Electrolytic bath state monitoring method and system based on generalized load equivalent model

Abstract

The invention discloses an electrolytic cell state monitoring method and system based on a generalized load equivalent model, and the method comprises the steps: operating a typical electrolytic cell generalized load equivalent model, and fitting a parameter curve equation according to the operation data, fitting dip angle values of parameter continuous values of the typical electrolytic cell generalized load equivalent model at all moments are obtained according to the parameter curve equation, adaptive discretization processing is conducted on sample values of the parameter continuous values of the model, and discretized parameter values are obtained; and analyzing and judging the current running state of the electrolytic cell by using a real-time generalized load equivalent model of the electrolytic cell based on the discretized parameter value. In the generalized load equivalent model parameter value of the electrolytic cell, the continuous parameter value obtained based on the fitting inclination angle is discretized and is applied to the generalized load equivalent model of the electrolytic cell, so that the model operation can be simplified, and the current density distribution and change trend in the electrolytic cell can be analyzed more easily; and the state of the electrolytic cell can be judged more accurately and quickly.

Description

Technical Field

[0001] The present invention relates to the technical field of electrolytic cell monitoring, and in particular to an electrolytic cell state monitoring method and system based on a generalized load equivalent model. Background Art

[0002] In the production process of electrolytic aluminum, the working state of the electrolytic cell directly affects the production efficiency and product quality. In order to monitor the working state of the electrolytic cell in real time, it is necessary to establish a generalized load equivalent model of the electrolytic cell. The model reflects the working state of the electrolytic cell through parameters such as the voltage and current of the electrolytic cell. However, the working environment of the electrolytic cell is complex and changeable. Various factors will affect the working state of the electrolytic cell, resulting in the instability of the parameters of the generalized load equivalent model, which affects the accuracy and reliability of the model. The traditional parameter optimization method is based on fitting the inclination angle to obtain continuous parameter values, but the parameter values ​​obtained by this method are not adaptable to the changes in the working conditions of the electrolytic cell and cannot adapt to the changing working environment.

[0003] It can be seen that how to effectively monitor the working status of the electrolytic cell has become a technical problem that needs to be urgently solved by those skilled in the art. Summary of the invention

[0004] The invention provides an electrolytic cell state monitoring method based on a generalized load equivalent model to improve the monitoring effect of the electrolytic cell state.

[0005] In order to solve the above technical problems, an embodiment of the present invention provides an electrolytic cell state monitoring method based on a generalized load equivalent model, comprising: Based on the historical operation data of the electrolytic cell to be monitored, a generalized load equivalent model of a typical electrolytic cell is constructed.

[0006] Run the generalized load equivalent model of the typical electrolytic cell, obtain sample values ​​of continuous parameter values ​​of the generalized load equivalent model of the typical electrolytic cell based on the time series, and fit the parameter curve equation according to the sample values; the continuous parameter values ​​include at least current density parameters, cell voltage parameters and electrolyte concentration parameters.

[0007] The fitting inclination values ​​of the continuous values ​​of the parameters of the generalized load equivalent model of the typical electrolytic cell at each moment are obtained according to the parameter curve equation.

[0008] The density peak clustering method is used to cluster the fitting inclination values, and the corresponding sample values ​​are divided into several parameter classes according to the clustering results.

[0009] The sample values ​​under each parameter class are adaptively discretized to obtain discretized parameter values; the adaptive discretization is designed to dynamically adjust the discretization interval according to the distribution range and change trend of the continuous values ​​of the parameters in the generalized load equivalent model of the typical electrolytic cell.

[0010] The real-time operation data of the electrolytic cell to be monitored is obtained, and a real-time electrolytic cell generalized load equivalent model is constructed according to the real-time operation data.

[0011] The discretized parameter value is applied to the real-time electrolytic cell generalized load equivalent model, and the real-time electrolytic cell generalized load equivalent model is run to obtain the parameter fitting inclination angle corresponding to the current electrolytic cell state, and the current electrolytic cell state is calculated and judged according to the parameter class corresponding to the parameter fitting inclination angle.

[0012] Furthermore, the generalized load equivalent model of a typical electrolytic cell is constructed based on the historical operation data of the electrolytic cell to be monitored, including: The historical operation data of the electrolytic cell to be monitored under normal operation and abnormal operation are obtained to obtain normal operation data and abnormal operation data.

[0013] The abnormal operation data and the normal operation data are compared and analyzed to obtain an abnormal characteristic value of the abnormal operation data relative to the normal operation data.

[0014] A parameter continuous value is determined according to the abnormal characteristic value, and a generalized load equivalent model of a typical electrolytic cell is constructed according to the parameter continuous value.

[0015] Furthermore, the running of the generalized load equivalent model of the typical electrolytic cell, obtaining sample values ​​of continuous values ​​of parameters of the generalized load equivalent model of the typical electrolytic cell based on a time series, and fitting a parameter curve equation according to the sample values, comprises: The continuous values ​​of the parameters based on the time series collected by the electrolytic cell load data collector are obtained, and the continuous values ​​of the parameters are standardized to obtain sample values ​​of the continuous values ​​of the parameters.

[0016] A parameter curve equation is obtained by adopting a radial basis neural network training method according to the sample values.

[0017] Furthermore, after obtaining the parameter curve equation by using the radial basis neural network training method according to the sample values, the method further includes: The integral value of the parameter curve equation is calculated by using the rectangular method. If the deviation between the integral value and the integral value of the actual parameter is greater than a preset deviation threshold, the parameter curve equation is adjusted by using a feedback compensator.

[0018] Furthermore, the step of obtaining the fitting inclination value of the continuous value of the parameter of the generalized load equivalent model of the typical electrolytic cell at each moment according to the parameter curve equation includes: A sampling window is set to sample the parameter curve equation to obtain a plurality of sampling points.

[0019] A parameter variation is obtained by a second-order central difference calculator according to the sampling points, a derivative value is obtained according to the parameter variation, and the inclination angle between the sampling points is calculated according to the derivative value.

[0020] A control point sequence is constructed according to each of the inclination angles by using a Bezier curve fitter, and an inclination continuous value curve is obtained by using a cubic spline interpolator according to the control point sequence.

[0021] The inclination continuous value curve is reconstructed to obtain the fitting inclination value of the parameter continuous value of the generalized load equivalent model of the typical electrolytic cell at each moment.

[0022] Furthermore, the density peak clustering method is used to cluster the fitting inclination values, and the corresponding sample values ​​are divided into several parameter classes according to the clustering results, including: The sample values ​​are subjected to zero-mean normalization processing according to each of the fitting inclination angle values ​​to obtain a normalized parameter sequence.

[0023] The local density value and relative distance value of each of the fitting inclination values ​​are calculated using a density peak clustering calculator according to the standardized parameter sequence to generate a plurality of parameter classes.

[0024] Furthermore, the adaptive discretization processing is performed on the sample values ​​under each parameter class to obtain the discretized parameter value, including: The sample values ​​under each of the parameter classes are discretized, and the mean and variance indicators under the current discretization interval are calculated according to the results of the discretization processing. It is determined whether the discretization effect meets the preset discretization accuracy value according to the calculation results; if not, the discretization interval is adjusted, and the sample values ​​under each of the parameter classes are discretized again until the discretization effect meets the preset requirements.

[0025] Further, the discretized parameter value is applied to the real-time electrolytic cell generalized load equivalent model, the real-time electrolytic cell generalized load equivalent model is run, and the parameter fitting inclination angle corresponding to the current electrolytic cell state is obtained, and the current electrolytic cell state is calculated and judged according to the parameter class corresponding to the parameter fitting inclination angle, including: The real-time electrolytic cell generalized load equivalent model is run based on the discretized parameter value, and the parameter fitting inclination is calculated according to each fitting inclination value stored in the circular queue by recursive least squares method.

[0026] Obtain a parameter class corresponding to the parameter fitting inclination, obtain a real-time parameter continuous value corresponding to the real-time operation data according to the parameter class, and calculate a value reflecting the current state of the electrolytic cell through a mean square error calculator according to the real-time parameter continuous value.

[0027] Furthermore, the method further comprises: The calculation accuracy of the real-time electrolytic cell generalized load equivalent model is evaluated, and the monitoring effect of the real-time electrolytic cell generalized load equivalent model on the electrolytic cell state is verified based on the evaluation result.

[0028] Another embodiment of the present invention provides an electrolytic cell state monitoring system based on a generalized load equivalent model, comprising: The typical model building module is used to build a generalized load equivalent model of a typical electrolytic cell based on the historical operation data of the electrolytic cell to be monitored.

[0029] A curve fitting module is used to run the generalized load equivalent model of the typical electrolytic cell, obtain sample values ​​of continuous parameter values ​​of the generalized load equivalent model of the typical electrolytic cell based on a time series, and fit a parameter curve equation according to the sample values; the continuous parameter values ​​include at least a current density parameter, a cell voltage parameter and an electrolyte concentration parameter.

[0030] The inclination angle calculation module is used to obtain the fitting inclination angle value of the continuous value of the parameter of the generalized load equivalent model of the typical electrolytic cell at each moment according to the parameter curve equation.

[0031] The parameter classification module is used to cluster the fitting inclination values ​​by using the density peak clustering method, and divide the corresponding sample values ​​into several parameter classes according to the clustering results.

[0032] The parameter discretization module is used to perform adaptive discretization processing on the sample values ​​under each parameter class to obtain discretized parameter values; the adaptive discretization processing is designed to dynamically adjust the discretization interval according to the distribution range and change trend of the continuous value of the parameter in the generalized load equivalent model of the typical electrolytic cell.

[0033] The real-time model building module is used to obtain the real-time operation data of the electrolytic cell to be monitored and to build a real-time electrolytic cell generalized load equivalent model based on the real-time operation data.

[0034] The state monitoring module is used to apply the discretized parameter value to the real-time electrolytic cell generalized load equivalent model, run the real-time electrolytic cell generalized load equivalent model, obtain the parameter fitting inclination corresponding to the current electrolytic cell state, and calculate and judge the current electrolytic cell state according to the parameter class corresponding to the parameter fitting inclination.

[0035] Compared with the prior art, the embodiments of the present invention have the following advantages: In the parameter selection of the generalized load equivalent model of the electrolytic cell, the continuous parameter values ​​obtained based on the fitting inclination are discretized and applied to the generalized load equivalent model of the electrolytic cell. This can simplify the model calculation, make it easier to analyze the current density distribution and changing trend in the electrolytic cell, and make a more accurate and rapid judgment on the state of the electrolytic cell. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] Figure 1 A flow chart of the steps of a method for monitoring the state of an electrolytic cell based on a generalized load equivalent model provided by an embodiment of the present invention; Figure 2 A structural block diagram of an electrolytic cell state monitoring system based on a generalized load equivalent model provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0037] 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. The purpose of providing these embodiments is to make the disclosure of the present invention more thorough and comprehensive. 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.

[0038] In the description of this application, the terms "first", "second", "third", etc. are used for descriptive purposes only and are not to be understood as indicating or implying relative importance or implicitly indicating the number of the indicated technical features. Thus, a feature defined as "first", "second", "third", etc. may explicitly or implicitly include one or more of the feature. In the description of this application, unless otherwise specified, "plurality" means two or more.

[0039] In the description of the present application, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense. For example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be a connection between the two elements. The terms "vertical", "horizontal", "left", "right", "upper", "lower" and similar expressions used herein are only for illustrative purposes, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore cannot be understood as a limitation on the present invention. The term "and / or" used herein includes any and all combinations of one or more related listed items. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood according to specific circumstances.

[0040] In the description of this application, it should be noted that, unless otherwise defined, all technical and scientific terms used in the present invention have the same meaning as those commonly understood by those skilled in the art. The terms used in the specification of the present invention are only for the purpose of describing specific embodiments and are not intended to limit the present invention. For those of ordinary skill in the art, the specific meanings of the above terms in this application can be understood by specific circumstances.

[0041] An embodiment of the present invention provides a method for monitoring the state of an electrolytic cell based on a generalized load equivalent model. For details, see Figure 1 , Figure 1 The flowchart is a step flow chart of an electrolytic cell state monitoring method based on a generalized load equivalent model in one embodiment of the present invention, including steps S11 to S17: Step S11, constructing a generalized load equivalent model of a typical electrolytic cell based on the historical operation data of the electrolytic cell to be monitored.

[0042] Specifically, historical operation data of the electrolyzer to be monitored under normal operating conditions and abnormal operating conditions are obtained to obtain normal operating data and abnormal operating data. Normal operating data provides a performance benchmark for the electrolyzer under standard or expected conditions. By collecting these data, key parameters such as voltage, current, and temperature of the electrolyzer under normal operating conditions can be understood, and abnormal operating data reveals the performance of the electrolyzer under fault or abnormal conditions.

[0043] The abnormal operation data and the normal operation data are compared and analyzed to obtain the abnormal characteristic value of the abnormal operation data relative to the normal operation data. By comparing the normal operation data, it is easier to identify the abnormal characteristics and improve the accuracy of the subsequent monitoring model.

[0044] The continuous values ​​of the parameters are determined according to the abnormal characteristic values, and the generalized load equivalent model of the typical electrolyzer is constructed according to the continuous values ​​of the parameters. Based on the abnormal characteristic values, a series of continuous values ​​of parameters related to the performance of the electrolyzer can be determined. These continuous values ​​of parameters can more comprehensively describe the performance of the electrolyzer under different conditions and provide key inputs for subsequent modeling.

[0045] Step S12, running the generalized load equivalent model of the typical electrolytic cell, obtaining sample values ​​of continuous parameter values ​​of the generalized load equivalent model of the typical electrolytic cell based on the time series, and fitting the parameter curve equation according to the sample values; the continuous parameter values ​​include at least current density parameters, cell voltage parameters and electrolyte concentration parameters.

[0046] The continuous values ​​of the parameters based on the time series collected by the electrolytic cell load data collector are obtained. The continuous values ​​based on the time series can comprehensively reflect the operating status of the electrolytic cell at different time points, as well as the changes in the operating status. The continuous values ​​of the parameters are standardized to eliminate the unit and dimension differences between different parameters so that they have a unified scale and distribution, and the sample values ​​of the continuous values ​​of the parameters are obtained.

[0047] The parameter curve equation is obtained according to the sample values ​​using the radial basis function neural network (RBFNN) training method. RBFNN has a strong nonlinear approximation capability and can accurately capture the complex laws of changes in electrolytic cell parameters over time.

[0048] The integral value of the parameter curve equation is calculated using the rectangular method. If the deviation between the integral value and the integral value of the actual parameter is greater than the preset deviation threshold, the parameter curve equation is adjusted using a feedback compensator. By comparing the integral value and adjusting the parameter curve equation, the prediction accuracy of the electrolyzer performance can be continuously improved. This is crucial for the optimized operation, fault diagnosis and performance prediction of the electrolyzer.

[0049] Step S13, obtaining the fitting inclination value of the continuous value of the parameter of the generalized load equivalent model of the typical electrolytic cell at each moment according to the parameter curve equation.

[0050] The sampling window is set to sample the parameter curve equation and obtain several sampling points. By setting the sampling window, we can determine at which time points or intervals to sample the parameter curve equation. Through reasonable sampling, we can ensure that the main features of the parameter curve equation are captured while reducing the amount of calculation.

[0051] The parameter change is obtained through the second-order central difference calculator according to the sampling points, and the derivative value is obtained according to the parameter change. By calculating the derivative value, the changing trend of the parameter curve at each sampling point can be understood. The inclination angle between each sampling point is calculated according to the derivative value. The inclination angle is a measure of the slope change between adjacent sampling points, which reflects the degree of curvature of the parameter curve at these points.

[0052] A control point sequence is constructed according to each inclination angle through the Bezier curve fitter. The Bezier curve fitter uses the given inclination angle data to construct a control point sequence. These control points define the shape of the curve. Then, the cubic spline interpolator uses these control points to generate a smooth and continuous inclination continuous value curve. This process ensures that the inclination curve is continuous between sampling points and has a smooth change trend.

[0053] The high-frequency noise and interference in the inclination continuous value curve are removed by a bandpass filter, and the low-frequency components related to the changes in parameters such as the generalized load of the electrolytic cell are retained to improve the accuracy and reliability of the data. Then, the inclination continuous value curve is scaled using the wavelet transform method to decompose the inclination continuous value curve into components of different scales, which represent the characteristics of the curve in different frequency ranges.

[0054] Based on the scale decomposition results of wavelet transform, components of different scales are recombined through reconstruction operation to obtain more accurate and reliable fitting inclination values ​​of electrolytic cell generalized load and other parameters. These fitting inclination values ​​not only reflect the actual situation of parameter changes, but also have high time and spatial resolution, which helps to achieve more accurate electrolytic cell state monitoring and performance prediction.

[0055] Step S14, clustering the fitting inclination values ​​by using the density peak clustering method, and dividing the corresponding sample values ​​into a plurality of parameter classes according to the clustering results.

[0056] According to each fitting inclination value, the sample values ​​of the continuous values ​​of the parameter dimensions including the current density parameter, the cell voltage parameter and the electrolyte concentration parameter are normalized to zero mean, eliminating the dimensional differences and numerical range differences between different parameters and obtaining a standardized parameter sequence.

[0057] According to the standardized parameter sequence, the local density value and relative distance value of each fitted inclination value are calculated using the density peak clustering calculator to generate several parameter classes. The density peak clustering algorithm is an algorithm for clustering based on the local density and relative distance of data points. Specifically, it is assumed that the cluster centers are those points with high local density and relatively long distances from other high-density points. By calculating the local density value and relative distance value of each fitted inclination value, potential cluster centers can be identified, and the parameter sequence can be divided into several parameter classes based on these centers. These parameter classes represent the characteristic types of different electrolyzer operating states pointed to by the sample values ​​of different parameter continuous values.

[0058] Step S15, adaptively discretize the sample values ​​under each parameter class to obtain discretized parameter values; the adaptive discretization process is designed to dynamically adjust the discretization interval according to the distribution range and change trend of the continuous values ​​of the parameters in the generalized load equivalent model of the typical electrolytic cell.

[0059] The sample values ​​under each parameter class are discretized. Discretization converts continuous sample values ​​into discrete, finite categories or intervals, thereby simplifying data representation. This not only reduces data storage requirements, but also extracts the main features of the data, which is helpful for subsequent analysis and processing.

[0060] The discretized parameter values ​​are divided into several sub-layers according to the numerical values. The sample points extracted from each sub-layer are counted by the kernel density calculator. The parameter continuous value distribution curve is obtained according to the distribution characteristics of the extracted sample points. The inflection point position where the tangent slope of the curve is greater than the preset slope threshold is calculated by fitting the parameter change trend through the Chebyshev polynomial, and the initial interval boundary sequence is obtained. Among them, the preset slope threshold is set based on the fluctuation critical threshold when the fluctuation state of the electrolytic cell data is judged to be abnormal.

[0061] A sliding time window is used to calculate the mean and variance of the initial interval boundary sequence. If the mean and variance exceed the preset range, it is judged that the indicator under the current discretization interval does not meet the preset accuracy requirements. The discretization interval is adjusted for resampling and re-discretization processing is performed until the final discretization effect meets the preset accuracy requirements or the number of discretizations reaches the preset maximum number.

[0062] Step S16, obtaining the real-time operation data of the electrolytic cell to be monitored, and constructing a real-time electrolytic cell generalized load equivalent model according to the real-time operation data.

[0063] In the actual application of electrolytic cell status monitoring, the status of the electrolytic cell is analyzed according to the actual values ​​to achieve the electrolytic cell monitoring effect.

[0064] Step S17, applying the discretized parameter value to the real-time electrolytic cell generalized load equivalent model, running the real-time electrolytic cell generalized load equivalent model, obtaining the parameter fitting inclination corresponding to the current electrolytic cell state, and calculating and judging the current electrolytic cell state according to the parameter class corresponding to the parameter fitting inclination.

[0065] The fitted inclination values ​​are stored in a circular queue. The circular queue can effectively store and manage a series of data points. Due to its cyclic characteristics, old data will be overwritten by new data, thereby ensuring that the queue always contains the latest data points.

[0066] The real-time electrolytic cell generalized load equivalent model is run based on the discretized parameter values, and each fitting inclination value is calculated by the recursive least squares method to obtain the parameter fitting inclination that best represents the current electrolytic cell operation data trend.

[0067] The parameter clustering category corresponding to the parameter fitting inclination is obtained, and the real-time parameter continuous value corresponding to the real-time operation data is obtained according to the parameter class. The value reflecting the current state of the electrolytic cell is calculated by the mean square error calculator based on the real-time parameter continuous value, so as to determine whether the electrolytic cell has a fault and the type of fault of the electrolytic cell.

[0068] Preferably, in order to further improve the effect of electrolytic cell state monitoring, this embodiment also includes: Step S18, evaluate the calculation accuracy of the real-time electrolytic cell generalized load equivalent model, and verify the monitoring effect of the real-time electrolytic cell generalized load equivalent model on the electrolytic cell state based on the evaluation result. Specifically, use a long short-term memory network to predict the load parameters of the real-time electrolytic cell generalized load equivalent model in time series to obtain a prediction parameter sequence. Based on the prediction parameter sequence, a root mean square error calculator is used in the parameter continuous value space to calculate the prediction accuracy value of the prediction parameter sequence.

[0069] The electrolytic cell state monitoring method based on the generalized load equivalent model of the present invention discretizes the continuous parameter values ​​obtained based on the fitting inclination in the parameter selection of the generalized load equivalent model of the electrolytic cell, and applies it to the generalized load equivalent model of the electrolytic cell. This can simplify the model operation, more easily analyze the current density distribution and changing trend in the electrolytic cell, and make a more accurate and rapid judgment on the state of the electrolytic cell.

[0070] The embodiment of the present invention further provides an electrolytic cell state monitoring system based on a generalized load equivalent model, which is used to execute the electrolytic cell state monitoring method based on a generalized load equivalent model as described above. Figure 2 The structure block diagram of the electrolytic cell state monitoring system based on the generalized load equivalent model according to an embodiment of the present invention, the system comprises: The typical model building module 21 is used to build a generalized load equivalent model of a typical electrolytic cell based on the historical operation data of the electrolytic cell to be monitored.

[0071] The curve fitting module 22 is used to run the generalized load equivalent model of the typical electrolytic cell, obtain sample values ​​of the continuous values ​​of parameters of the generalized load equivalent model of the typical electrolytic cell based on the time series, and fit the parameter curve equation according to the sample values; the continuous values ​​of the parameters include at least current density parameters, cell voltage parameters and electrolyte concentration parameters.

[0072] The inclination angle calculation module 23 is used to obtain the fitting inclination angle value of the continuous value of the parameter of the generalized load equivalent model of the typical electrolytic cell at each moment according to the parameter curve equation.

[0073] The parameter classification module 24 is used to cluster the fitting inclination values ​​by using the density peak clustering method, and divide the corresponding sample values ​​into a plurality of parameter classes according to the clustering results.

[0074] The parameter discretization module 25 is used to perform adaptive discretization processing on the sample values ​​under each parameter class to obtain discretized parameter values; the adaptive discretization processing is designed to dynamically adjust the discretization interval according to the distribution range and change trend of the continuous value of the parameter in the generalized load equivalent model of the typical electrolytic cell.

[0075] The real-time model building module 26 is used to obtain the real-time operation data of the electrolytic cell to be monitored, and to build a real-time electrolytic cell generalized load equivalent model according to the real-time operation data.

[0076] The state monitoring module 27 is used to apply the discretized parameter value to the real-time electrolytic cell generalized load equivalent model, run the real-time electrolytic cell generalized load equivalent model, obtain the parameter fitting inclination corresponding to the current electrolytic cell state, and calculate and judge the current electrolytic cell state according to the parameter class corresponding to the parameter fitting inclination.

[0077] The technical features and technical effects of the system proposed in the embodiment of the present invention are the same as those of the method proposed in the embodiment of the present invention, and are not described in detail here. Each module in the above system can be implemented in whole or in part by software, hardware, and a combination thereof. Each of the above modules can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each of the above modules.

[0078] The above-mentioned embodiments only express several implementation methods of the present invention, and the description thereof is relatively specific and detailed, but it cannot be understood as limiting the scope of the patent of the present invention. It should be pointed out that, for ordinary technicians in this field, several variations and improvements can be made without departing from the concept of the present invention, which all belong to the protection scope of the present invention. Therefore, the protection scope of the patent of the present invention shall be subject to the attached claims.

Claims

1. A method for monitoring the state of an electrolytic cell based on a generalized load equivalent model, characterized in that: include: Based on the historical operation data of the electrolytic cell to be monitored, a generalized load equivalent model of a typical electrolytic cell is constructed; Running the generalized load equivalent model of the typical electrolytic cell, obtaining sample values ​​of continuous parameter values ​​of the generalized load equivalent model of the typical electrolytic cell based on a time series, and fitting a parameter curve equation according to the sample values; the continuous parameter values ​​at least include a current density parameter, a cell voltage parameter, and an electrolyte concentration parameter; Obtaining the fitting inclination value of the parameter continuous value of the generalized load equivalent model of the typical electrolytic cell at each moment according to the parameter curve equation; Clustering the fitted inclination values ​​using a density peak clustering method, and dividing the corresponding sample values ​​into a number of parameter classes according to the clustering results; Adaptively discretizing the sample values ​​under each parameter class to obtain discretized parameter values; the adaptive discretization process is designed to dynamically adjust the discretization interval according to the distribution range and change trend of the continuous value of the parameter in the generalized load equivalent model of the typical electrolytic cell; Acquire the real-time operation data of the electrolytic cell to be monitored, and construct a real-time electrolytic cell generalized load equivalent model according to the real-time operation data; The discretized parameter value is applied to the real-time electrolytic cell generalized load equivalent model, and the real-time electrolytic cell generalized load equivalent model is run to obtain the parameter fitting inclination angle corresponding to the current electrolytic cell state, and the current electrolytic cell state is calculated and judged according to the parameter class corresponding to the parameter fitting inclination angle.

2. The electrolytic cell state monitoring method based on the generalized load equivalent model according to claim 1, characterized in that: The method of constructing a generalized load equivalent model of a typical electrolytic cell based on the historical operation data of the electrolytic cell to be monitored includes: Acquire historical operation data of the electrolytic cell to be monitored in normal operation state and abnormal operation state, and obtain normal operation data and abnormal operation data; Comparing and analyzing the abnormal operation data with the normal operation data, to obtain an abnormal characteristic value of the abnormal operation data relative to the normal operation data; A parameter continuous value is determined according to the abnormal characteristic value, and a generalized load equivalent model of a typical electrolytic cell is constructed according to the parameter continuous value.

3. The electrolytic cell state monitoring method based on the generalized load equivalent model according to claim 1, characterized in that: The step of running the generalized load equivalent model of the typical electrolytic cell, obtaining sample values ​​of continuous values ​​of parameters of the generalized load equivalent model of the typical electrolytic cell based on a time series, and fitting a parameter curve equation according to the sample values ​​includes: Acquire the continuous values ​​of the parameters based on the time series collected by the electrolytic cell load data collector, perform standardization on the continuous values ​​of the parameters, and obtain sample values ​​of the continuous values ​​of the parameters; A parameter curve equation is obtained by adopting a radial basis neural network training method according to the sample values.

4. The electrolytic cell state monitoring method based on the generalized load equivalent model according to claim 3 is characterized in that: After obtaining the parameter curve equation by using the radial basis neural network training method according to the sample values, the method further includes: The integral value of the parameter curve equation is calculated by using the rectangular method. If the deviation between the integral value and the integral value of the actual parameter is greater than a preset deviation threshold, the parameter curve equation is adjusted by using a feedback compensator.

5. The electrolytic cell state monitoring method based on the generalized load equivalent model according to claim 1, characterized in that: The step of obtaining the fitting inclination value of the continuous value of the parameter of the generalized load equivalent model of the typical electrolytic cell at each moment according to the parameter curve equation includes: Setting a sampling window to sample the parameter curve equation to obtain a plurality of sampling points; Obtaining a parameter change amount through a second-order central difference calculator according to the sampling points and obtaining a derivative value according to the parameter change amount, and calculating the inclination angle between each of the sampling points according to the derivative value; Constructing a control point sequence by a Bezier curve fitter according to each of the inclination angles, and obtaining an inclination continuous value curve by a cubic spline interpolator according to the control point sequence; The inclination continuous value curve is reconstructed to obtain the fitting inclination value of the parameter continuous value of the generalized load equivalent model of the typical electrolytic cell at each moment.

6. The electrolytic cell state monitoring method based on the generalized load equivalent model according to claim 5, characterized in that: The density peak clustering method is used to cluster the fitting inclination values, and the corresponding sample values ​​are divided into several parameter classes according to the clustering results, including: Performing zero-mean normalization processing on the sample values ​​according to the fitting inclination values ​​to obtain a normalized parameter sequence; The local density value and relative distance value of each of the fitting inclination values ​​are calculated using a density peak clustering calculator according to the standardized parameter sequence to generate a plurality of parameter classes.

7. The electrolytic cell state monitoring method based on the generalized load equivalent model according to claim 1, characterized in that: The step of adaptively discretizing the sample values ​​under each parameter class to obtain a discretized parameter value includes: The sample values ​​under each of the parameter classes are discretized, and the mean and variance indicators under the current discretization interval are calculated according to the results of the discretization processing. It is determined whether the discretization effect meets the preset discretization accuracy value according to the calculation results; if not, the discretization interval is adjusted, and the sample values ​​under each of the parameter classes are discretized again until the discretization effect meets the preset requirements.

8. The electrolytic cell state monitoring method based on the generalized load equivalent model according to claim 1, characterized in that: The step of applying the discretized parameter value to the real-time electrolytic cell generalized load equivalent model, running the real-time electrolytic cell generalized load equivalent model, obtaining a parameter fitting inclination angle corresponding to the current electrolytic cell state, and calculating and judging the current electrolytic cell state according to the parameter class corresponding to the parameter fitting inclination angle includes: Running the real-time electrolytic cell generalized load equivalent model based on the discretized parameter value, and calculating the parameter fitting inclination angle according to each fitting inclination angle value stored in the circular queue by recursive least square method; Obtain a parameter class corresponding to the parameter fitting inclination, obtain a real-time parameter continuous value corresponding to the real-time operation data according to the parameter class, and calculate a value reflecting the current state of the electrolytic cell through a mean square error calculator according to the real-time parameter continuous value.

9. The electrolytic cell state monitoring method based on the generalized load equivalent model according to claim 1, characterized in that: The method further comprises: The calculation accuracy of the real-time electrolytic cell generalized load equivalent model is evaluated, and the monitoring effect of the real-time electrolytic cell generalized load equivalent model on the electrolytic cell state is verified based on the evaluation result.

10. An electrolytic cell state monitoring system based on a generalized load equivalent model, characterized in that: include: A typical model building module is used to build a generalized load equivalent model of a typical electrolyzer based on the historical operation data of the electrolyzer to be monitored; A curve fitting module is used to run the generalized load equivalent model of the typical electrolytic cell, obtain sample values ​​of continuous values ​​of parameters of the generalized load equivalent model of the typical electrolytic cell based on a time series, and fit a parameter curve equation according to the sample values; the continuous values ​​of parameters include at least a current density parameter, a cell voltage parameter, and an electrolyte concentration parameter; An inclination calculation module is used to obtain the fitting inclination value of the continuous value of the parameter of the generalized load equivalent model of the typical electrolytic cell at each moment according to the parameter curve equation; A parameter classification module, used for clustering the fitting inclination values ​​by using a density peak clustering method, and dividing the corresponding sample values ​​into a number of parameter classes according to the clustering results; A parameter discretization module is used to perform adaptive discretization processing on the sample values ​​under each parameter class to obtain a discretized parameter value; the adaptive discretization processing is designed to dynamically adjust the discretization interval according to the distribution range and change trend of the continuous value of the parameter in the generalized load equivalent model of the typical electrolytic cell; A real-time model building module, used to obtain the real-time operation data of the electrolytic cell to be monitored, and to build a real-time electrolytic cell generalized load equivalent model according to the real-time operation data; The state monitoring module is used to apply the discretized parameter value to the real-time electrolytic cell generalized load equivalent model, run the real-time electrolytic cell generalized load equivalent model, obtain the parameter fitting inclination corresponding to the current electrolytic cell state, and calculate and judge the current electrolytic cell state according to the parameter class corresponding to the parameter fitting inclination.

Images