knowledge, and fuzzy division is performed; then, the production rules are obtained according to the experience knowledge of the aluminum electrolysis process personnel and expressed as IT2FPNs. In order to integrate the experience of multiple experts, the type II fuzzy membership function is introduced, and the entropy weight-based approximation to ideal solution sorting method is used to integrate the cognition of multiple experts in each fuzzy division. Thus, a knowledge decision-making model based on IT2FPNs and the approximation to ideal solution sorting method is constructed, such as Figure 2 shown.
[0034] Example 3
[0035] Anode lifting control with unknown and uncertain dynamic compensation
[0036] The pole pitch is a critical control variable in the aluminum electrolysis production process. However, the influence of mechanical structures such as worm gear reducers and ball screws in the electrolytic cell anode lifting device leads to many non-smooth dynamics (such as friction, hysteresis, dead zone or saturation, etc.), which brings many difficulties to the precise control of the pole pitch. Therefore, a simple unknown dynamic estimator will be studied to realize the control of the anode lifting system under external disturbances by means of PID feedback plus feedforward compensation. The control strategy will not rely on traditional disturbance observation and other measurements, which can reduce the demand for system models and only use integrated estimation and control driven by measurable input / output data (such as Figure 3 ).
[0037] In summary, 1. The present invention provides a variable control method for aluminum oxide concentration, which constructs a decision model based on production data and cognitive knowledge, and uses actual production information as a basis to bring the model into the calculation of the amount of aluminum chloride added. Compared with the traditional method of adding aluminum chloride based on human experience, the amount of aluminum chloride added is more accurate, which is beneficial to improving the stability of the electrolytic cell and the product quality.
[0038] 2. The variable control method for aluminum fluoride addition provided by the present invention is also based on expert experience and knowledge to construct a decision-making model, thereby improving the accuracy of aluminum fluoride addition and eliminating the uncertainty of manual addition.
[0039] In the description of this specification, the description with reference to the terms "one embodiment", "example", "specific example", etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described can be combined in any one or more embodiments or examples in a suitable manner.
[0040] The preferred embodiments of the present invention disclosed above are only used to help illustrate the present invention. The preferred embodiments do not describe all the details in detail, nor do they limit the invention to the specific implementation methods described. Obviously, many modifications and changes can be made according to the content of this specification. This specification selects and specifically describes these embodiments in order to better explain the principles and practical applications of the present invention, so that those skilled in the art can understand and use the present invention well. The present invention is limited only by the claims and their full scope and equivalents.
Claims
1. An intelligent control method for key variables in aluminum electrolysis process, characterized in that: Includes three variable controls: alumina concentration, aluminum fluoride addition, and anode lifting and lowering; A method for controlling the variable concentration of aluminum oxide comprises the following steps: S1. First, extract knowledge from production data to form a knowledge representation framework; S2. Based on the knowledge representation framework in S1 above, a fuzzy decision tree is proposed to extract relationship rules from the historical production data of aluminum electrolysis; S3, using fuzzy cognitive graph as a model for knowledge representation and reasoning, and obtaining the fuzzy cognitive graph model through optimization methods; S4, inputting the production process information into the above-mentioned fuzzy cognitive graph model, and after iterative calculation and denormalization, obtaining the amount of aluminum chloride added; S5, using the amount of aluminum chloride added in S4 in the production process or process equipment to obtain feedback information, and updating the relationship rules in S2 based on the information; S6. Combined with the specific measures for the standard feeding of the electrolytic cell, a feeding plan for the electrolytic cell with different alumina concentration intervals is given to control the alumina concentration to reach the desired interval value; The variable control method of aluminum fluoride addition comprises the following steps: S1. Select model feature variables based on expert experience and knowledge, and perform fuzzy division; S2, then obtain production rules based on the experience knowledge of aluminum electrolysis process personnel and express them as IT2FPNs; S3, integrating expert cognition into the production rules in S2, introducing type II fuzzy membership function, using the entropy weight approximation ideal solution sorting method to fuse multiple expert cognitions for each fuzzy partition, and constructing a knowledge decision model based on IT2FPNs and approximation ideal solution sorting method; S4, input the production information into the knowledge decision model of S3, and obtain the amount of aluminum fluoride added based on IT2FPNs; The anode lifting variable control method includes constructing an unknown dynamic estimator and controlling the anode lifting system by means of PID feedback plus feedforward compensation.
