Fuzzy reasoning-based aluminum electrolysis alumina concentration online identification method and system

Abstract

The invention provides an aluminum electrolysis aluminum oxide concentration online identification method and system based on fuzzy reasoning, and the method comprises the steps: obtaining the apparent aluminum oxide concentration and a comprehensive blanking period through a time sequence knowledge graph, and calculating the PDHC value of the current comprehensive blanking period and the PDHC value of the previous comprehensive blanking period; inputting the PDHC value of the current comprehensive blanking period and the PDHC value of the previous comprehensive blanking period into a preset fuzzy inference engine to obtain an aluminum oxide concentration state of the current comprehensive blanking period and a fuzzy inference output variable of an aluminum oxide concentration change trend; and carrying out defuzzification processing on the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive blanking period to obtain an identification evaluation value of the alumina concentration state and the alumina concentration change trend of the current comprehensive blanking period.

Description

Technical Field

[0001] The present disclosure relates to the field of industrial semantics, and more specifically, to an online identification method and system for aluminum electrolysis alumina concentration based on fuzzy reasoning. Background Art

[0002] Alumina concentration is a core parameter in the aluminum electrolysis production process, which directly affects the material balance of the electrolytic cell. Alumina concentration monitoring is of great significance in reducing production energy consumption and improving product quality. However, with the large-scale development of aluminum electrolytic cells, aluminum electrolysis data presents the characteristics of massive, scattered, and coexisting in multiple formats, which brings challenges to the online identification of alumina concentration. Therefore, how to process the production knowledge in the aluminum electrolysis process and realize accurate online identification of alumina concentration is the key to improving the quality and efficiency of the aluminum electrolysis industry. When the alumina concentration is in a normal state, there is a "under-rise and over-fall" relationship between its normalized cell voltage and the feeding state, that is, when the feeding state is under-feeding, the cell voltage is in a rising state, and when the feeding state is over-feeding, the cell voltage is in a falling state. Its PDHC

[34] The value changes within the normal range; when the concentration is in an abnormal state allowed by the process, the relationship of "under-rise and over-fall" is destroyed, and the value of PDHC also changes accordingly. Therefore, the online identification of alumina concentration can be achieved based on the value of PDHC. In the multi-granularity time-series knowledge graph of aluminum electrolysis, the value of PDHC is stored at the process semantic level.

[0003] Traditional methods for identifying alumina concentration are mainly divided into mechanism model-based methods and purely data-driven methods. The mechanism model-based method extracts key parameters from process mechanism knowledge or experimental electrolytic cells to build a model to estimate and predict alumina concentration. In the model-based method, key parameters are mostly fixed values, which is inconsistent with the actual situation that key parameters in aluminum electrolysis production change in real time and cannot be measured quickly and continuously. Therefore, the molecular dynamics and fluid dynamics models established from a microscopic perspective are too computationally expensive and difficult to apply to industrial production. The second type of method is data-driven and attempts to obtain knowledge from parameters such as cell voltage to estimate and predict alumina concentration. However, the above methods lack process mechanism knowledge, are not sensitive enough to abnormal cell conditions of the electrolytic cell, and are difficult to adapt to frequent changes in operating conditions in actual production.

[0004] In summary, many key parameters in the existing alumina concentration identification methods, such as aluminum output, aluminum level, electrolyte level, etc., cannot be measured continuously and quickly, and are mostly stored in process reports after manual offline measurements. The production operation data such as cell voltage and cell current collected by industrial sensors are time series data and stored in a time series database. Therefore, the identification of alumina concentration involves a large amount of multi-source heterogeneous data, which is difficult to achieve online identification. In addition, the existing methods can only recognize the normal state of alumina concentration, and it is difficult to identify the abnormally low and abnormally high states of alumina concentration online. Too low alumina concentration can easily cause anode effect and destroy the material balance of the electrolytic cell. Therefore, the abnormally low state of alumina concentration can be identified by predicting the anode effect. Too high alumina concentration will reduce current efficiency and endanger the stability of the electrolytic cell, but there are currently very few online detection methods for high alumina concentration in electrolytic cells, which is difficult to meet the needs of comprehensive identification of alumina concentration states. Summary of the invention

[0005] The purpose of the embodiments of the present disclosure is to provide a method and system for online identification of aluminum oxide concentration in aluminum electrolysis based on fuzzy reasoning, so as to improve the accuracy of online detection of high aluminum oxide concentration in the electrolytic cell.

[0006] In a first aspect, the present invention provides an online identification method for aluminum electrolysis aluminum oxide concentration based on fuzzy reasoning, comprising: an online identification method for aluminum electrolysis aluminum oxide concentration based on fuzzy reasoning, characterized in that it comprises:

[0007] Obtain the apparent alumina concentration and comprehensive feeding cycle through the time series knowledge graph, and calculate the PDHC value of the current comprehensive feeding cycle and the previous comprehensive feeding cycle;

[0008] Inputting the PDHC values ​​of the current comprehensive feeding cycle and the previous comprehensive feeding cycle into a preset fuzzy inference engine to obtain fuzzy inference output variables of the aluminum oxide concentration state and the aluminum oxide concentration change trend of the current comprehensive feeding cycle;

[0009] The fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive material unloading cycle are defuzzified to obtain the identification evaluation values ​​of the alumina concentration state and the alumina concentration change trend of the current comprehensive material unloading cycle.

[0010] Furthermore, the preset fuzzy inference engine establishes fuzzy rules through the following steps:

[0011] According to the historical data of various electrolytic cells, the numerical range of the PDHC value under various states of alumina concentration is obtained, and the first threshold, the second threshold, the third threshold and the fourth threshold are selected, and a fuzzy threshold range is preset for each type of electrolytic cell; wherein the second threshold < the first threshold < the third threshold < the fourth threshold; the fuzzy threshold range includes a normal threshold range of alumina concentration for characterizing that the PDHC value is greater than the first threshold and less than the third threshold, a negative small threshold range for characterizing that the PDHC value is less than the first threshold and greater than the second threshold, a negative large threshold range for characterizing that the PDHC value is less than the second threshold, a positive small threshold range for characterizing that the PDHC value is greater than the third threshold and less than the fourth threshold, and a positive large threshold range for characterizing that the PDHC value is greater than the fourth threshold;

[0012] Establishing PDHC membership functions corresponding to the normal threshold range, the negative small threshold range, the negative large threshold range, the positive small threshold range, and the positive large threshold range;

[0013] According to the PDHC value of the current comprehensive unloading cycle, a fuzzy inference rule of the alumina concentration is established, and the fuzzy inference rule of the alumina concentration change trend includes a normal state of alumina concentration for characterizing that the PDHC value is within the normal threshold range, a high state of alumina concentration for characterizing that the PDHC value is within the negative small threshold range, an extremely high state of alumina concentration for characterizing that the PDHC value is within the negative large threshold range, a low state of alumina concentration for characterizing that the PDHC value is within the positive small threshold range, and an extremely low state of alumina concentration for characterizing that the PDHC value is within the positive large threshold range;

[0014] Establishing aluminum oxide concentration membership functions corresponding to the normal aluminum oxide concentration state, the low aluminum oxide concentration state, the high aluminum oxide concentration state, the extremely low aluminum oxide concentration state, and the extremely high aluminum oxide concentration state;

[0015] According to the PDHC values ​​of the two comprehensive feeding cycles, a fuzzy inference rule for the change trend of the alumina concentration is established, wherein the fuzzy inference rule for the change trend of the alumina concentration includes a first alumina concentration change trend for representing that the alumina concentration remains unchanged, a second alumina concentration change trend for representing that the alumina concentration increases, a third alumina concentration change trend for representing that the alumina concentration decreases, a fourth alumina concentration change trend for representing that the alumina concentration changes from low to normal, and a fifth alumina concentration change trend for representing that the alumina concentration changes from high to normal;

[0016] Establish an aluminum oxide concentration variation trend membership function corresponding to the first aluminum oxide concentration variation trend, the second aluminum oxide concentration variation trend, the third aluminum oxide concentration variation trend, the fourth aluminum oxide concentration variation trend, and the fifth aluminum oxide concentration variation trend.

[0017] Furthermore, the step of inputting the PDHC values ​​of the current integrated material discharging cycle and the previous integrated material discharging cycle into a preset fuzzy inference engine to obtain the fuzzy inference output variables of the alumina concentration state and the alumina concentration change trend of the current integrated material discharging cycle includes:

[0018] According to each PDHC membership function, the PDHC fuzziness value corresponding to the PDHC value of the current comprehensive material unloading cycle and the previous comprehensive material unloading cycle is obtained;

[0019] According to the PDHC fuzziness value corresponding to the PDHC value of the current comprehensive unloading cycle and the previous comprehensive unloading cycle, the fuzzy reasoning rule of the alumina concentration and the fuzzy reasoning rule of the alumina concentration change trend, the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle are obtained.

[0020] Furthermore, the step of performing defuzzification processing on the fuzzy inference output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive feeding cycle to obtain the identification evaluation value of the alumina concentration state and the alumina concentration change trend of the current comprehensive feeding cycle also includes:

[0021] According to the alumina concentration membership function and the alumina concentration change trend membership function, the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle are defuzzified using the centroid method to obtain the identification evaluation values ​​of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle.

[0022] Further, the PDHC membership functions of the normal threshold range, the negative small threshold range, and the positive small threshold range are generalized bell-shaped membership functions; the PDHC membership function of the negative large threshold range is a Z-shaped membership function, and the PDHC membership function of the positive large threshold range is an S-shaped membership function; the aluminum oxide concentration membership function and the aluminum oxide concentration change trend membership function are triangular membership functions;

[0023] After the step of obtaining the identification evaluation value of the aluminum oxide concentration state and the aluminum oxide concentration change trend of the current comprehensive material feeding cycle, the method further includes:

[0024] The identification and evaluation values ​​of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle are stored in the time series knowledge graph, and a corresponding relationship is established between the electrolytic cell and the alumina concentration according to the identification and evaluation values ​​of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle to reflect the online working conditions in the aluminum electrolysis production process.

[0025] In a second aspect, the present invention provides an online identification system for aluminum electrolysis alumina concentration based on fuzzy reasoning, comprising:

[0026] The PDHC acquisition module is used to obtain the apparent alumina concentration and the comprehensive feeding cycle through the time series knowledge graph, and calculate the PDHC value of the current comprehensive feeding cycle and the previous comprehensive feeding cycle;

[0027] A fuzzy reasoning module, used for inputting the PDHC values ​​of the current comprehensive feeding cycle and the previous comprehensive feeding cycle into a preset fuzzy reasoning engine, and obtaining fuzzy reasoning output variables of the aluminum oxide concentration state and the aluminum oxide concentration change trend of the current comprehensive feeding cycle;

[0028] The online identification module is used to defuzzify the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive feeding cycle to obtain the identification evaluation value of the alumina concentration state and the alumina concentration change trend of the current comprehensive feeding cycle.

[0029] Furthermore, the preset fuzzy inference engine includes:

[0030] A PDHC membership function modeling module is used to obtain the numerical range of PDHC values ​​under various states of alumina concentration according to historical data of various electrolytic cells, select a first threshold, a second threshold, a third threshold and a fourth threshold, and preset a fuzzy threshold range for each type of electrolytic cell; wherein the second threshold < the first threshold < the third threshold < the fourth threshold; the fuzzy threshold range includes a normal threshold range of alumina concentration for characterizing that the PDHC value is greater than the first threshold and less than the third threshold, a negative small threshold range for characterizing that the PDHC value is less than the first threshold and greater than the second threshold, a negative large threshold range for characterizing that the PDHC value is less than the second threshold, a positive small threshold range for characterizing that the PDHC value is greater than the third threshold and less than the fourth threshold, and a positive large threshold range for characterizing that the PDHC value is greater than the fourth threshold; establish each PDHC membership function corresponding to the normal threshold range, negative small threshold range, negative large threshold range, positive small threshold range and positive large threshold range;

[0031] Alumina concentration membership function modeling module, used for establishing fuzzy reasoning rules of the alumina concentration according to the PDHC value of the current comprehensive unloading cycle, the fuzzy reasoning rules of the alumina concentration change trend include a normal state of alumina concentration for characterizing that the PDHC value is within the normal threshold range, a high state of alumina concentration for characterizing that the PDHC value is within the negative small threshold range, an extremely high state of alumina concentration for characterizing that the PDHC value is within the negative large threshold range, a low state of alumina concentration for characterizing that the PDHC value is within the positive small threshold range, and an extremely low state of alumina concentration for characterizing that the PDHC value is within the positive large threshold range; establishing alumina concentration membership functions corresponding to the normal state of alumina concentration, the low state of alumina concentration, the high state of alumina concentration, the extremely low state of alumina concentration, and the extremely high state of alumina concentration;

[0032] The alumina concentration change trend membership function modeling module is used to establish fuzzy reasoning rules for the alumina concentration change trend according to the PDHC values ​​of two comprehensive feeding cycles, wherein the fuzzy reasoning rules for the alumina concentration change trend include a first alumina concentration change trend for representing a constant alumina concentration, a second alumina concentration change trend for representing an increase in alumina concentration, a third alumina concentration change trend for representing a decrease in alumina concentration, a fourth alumina concentration change trend for representing a change from low to normal alumina concentration, and a fifth alumina concentration change trend for representing a change from high to normal alumina concentration; and establish alumina concentration change trend membership functions corresponding to the first alumina concentration change trend, the second alumina concentration change trend, the third alumina concentration change trend, the fourth alumina concentration change trend, and the fifth alumina concentration change trend.

[0033] Furthermore, the fuzzy reasoning module is specifically used to obtain, according to each PDHC membership function, the PDHC fuzziness value corresponding to the PDHC value of the current comprehensive unloading cycle and the previous comprehensive unloading cycle; and according to the PDHC fuzziness value corresponding to the PDHC value of the current comprehensive unloading cycle and the previous comprehensive unloading cycle, the fuzzy reasoning rule of the alumina concentration and the fuzzy reasoning rule of the alumina concentration change trend, to obtain the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle.

[0034] Furthermore, the online identification module is specifically used to defuzzify the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle according to the alumina concentration membership function and the alumina concentration change trend membership function using the centroid method to obtain the identification evaluation value of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle.

[0035] Further, the PDHC membership functions of the normal threshold range, the negative small threshold range, and the positive small threshold range are generalized bell-shaped membership functions; the PDHC membership function of the negative large threshold range is a Z-shaped membership function, and the PDHC membership function of the positive large threshold range is an S-shaped membership function;

[0036] The aluminum oxide concentration membership function and the aluminum oxide concentration change trend membership function are triangular membership functions.

[0037] The fuzzy reasoning-based online identification method and system for aluminum electrolysis alumina concentration of the present invention utilizes an online identification method for alumina concentration based on a multi-granularity time series knowledge graph, obtains the value of PDHC through the time series knowledge graph to perform online identification of alumina concentration, stores the identification results in the time series knowledge graph, and establishes a corresponding relationship between the electrolytic cell and the alumina concentration based on the results to reflect the online working conditions in the aluminum electrolysis production process. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0039] Figure 1 It is a flow chart of an online identification method of aluminum oxide concentration in aluminum electrolysis based on fuzzy reasoning according to an embodiment of the present disclosure.

[0040] Figure 2 It is a graph of the PDHC membership function in the online identification method of aluminum electrolysis alumina concentration based on fuzzy reasoning according to an embodiment of the present disclosure.

[0041] Figure 3 It is a surface schematic diagram of the variation trend of PDHC and alumina concentration in the online identification method of alumina concentration in aluminum electrolysis based on fuzzy reasoning according to an embodiment of the present disclosure.

[0042] Figure 4 It is a diagram showing the results of identifying the state of alumina concentration in the online identification method of alumina concentration in aluminum electrolysis based on fuzzy reasoning according to an embodiment of the present disclosure.

[0043] Figure 5 It is a structural schematic diagram of an online identification system for aluminum electrolysis alumina concentration based on fuzzy reasoning according to an embodiment of the present disclosure. DETAILED DESCRIPTION

[0044] The embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0045] It should be noted that the following embodiments and features in the embodiments may be combined with each other in the absence of conflict; and, based on the embodiments in the present disclosure, all other embodiments obtained by ordinary technicians in the field without making any creative work are within the scope of protection of the present disclosure.

[0046] It should be noted that various aspects of the embodiments within the scope of the appended claims are described below. It should be apparent that the aspects described in the present embodiment may be embodied in a wide variety of forms, and any specific structure and / or function described in the present embodiment is merely illustrative. Based on the present disclosure, it should be understood by those skilled in the art that an aspect described in the present embodiment may be implemented independently of any other aspect, and two or more of these aspects may be combined in various ways. For example, any number of aspects described in the present embodiment may be used to implement the device and / or practice the method. In addition, other structures and / or functionalities other than one or more of the aspects described in the present embodiment may be used to implement this device and / or practice this method.

[0047] Figure 1 It is a flow chart of an online identification method of aluminum oxide concentration in aluminum electrolysis based on fuzzy reasoning according to an embodiment of the present disclosure.

[0048] like Figure 1 As shown, the online identification method of aluminum electrolysis alumina concentration based on fuzzy reasoning includes:

[0049] Step 101: Obtain the apparent alumina concentration and the comprehensive feeding cycle through the time series knowledge graph, and calculate the PDHC value of the current comprehensive feeding cycle and the previous comprehensive feeding cycle;

[0050] Step 102: inputting the PDHC values ​​of the current comprehensive material discharging cycle and the previous comprehensive material discharging cycle into a preset fuzzy inference engine to obtain fuzzy inference output variables of the aluminum oxide concentration state and aluminum oxide concentration change trend of the current comprehensive material discharging cycle;

[0051] Step 103: Defuzzify the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive material unloading cycle to obtain the identification evaluation value of the alumina concentration state and the alumina concentration change trend of the current comprehensive material unloading cycle.

[0052] In this embodiment, the apparent alumina concentration PAC and the comprehensive feeding cycle IFP are first obtained through the time series knowledge graph, the PDHC value of the comprehensive feeding cycle is calculated, the PDHC value is input into the set fuzzy inference engine, the output variable of the fuzzy inference is obtained, and the identification evaluation value of the alumina concentration state and change trend is obtained after defuzzification, and finally the alumina concentration identification result represented by the evaluation value is entered into the time series knowledge graph.

[0053] There are also multiple preferred embodiments of the above-mentioned online identification method of aluminum oxide concentration in aluminum electrolysis based on fuzzy reasoning. The preset fuzzy reasoning engine establishes fuzzy rules through the following steps: according to the historical data of various types of electrolytic cells, the numerical range of the PDHC value under various states of aluminum oxide concentration is obtained to select the first threshold, the second threshold, the third threshold and the fourth threshold, and the fuzzy threshold range is preset for each type of electrolytic cell; wherein, the second threshold < the first threshold < the third threshold < the fourth threshold; the fuzzy threshold range includes a normal threshold range of aluminum oxide concentration for characterizing that the PDHC value is greater than the first threshold and less than the third threshold, a negative small threshold range for characterizing that the PDHC value is less than the first threshold and greater than the second threshold, a negative large threshold range for characterizing that the PDHC value is less than the second threshold, a positive small threshold range for characterizing that the PDHC value is greater than the third threshold and less than the fourth threshold, and a positive large threshold range for characterizing that the PDHC value is greater than the fourth threshold;

[0054] Establishing PDHC membership functions corresponding to the normal threshold range, the negative small threshold range, the negative large threshold range, the positive small threshold range, and the positive large threshold range;

[0055] According to the PDHC value of the current comprehensive unloading cycle, a fuzzy inference rule of the alumina concentration is established, and the fuzzy inference rule of the alumina concentration change trend includes a normal state of alumina concentration for characterizing that the PDHC value is within the normal threshold range, a high state of alumina concentration for characterizing that the PDHC value is within the negative small threshold range, an extremely high state of alumina concentration for characterizing that the PDHC value is within the negative large threshold range, a low state of alumina concentration for characterizing that the PDHC value is within the positive small threshold range, and an extremely low state of alumina concentration for characterizing that the PDHC value is within the positive large threshold range;

[0056] Establishing aluminum oxide concentration membership functions corresponding to the normal aluminum oxide concentration state, the low aluminum oxide concentration state, the high aluminum oxide concentration state, the extremely low aluminum oxide concentration state, and the extremely high aluminum oxide concentration state;

[0057] According to the PDHC values ​​of the two comprehensive feeding cycles, a fuzzy inference rule for the change trend of the alumina concentration is established, wherein the fuzzy inference rule for the change trend of the alumina concentration includes a first alumina concentration change trend for representing that the alumina concentration remains unchanged, a second alumina concentration change trend for representing that the alumina concentration increases, a third alumina concentration change trend for representing that the alumina concentration decreases, a fourth alumina concentration change trend for representing that the alumina concentration changes from low to normal, and a fifth alumina concentration change trend for representing that the alumina concentration changes from high to normal;

[0058] Establish an aluminum oxide concentration variation trend membership function corresponding to the first aluminum oxide concentration variation trend, the second aluminum oxide concentration variation trend, the third aluminum oxide concentration variation trend, the fourth aluminum oxide concentration variation trend, and the fifth aluminum oxide concentration variation trend.

[0059] The above preferred embodiment specifically includes a plurality of specific optimization steps, firstly, the PDHC fuzzy membership function and its parameters are determined, and the PDHC is converted into an identification evaluation index by fuzzifying the PDHC and adopting reasonable fuzzy rules. The accuracy of the identification process is further improved by fuzzifying the input, determining the fuzzy membership function and determining the relevant parameters.

[0060] (I) Input parameter determination

[0061] In order to better identify the concentration of alumina in the aluminum electrolysis process, it is necessary not only to consider the state of alumina concentration in the current comprehensive feeding cycle, but also to consider the change trend of alumina concentration. If only the current PDHC value is used as the input of fuzzy reasoning, it is easy to misjudge the current state of the electrolytic cell, thereby providing an incorrect feeding decision. Therefore, this embodiment uses the PDHC value of the current comprehensive feeding cycle and the PDHC value of the previous comprehensive feeding cycle as the input of fuzzy reasoning. Since the time range covered by a comprehensive feeding cycle is about 2000 seconds in the industrial data used, the PDHC of two adjacent comprehensive feeding cycles can basically reflect the change trend of alumina concentration. The key parameter setting of fuzzy reasoning includes the determination of PDHC threshold, the determination of input parameters, and the determination of fuzzy membership function and its parameters. For the PDHC threshold, the identification of alumina concentration is mainly achieved by comparing whether the PDHC value exceeds the preset threshold in positive and negative conditions. Due to the different models and process parameters of the electrolytic cells, the PDHC value range and threshold under each cell condition change with the change of electrolyte components, there is uncertainty, and it is difficult to give an accurate threshold. This embodiment obtains the approximate range of PDHC under normal alumina concentration conditions from historical data of various types of electrolytic cells, and presets a fuzzy threshold for each type of electrolytic cell, that is, the set threshold is a range rather than an exact value. On this basis, the PDHC value is converted into an identification evaluation index, thereby realizing online identification of alumina concentration.

[0062] (II) Fuzzy processing of input parameters

[0063] After the fuzzy inference input parameters are set, these input parameters need to be fuzzified. In this stage, the first thing to do is to choose a suitable membership function.

[0064] Common fuzzy membership functions are mainly divided into triangular membership functions, trapezoidal membership functions, Gaussian membership functions, generalized bell-shaped membership functions, S-shaped membership functions and Z-shaped membership functions. This embodiment selects three membership functions with relatively smooth curves: generalized bell-shaped membership functions, S-shaped membership functions and Z-shaped membership functions. Common fuzzification methods generally divide input parameters into central area fuzzy sets and edge area fuzzy sets, wherein the central area fuzzy set represents that the variable is in the optimal state or stable trend, and the edge area fuzzy set represents that the state of the variable is not ideal or has an increasing or decreasing trend, and the closer the fuzzy set is to the edge, the more extreme its state and trend are. According to the characteristics of the aluminum electrolysis production process and the possible state of alumina concentration, the fuzzy set division of the input parameter PDHC value is carried out, and the PDHC is divided into five fuzzy sets: normal threshold range (N), negative small threshold range (NS), negative large threshold range (NB), positive small threshold range (PS) and positive large threshold range (PB). Among them, N, NS, and PS are located in the central area, indicating that the PDHC value is in the normal, small, and large ranges respectively; NB and PB are located at the left and right ends respectively, indicating that the PDHC value is in the extremely small or extremely large area.

[0065] Among them, since the three fuzzy sets of N, NS, and PS are located at the center of all fuzzy sets, and the PDHC ranges covered by these three categories (normal, slightly smaller, and slightly larger) account for a relatively small part of the overall PDHC range, these fuzzy sets are determined using a generalized bell-shaped membership function that can be precisely adjusted. The form of the bell-shaped membership function is shown below.

[0066]

[0067] Among them, c i and a i are the center value positions and curve width parameters of the three fuzzy function curves, b i is the curve slope parameter.

[0068] Since the NB fuzzy set indicates that the PDHC value is in an extremely small area and is located at the leftmost end of all fuzzy sets, a monotonically decreasing Z-type membership function is selected to determine the fuzzy set, which represents the concept of "negative large". The form of the Z-type membership function is shown below.

[0069]

[0070] Among them, a i and b i The initial point where the function curve drops and the point where the function intersects the horizontal line are determined respectively, which together determine the downward slope of the curve.

[0071] Since the PB fuzzy set indicates that the PDHC value is in an extremely large area and is located at the rightmost end of all fuzzy sets, a monotonically increasing S-type membership function is selected to determine the fuzzy set to represent the concept of "positive and large". The form of the S-type membership function is shown below.

[0072]

[0073] Where c i and d i They respectively determine the steepness of the curve and the initial value of the horizontal translation.

[0074] (III) Determination of membership function parameters

[0075] In the process of fuzzifying the input parameters, the parameters of the membership function can directly affect the size of each fuzzy set area and the shape of the function curve, and thus affect the results of fuzzy reasoning. Therefore, when setting the membership function parameters, it is necessary to determine the approximate range of normal PDHC values ​​based on the distribution of PDHC values ​​of different electrolytic cells. On this basis, combined with the process mechanism knowledge, the size of the fuzzy set and the shape of the membership function curve are as consistent as possible with the actual situation in industrial production.

[0076] Assume that the normal PDHC value range is The critical point of abnormally low PDHC value is The critical point of abnormally high PDHC value is Firstly, according to the process mechanism knowledge, the membership function parameters of the three types of fuzzy sets located in the central area, namely normal, negative small and positive small, are set.

[0077] First, determine the center value position c of the membership function curve i From the process mechanism knowledge, we know that when the PDHC value is It indicates that the current aluminum oxide concentration is basically normal. At this time, the center point parameter c of the membership function curve corresponding to the N fuzzy set is N Can be and The calculation results are:

[0078]

[0079] Among them, ω N is the adjustment coefficient, which is used to correct the center point position.

[0080] When the PDHC value is lower than and higher than When , it indicates that the current aluminum oxide concentration is slightly higher than the normal value, but not abnormally high. The PDHC value at this time is assumed to belong to the NS fuzzy set. Its membership function f NS The center point C of (x) NS Can be based on and The calculation results are:

[0081]

[0082] Among them, ω Ns f NS The adjustment coefficient of the center point of (x).

[0083] Similarly, when the PDHC value is higher than and lower than When , it indicates that the current aluminum oxide concentration is slightly lower than the normal value, but not abnormally low. The PDHC value at this time is set as the PS fuzzy set. Its membership function f PS The center point C of (x) PS Can be based on and The calculation results are:

[0084]

[0085] Among them, ω PS f PS (x) Adjustment coefficient of the center point.

[0086] When the PDHC value is far beyond the normal range and break through the critical point or , indicating that the current aluminum oxide concentration is abnormally high or low. In this case, the PDHC value belongs to the NB and PB fuzzy sets respectively. Therefore, according to the critical point or The initial point a of the descending of the membership function curves γ(x) and δ(x) corresponding to NB and PB PDHC and initial translation value d PDHC To confirm:

[0087]

[0088] Among them, ω NB and ω PB γ(x) and δ(x) are adjustment coefficients used to adjust the parameter position.

[0089] After the above parameters are determined, the width of the bell-shaped membership function and the slopes of the S-shaped and Z-shaped membership functions can be determined based on these parameters.

[0090] For the membership function corresponding to the fuzzy set of category N, its width can be expressed by the center position c of the function. N And two adjacent membership function curves f NS (x) and f PS The center point C of (x) NS and C PS From the historical data, it can be seen that the normal value range of PDHC accounts for a small proportion of the overall range. Therefore, the width of the membership function corresponding to the N fuzzy set is also small. This embodiment uses the minimum value function to calculate f N The width of (x) is calculated as follows:

[0091]

[0092] in, For the curve f N Width correction factor for (x).

[0093] For the fuzzy sets corresponding to NS and PS, the maximum value function is used to calculate the width of its membership function curve. The calculation method is as follows:

[0094]

[0095] in, and They are curves f NS (x) and f PS Width correction factor for (x).

[0096] The horizontal line intersection point b of the membership function corresponding to the fuzzy sets corresponding to NB and PB PDHC and the steepness parameter c PDHC The horizontal line intersection point b of γ(x) is determined by the known parameters of the membership function and the width parameters of the adjacent function. PDHC for:

[0097]

[0098] in, is the intersection position adjustment coefficient.

[0099] Similarly, the steepness parameter c of δ(x) PDHC for:

[0100]

[0101] in, Coefficient that adjusts the steepness parameter for δ(x).

[0102] After parameter setting, the PDHC membership function parameter summary table is shown in Table 1, and the PDHC membership function curve is shown in Figure 2 shown.

[0103] Table 1

[0104]

[0105] The identification of alumina concentration should not only consider the state of alumina concentration in the current comprehensive feeding cycle, but also the change trend of alumina concentration. A good change trend can not only indicate that the current electrolytic cell is in a stable state, but under certain conditions, a good trend can improve the current state of alumina concentration and make the alumina concentration develop from high or low to normal. Therefore, it is necessary to set the fuzzy reasoning rules for alumina concentration identification based on the PDHC values ​​of two adjacent comprehensive feeding cycles.

[0106] According to the characteristics of the aluminum electrolysis process, the current state of alumina concentration is divided into five categories: normal alumina concentration state (Z, normal alumina concentration state), high alumina concentration state (H, i.e. high alumina concentration state), abnormally high alumina concentration state (AH, i.e. extremely high alumina concentration state), low alumina concentration state (L, i.e. low alumina concentration state), abnormally low alumina concentration state (AL, i.e. extremely low alumina concentration state).

[0107] The alumina concentration change trend can be divided into five categories: the alumina concentration state remains basically unchanged (S, i.e. the first alumina concentration change trend), the alumina concentration changes from high to normal (HN, i.e. the fifth alumina concentration change trend), the alumina concentration changes from low to normal (LN, i.e. the fourth alumina concentration change trend), the alumina concentration decreases (TL, i.e. the third alumina concentration change trend), and the alumina concentration increases (TH, i.e. the second alumina concentration change trend).

[0108] There is a corresponding relationship between the PDHC value in the current comprehensive feeding cycle and the alumina concentration state. According to this corresponding relationship, the fuzzy reasoning rules are as follows:

[0109] IF PDHC1=N THEN P=Z

[0110] IF PDHC1=NS THEN P=H

[0111] IF PDHC1=PS THEN P=L

[0112] IF PDHC1=NB THEN P=AH IF PDHC1=PB THEN P=AL

[0113] Among them, PDHC1 is the value of PDHC in the current comprehensive feeding cycle, and P is the alumina concentration state in the current comprehensive feeding cycle.

[0114] The alumina concentration change trend T is inferred by the PDHC value PDHC2 of the previous comprehensive feeding cycle and the PDHC value PDHC1 of the current comprehensive feeding cycle. If the PDHC2 value of the previous feeding cycle and the PDHC1 value of the current comprehensive feeding cycle are in the same fuzzy set, that is, the alumina concentration state of the previous comprehensive feeding cycle is basically the same as the alumina concentration state of the current comprehensive feeding cycle, indicating that the alumina concentration state is basically unchanged. The reasoning rule for this situation is:

[0115] IF(PDHC1=N and PDHC2=N) or (PDHC1=NS and PDHC2=NS) or (PDHC1=PS and PDHC2=PS) or (PDHC1=NB and PDHC2=NB) or (PDHC1=PB and PDHC2=PB), THEN T=S

[0116] When the PDHC value returns to the normal range from a smaller value (NS) or an abnormally small value (NB) outside the normal range, it indicates that the aluminum oxide concentration has changed from high to normal. The reasoning rule for this situation is:

[0117] IF(PDHC2=NS and PDHC1=N)or(PDHC2=NB and PDHC1=N),THEN T=HN

[0118] Among them, the weights of PDHC2=NB and PDHC1=N are relatively small. This is because in the actual aluminum electrolysis production process, the alumina concentration is difficult to adjust from an abnormally high state to a normal state within a comprehensive feeding cycle.

[0119] Similarly, when the PDHC value returns to the normal range from a larger value (PS) or an abnormally large value (PB) outside the normal range, it indicates that the aluminum oxide concentration has returned to normal from low. The reasoning rule for this situation is:

[0120] IF(PDHC2=PS and PDHC1=N)or(PDHC2=PB and PDHC1=N),THEN T=LN

[0121] Similarly, in Among them, PDHC2=PB and PDHC1=N have smaller weights.

[0122] The PDHC value of the current comprehensive material cutting cycle does not belong to the same fuzzy set as the PDHC value of the current comprehensive material cutting cycle, and the fuzzy set of the PDHC value of the current comprehensive material cutting cycle is Figure 2 If the given membership function curve is biased to the right, it means that the aluminum oxide concentration is reduced. The inference rule for this situation is:

[0123] IF(PDHC2=NB and PDHC1=NS) or (PDHC2=NB and PDHC1=PS) or (PDHC2=Nand PDHC1=PS) or (PDHC2=N and PDHC1=PB) or (PDHC2=PS and PDHC1=PB) or (PDHC2=NS and PDHC1=PS) or (PDHC2=NS and PDHC1=PB),THENT=TL

[0124] Similarly, when the PDHC2 value and the PDHC1 value do not belong to the same fuzzy set, and the fuzzy set of PDHC1 is Figure 2 If the given membership function curve is biased to the left, it means that the aluminum oxide concentration increases. The inference rule for this situation is:

[0125] IF(PDHC2=PB and PDHC1=PS) or (PDHC2=PB and PDHC1=NS) or (PDHC2=Nand PDHC1=NS) or (PDHC2=N and PDHC1=NB) or (PDHC2=NS and PDHC1=NB) or (PDHC2=PS and PDHC1=NS) or (PDHC2=PS and PDHC1=NB), THEN T=TH

[0126] The above five fuzzy inference rules are summarized in a table, and the results are shown in Table 2. Among them, S means that the alumina concentration is basically unchanged, TH means that the alumina concentration increases, TL means that the alumina concentration decreases, HN and LN mean that the alumina concentration changes from low to normal and from high to normal, respectively.

[0127] Table 2 Alumina concentration identification rules based on PDHC

[0128]

[0129] In the actual aluminum electrolysis process, it is basically impossible that the alumina concentration in the previous comprehensive feeding cycle is abnormally high (abnormally low), while the alumina concentration in the next comprehensive feeding cycle is abnormally low (abnormally high). Therefore, when PDHC2=PB and PDHC1=NB or PDHC2=NB and PDHC1=PB appears, the fuzzy inference engine will not output the change trend of alumina concentration.

[0130] After obtaining the identification variables through the fuzzy reasoning method, it is necessary to convert the identification variables into real variables through the defuzzification method. Before defuzzification, it is necessary to set the value range of the identification variables, and set the 5 types of membership functions corresponding to the current aluminum oxide concentration state identification results and the aluminum oxide concentration change trend respectively on the value range. In this embodiment, the identification value range of the current aluminum oxide concentration state and the aluminum oxide concentration change trend is determined as [-3, 3]. For the current aluminum oxide concentration state (P), the identification value is close to 0, then the current aluminum oxide concentration is in a normal state. When the identification value is larger and closer to 3, it indicates that the aluminum oxide concentration is high; when the identification value is smaller and closer to -3, it indicates that the aluminum oxide concentration is low. For the aluminum oxide concentration change trend (Z), when the identification value is close to 0, it indicates that the aluminum oxide concentration is basically unchanged. When the identification value is larger and closer to 3, it indicates that the aluminum oxide concentration increases; when the identification value is smaller and closer to -3, it indicates that the aluminum oxide concentration decreases. This embodiment uses a triangular membership function to establish a membership function for the current state and change trend of aluminum oxide concentration. The form of the triangular membership function is:

[0131]

[0132] Among them, a i and c i is i When the horizontal axis is 0, b i is i The horizontal axis at the maximum value.

[0133] According to the characteristics of aluminum electrolysis production data, the parameters of each membership function are set, and the parameters are shown in Table 3.

[0134] Table 3 Membership function parameters of alumina concentration state and change trend

[0135]

[0136] After determining the membership function of the fuzzy system output value, the online identification semantic value of the aluminum oxide concentration is converted into an accurate value through a defuzzification method according to the fuzzy reasoning result. This embodiment uses the centroid method for defuzzification, and uses the centroid of the fuzzy set of the identification result as the defuzzified result. The calculation formula is:

[0137]

[0138] Among them, v0 is the exact value after defuzzification, v k is the identification variable, σ i Represents the membership function of the identification result.

[0139] After fuzzification, fuzzy rule reasoning and defuzzification, the relationship between the PDHC value and the current state of alumina concentration and the change trend of alumina concentration can be obtained. Since the state of alumina concentration is only related to the PDHC1 value of the current feeding state, this embodiment only gives a surface diagram of the PDHC1 value and the change trend of alumina concentration, such as Figure 3 As shown. For the experimental results, the evaluation indicators used in this embodiment mainly include: Accuracy, Precision, Recall and F1-Score. Among them, accuracy is the most common evaluation indicator, which calculates the ratio of the number of correctly classified samples to the total number of samples, but this indicator is less applicable to datasets with imbalanced categories. The experimental data used for aluminum oxide concentration identification is real-time industrial data obtained from industrial sensors, in which most of the aluminum oxide concentrations are in normal conditions. Therefore, the accuracy cannot comprehensively evaluate the classification performance and is only used as a reference in this embodiment.

[0140] Precision indicates the proportion of samples that are actually positive among all samples classified as positive. It focuses on the accuracy of the classifier for the positive class, that is, the reliability of the classifier when classifying as positive.

[0141] The recall rate indicates the proportion of samples that are correctly classified as positive among all samples that are actually positive. It focuses on the coverage of all true positive classes by the classifier, that is, the completeness of the classifier in classifying all samples that are actually positive.

[0142] The F1 score is the harmonic mean of precision and recall, which takes into account the balance between precision and recall and can better evaluate the overall performance of the classifier.

[0143] The calculation methods of the four types of evaluation indicators are as follows:

[0144]

[0145] Among them, T and F in TP, TN, FP and FN indicate whether the classification result matches the actual value, and P and N indicate whether the classification result is positive or negative. For example, in a multi-classification task, suppose there are three types A, B, and C. For A, its positive class is A, and its negative classes are B and C.

[0146] This embodiment adopts a fuzzy identification method of alumina concentration based on a multi-granularity time series knowledge graph, and uses PDHC data for online identification of alumina concentration. The PDHC data used in the experiment is derived from the multi-granularity knowledge graph, which contains all PDHC values ​​and corresponding comprehensive material discharge cycles of the No. 5201 electrolytic cell from August 2021 to June 2022, totaling 8935. The above data is used as the input of the fuzzy inference machine, and the fuzzy inference method is used to identify the state of alumina concentration. The final identification result is as follows: Figure 4 shown.

[0147] The results of fuzzy identification of aluminum oxide concentration are represented by confusion matrix. Among them, AL, L, Z, H, and AH are aluminum oxide concentration state labels, which respectively indicate abnormally low, low, normal, high, and abnormally high aluminum oxide concentrations. The horizontal axis of the confusion matrix represents the actual value label, the vertical axis represents the classification result label, the elements in each cell represent the classification proportion, and the elements on the main diagonal represent the proportion of each category classification correct to the actual sample, that is, the recall rate of the category. The evaluation indicators of each category are summarized in Table 4.

[0148] Table 4 Summary of evaluation index values ​​for each alumina concentration state

[0149]

[0150] According to the analysis of experimental results, the following conclusions can be drawn:

[0151] (1) Overall, the fuzzy identification method of aluminum oxide concentration can accurately identify the aluminum oxide concentration state. The classification accuracy of this method can be calculated by the total number of correctly classified samples and the total number of actual samples in Table 3. Its value is about 0.995, which indicates that this method has high accuracy in identifying the aluminum oxide concentration state.

[0152] (2) For samples with a "normal" aluminum oxide concentration, this method has excellent recognition effect. Since the proportion of samples with a "normal" aluminum oxide concentration is very large, it is necessary to comprehensively consider its precision and recall rate to ensure that the classification method can capture all truly "normal" samples as much as possible while avoiding misclassifying other types of samples as "normal". As shown in Table 3, the precision and recall rate of the "normal" aluminum oxide concentration state are both 0.999, which shows that this method has high classification reliability and classification completeness when facing samples with a "normal" aluminum oxide concentration, and can achieve comprehensive coverage of "normal" samples.

[0153] (3) For samples in other states, the recognition effect of this method is good. For "abnormal" samples with aluminum oxide concentration states of "abnormally low", "low", "high" and "abnormally high", the precision, recall and F1 scores of this method are all at a high level, indicating that this method has good recognition performance even for samples with "abnormal" aluminum oxide concentration states.

[0154] (4) This method will basically not mistakenly identify samples with high aluminum oxide concentration as low, and vice versa. In the aluminum electrolysis process, when the aluminum oxide concentration is low or high, different control strategies are required to adjust the aluminum oxide concentration to a normal state. If the high state is mistakenly identified as low, the wrong control strategy will be adopted, causing the aluminum oxide concentration to be out of control. Figure 4 From the confusion matrix shown, it can be seen that the proportion of samples with a "low" state that are mistakenly identified as "high" by this method is 0.001, and the proportion of samples with a "high" state that are mistakenly identified as "low" is 0.004. This shows that when faced with samples with low or high alumina concentration states, this method can identify them more accurately, thereby providing correct decision support for aluminum electrolysis production unloading.

[0155] Table 5 Comparison among various methods

[0156]

[0157] The parameters such as PDHC and IFP used in the method proposed in this embodiment can be obtained online. Through fuzzy reasoning, online identification of normal and abnormal states of aluminum oxide concentration can be achieved. The time required for identification of each comprehensive feeding cycle is approximately one feeding interval, which can basically meet the needs of online identification of aluminum oxide concentration. The method proposed in this embodiment is superior to the existing methods in various dimensions such as identification range, real-time performance, and accuracy. The method needs to be optimized in the future to reduce the time required for each identification and improve the time performance of the method.

[0158] Figure 5 It is a structural diagram of an online identification system for aluminum electrolysis alumina concentration based on fuzzy reasoning according to an embodiment of the present disclosure. Figure 1-4 The embodiment shown can be used to explain this embodiment. Figure 5 As shown: An online identification system for aluminum electrolysis alumina concentration based on fuzzy reasoning, comprising:

[0159] The PDHC acquisition module 501 is used to obtain the apparent alumina concentration and the comprehensive feeding cycle through the time series knowledge graph, and calculate the PDHC value of the current comprehensive feeding cycle and the previous comprehensive feeding cycle;

[0160] The fuzzy reasoning module 502 is used to input the PDHC values ​​of the current comprehensive material discharging cycle and the previous comprehensive material discharging cycle into a preset fuzzy reasoning engine to obtain the fuzzy reasoning output variables of the aluminum oxide concentration state and the aluminum oxide concentration change trend of the current comprehensive material discharging cycle;

[0161] The online identification module 503 is used to defuzzify the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive feeding cycle to obtain the identification evaluation value of the alumina concentration state and the alumina concentration change trend of the current comprehensive feeding cycle.

[0162] Preferably, the preset fuzzy inference engine includes:

[0163] A PDHC membership function modeling module is used to obtain the numerical range of PDHC values ​​under various states of alumina concentration according to historical data of various electrolytic cells, select a first threshold, a second threshold, a third threshold and a fourth threshold, and preset a fuzzy threshold range for each type of electrolytic cell; wherein the second threshold < the first threshold < the third threshold < the fourth threshold; the fuzzy threshold range includes a normal threshold range of alumina concentration for characterizing that the PDHC value is greater than the first threshold and less than the third threshold, a negative small threshold range for characterizing that the PDHC value is less than the first threshold and greater than the second threshold, a negative large threshold range for characterizing that the PDHC value is less than the second threshold, a positive small threshold range for characterizing that the PDHC value is greater than the third threshold and less than the fourth threshold, and a positive large threshold range for characterizing that the PDHC value is greater than the fourth threshold;

[0164] Alumina concentration membership function modeling module, used for establishing fuzzy reasoning rules of the alumina concentration according to the PDHC value of the current comprehensive unloading cycle, the fuzzy reasoning rules of the alumina concentration change trend include a normal state of alumina concentration for characterizing that the PDHC value is within the normal threshold range, a high state of alumina concentration for characterizing that the PDHC value is within the negative small threshold range, an extremely high state of alumina concentration for characterizing that the PDHC value is within the negative large threshold range, a low state of alumina concentration for characterizing that the PDHC value is within the positive small threshold range, and an extremely low state of alumina concentration for characterizing that the PDHC value is within the positive large threshold range; establishing alumina concentration membership functions corresponding to the normal state of alumina concentration, the low state of alumina concentration, the high state of alumina concentration, the extremely low state of alumina concentration, and the extremely high state of alumina concentration;

[0165] The alumina concentration change trend membership function modeling module is used to establish fuzzy reasoning rules for the alumina concentration change trend according to the PDHC values ​​of two comprehensive feeding cycles, wherein the fuzzy reasoning rules for the alumina concentration change trend include a first alumina concentration change trend for representing a constant alumina concentration, a second alumina concentration change trend for representing an increase in alumina concentration, a third alumina concentration change trend for representing a decrease in alumina concentration, a fourth alumina concentration change trend for representing a change from low to normal alumina concentration, and a fifth alumina concentration change trend for representing a change from high to normal alumina concentration; and establish alumina concentration change trend membership functions corresponding to the first alumina concentration change trend, the second alumina concentration change trend, the third alumina concentration change trend, the fourth alumina concentration change trend, and the fifth alumina concentration change trend.

[0166] Preferably, the fuzzy reasoning module 502 is specifically used to obtain, according to each PDHC membership function, the PDHC fuzziness value corresponding to the PDHC value of the current comprehensive unloading cycle and the previous comprehensive unloading cycle; and according to the PDHC fuzziness value corresponding to the PDHC value of the current comprehensive unloading cycle and the previous comprehensive unloading cycle, the fuzzy reasoning rule of the alumina concentration and the fuzzy reasoning rule of the alumina concentration change trend, to obtain the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle.

[0167] The online identification module 503 is specifically used to defuzzify the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle using the centroid method according to the alumina concentration membership function and the alumina concentration change trend membership function, so as to obtain the identification evaluation value of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle.

[0168] This embodiment uses an online identification method for alumina concentration based on a multi-granularity time-series knowledge graph. The PDHC value is obtained through the time-series knowledge graph to perform online identification of the alumina concentration, and the identification result is stored in the time-series knowledge graph. According to the result, a corresponding relationship is established between the electrolytic cell and the alumina concentration to reflect the online working conditions in the aluminum electrolysis production process.

[0169] It should be noted that not all steps and modules in the above-mentioned processes and system structure diagrams are necessary, and some steps or modules can be ignored according to actual needs. The execution order of each step is not fixed and can be adjusted as needed. The system structure described in the above-mentioned embodiments can be a physical structure or a logical structure, that is, some modules may be implemented by the same physical entity, or some modules may be implemented by multiple physical entities, or some components in multiple independent devices may be implemented together.

[0170] In the above embodiments, the hardware unit can be realized by mechanical means or electrical means. For example, a hardware unit can include permanent dedicated circuits or logic (such as special processors, FPGA or ASIC) to complete the corresponding operation. The hardware unit can also include programmable logic or circuits (such as general-purpose processors or other programmable processors), which can be temporarily set by software to complete the corresponding operation. Concrete implementation (mechanical means or dedicated permanent circuits or temporarily set circuits) can be determined based on cost and time considerations.

[0171] The present invention is shown and described in detail above through the accompanying drawings and preferred embodiments. However, the present invention is not limited to these disclosed embodiments. Based on the above multiple embodiments, those skilled in the art can know that the code review methods in the above different embodiments can be combined to obtain more embodiments of the present invention, and these embodiments are also within the protection scope of the present invention.

Claims

1. A method for online identification of aluminum oxide concentration in aluminum electrolysis based on fuzzy reasoning, characterized in that: include: Obtain the apparent alumina concentration and comprehensive feeding cycle through the time series knowledge graph, and calculate the PDHC value of the current comprehensive feeding cycle and the previous comprehensive feeding cycle; Inputting the PDHC values ​​of the current comprehensive feeding cycle and the previous comprehensive feeding cycle into a preset fuzzy inference engine to obtain fuzzy inference output variables of the aluminum oxide concentration state and the aluminum oxide concentration change trend of the current comprehensive feeding cycle; The fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive material unloading cycle are defuzzified to obtain the identification evaluation values ​​of the alumina concentration state and the alumina concentration change trend of the current comprehensive material unloading cycle.

2. The method for online identification of aluminum oxide concentration in aluminum electrolysis based on fuzzy reasoning according to claim 1 is characterized in that: The preset fuzzy inference engine establishes fuzzy rules through the following steps: According to the historical data of various electrolytic cells, the numerical range of the PDHC value under various states of alumina concentration is obtained, and the first threshold, the second threshold, the third threshold and the fourth threshold are selected, and a fuzzy threshold range is preset for each type of electrolytic cell; wherein the second threshold < the first threshold < the third threshold < the fourth threshold; the fuzzy threshold range includes a normal threshold range of alumina concentration for characterizing that the PDHC value is greater than the first threshold and less than the third threshold, a negative small threshold range for characterizing that the PDHC value is less than the first threshold and greater than the second threshold, a negative large threshold range for characterizing that the PDHC value is less than the second threshold, a positive small threshold range for characterizing that the PDHC value is greater than the third threshold and less than the fourth threshold, and a positive large threshold range for characterizing that the PDHC value is greater than the fourth threshold; Establishing PDHC membership functions corresponding to the normal threshold range, the negative small threshold range, the negative large threshold range, the positive small threshold range, and the positive large threshold range; According to the PDHC value of the current comprehensive unloading cycle, a fuzzy inference rule of the alumina concentration is established, and the fuzzy inference rule of the alumina concentration change trend includes a normal state of alumina concentration for characterizing that the PDHC value is within the normal threshold range, a high state of alumina concentration for characterizing that the PDHC value is within the negative small threshold range, an extremely high state of alumina concentration for characterizing that the PDHC value is within the negative large threshold range, a low state of alumina concentration for characterizing that the PDHC value is within the positive small threshold range, and an extremely low state of alumina concentration for characterizing that the PDHC value is within the positive large threshold range; Establishing aluminum oxide concentration membership functions corresponding to the normal aluminum oxide concentration state, the low aluminum oxide concentration state, the high aluminum oxide concentration state, the extremely low aluminum oxide concentration state, and the extremely high aluminum oxide concentration state; According to the PDHC values ​​of the two comprehensive feeding cycles, a fuzzy inference rule for the change trend of the alumina concentration is established, wherein the fuzzy inference rule for the change trend of the alumina concentration includes a first alumina concentration change trend for representing that the alumina concentration remains unchanged, a second alumina concentration change trend for representing that the alumina concentration increases, a third alumina concentration change trend for representing that the alumina concentration decreases, a fourth alumina concentration change trend for representing that the alumina concentration changes from low to normal, and a fifth alumina concentration change trend for representing that the alumina concentration changes from high to normal; Establish an aluminum oxide concentration variation trend membership function corresponding to the first aluminum oxide concentration variation trend, the second aluminum oxide concentration variation trend, the third aluminum oxide concentration variation trend, the fourth aluminum oxide concentration variation trend, and the fifth aluminum oxide concentration variation trend.

3. The method for online identification of aluminum oxide concentration in aluminum electrolysis based on fuzzy reasoning according to claim 2 is characterized in that: The step of inputting the PDHC values ​​of the current comprehensive feeding cycle and the previous comprehensive feeding cycle into a preset fuzzy inference engine to obtain the fuzzy inference output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive feeding cycle includes: According to each PDHC membership function, the PDHC fuzziness value corresponding to the PDHC value of the current comprehensive material unloading cycle and the previous comprehensive material unloading cycle is obtained; According to the PDHC fuzziness values ​​corresponding to the PDHC values ​​of the current comprehensive unloading cycle and the previous comprehensive unloading cycle, the fuzzy reasoning rules of the alumina concentration and the fuzzy reasoning rules of the alumina concentration change trend, the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle are obtained.

4. The method for online identification of aluminum oxide concentration in aluminum electrolysis based on fuzzy reasoning according to claim 3 is characterized in that: The step of performing defuzzification processing on the fuzzy inference output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive feeding cycle to obtain the identification evaluation value of the alumina concentration state and the alumina concentration change trend of the current comprehensive feeding cycle also includes: According to the alumina concentration membership function and the alumina concentration change trend membership function, the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle are defuzzified using the centroid method to obtain the identification evaluation values ​​of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle.

5. The method for online identification of aluminum oxide concentration in aluminum electrolysis based on fuzzy reasoning according to claim 4 is characterized in that: The PDHC membership functions of the normal threshold range, the negative small threshold range, and the positive small threshold range are generalized bell-shaped membership functions; the PDHC membership function of the negative large threshold range is a Z-shaped membership function, and the PDHC membership function of the positive large threshold range is an S-shaped membership function; the aluminum oxide concentration membership function and the aluminum oxide concentration change trend membership function are triangular membership functions; After the step of obtaining the identification evaluation value of the aluminum oxide concentration state and the aluminum oxide concentration change trend of the current comprehensive material feeding cycle, the method further includes: The identification and evaluation values ​​of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle are stored in the time series knowledge graph, and a corresponding relationship is established between the electrolytic cell and the alumina concentration according to the identification and evaluation values ​​of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle to reflect the online working conditions in the aluminum electrolysis production process.

6. An online identification system for aluminum electrolysis alumina concentration based on fuzzy reasoning, characterized in that: include: The PDHC acquisition module is used to obtain the apparent alumina concentration and the comprehensive feeding cycle through the time series knowledge graph, and calculate the PDHC value of the current comprehensive feeding cycle and the previous comprehensive feeding cycle; A fuzzy reasoning module, used for inputting the PDHC values ​​of the current comprehensive feeding cycle and the previous comprehensive feeding cycle into a preset fuzzy reasoning engine, and obtaining fuzzy reasoning output variables of the aluminum oxide concentration state and the aluminum oxide concentration change trend of the current comprehensive feeding cycle; The online identification module is used to defuzzify the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive feeding cycle to obtain the identification evaluation value of the alumina concentration state and the alumina concentration change trend of the current comprehensive feeding cycle.

7. The online identification system of aluminum electrolysis alumina concentration based on fuzzy reasoning according to claim 6 is characterized in that: The preset fuzzy inference engine includes: A PDHC membership function modeling module is used to obtain the numerical range of PDHC values ​​under various states of alumina concentration according to historical data of various electrolytic cells, select a first threshold, a second threshold, a third threshold and a fourth threshold, and preset a fuzzy threshold range for each type of electrolytic cell; wherein the second threshold < the first threshold < the third threshold < the fourth threshold; the fuzzy threshold range includes a normal threshold range of alumina concentration for characterizing that the PDHC value is greater than the first threshold and less than the third threshold, a negative small threshold range for characterizing that the PDHC value is less than the first threshold and greater than the second threshold, a negative large threshold range for characterizing that the PDHC value is less than the second threshold, a positive small threshold range for characterizing that the PDHC value is greater than the third threshold and less than the fourth threshold, and a positive large threshold range for characterizing that the PDHC value is greater than the fourth threshold; establish each PDHC membership function corresponding to the normal threshold range, negative small threshold range, negative large threshold range, positive small threshold range and positive large threshold range; an alumina concentration membership function modeling module, for establishing a fuzzy inference rule of the alumina concentration according to the PDHC value of the current comprehensive unloading cycle, wherein the fuzzy inference rule of the alumina concentration change trend includes a normal state of alumina concentration for characterizing that the PDHC value is within the normal threshold range, a high state of alumina concentration for characterizing that the PDHC value is within the negative small threshold range, an extremely high state of alumina concentration for characterizing that the PDHC value is within the negative large threshold range, a low state of alumina concentration for characterizing that the PDHC value is within the positive small threshold range, and an extremely low state of alumina concentration for characterizing that the PDHC value is within the positive large threshold range; The alumina concentration change trend membership function modeling module is used to establish fuzzy reasoning rules for the alumina concentration change trend according to the PDHC values ​​of two comprehensive feeding cycles, wherein the fuzzy reasoning rules for the alumina concentration change trend include a first alumina concentration change trend for representing a constant alumina concentration, a second alumina concentration change trend for representing an increase in alumina concentration, a third alumina concentration change trend for representing a decrease in alumina concentration, a fourth alumina concentration change trend for representing a change from low to normal alumina concentration, and a fifth alumina concentration change trend for representing a change from high to normal alumina concentration; and establish alumina concentration change trend membership functions corresponding to the first alumina concentration change trend, the second alumina concentration change trend, the third alumina concentration change trend, the fourth alumina concentration change trend, and the fifth alumina concentration change trend.

8. The online identification system of aluminum electrolysis alumina concentration based on fuzzy reasoning according to claim 7 is characterized in that: The fuzzy reasoning module is specifically used to obtain, according to each PDHC membership function, the PDHC fuzziness value corresponding to the PDHC value of the current comprehensive unloading cycle and the previous comprehensive unloading cycle; and according to the PDHC fuzziness value corresponding to the PDHC value of the current comprehensive unloading cycle and the previous comprehensive unloading cycle, the fuzzy reasoning rule of the alumina concentration and the fuzzy reasoning rule of the alumina concentration change trend, to obtain the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle.

9. The online identification system of aluminum electrolysis alumina concentration based on fuzzy reasoning according to claim 8 is characterized in that: The online identification module is specifically used to defuzzify the fuzzy reasoning output variables of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle according to the alumina concentration membership function and the alumina concentration change trend membership function using the center of gravity method to obtain the identification evaluation value of the alumina concentration state and the alumina concentration change trend of the current comprehensive unloading cycle.

10. The online identification system of aluminum electrolysis alumina concentration based on fuzzy reasoning according to claim 9 is characterized in that: The PDHC membership functions of the normal threshold range, the negative small threshold range, and the positive small threshold range are generalized bell-shaped membership functions; the PDHC membership function of the negative large threshold range is a Z-shaped membership function, and the PDHC membership function of the positive large threshold range is an S-shaped membership function; The aluminum oxide concentration membership function and the aluminum oxide concentration change trend membership function are triangular membership functions.

Images