Carbon dioxide electrolysis system based on pulse voltage regulation and control

Abstract

The invention discloses a carbon dioxide electrolysis system based on pulse voltage regulation and control. The system comprises a pulse voltage source module, an electrolytic reaction module, a parameter regulation and control module, a time sequence optimization module, a reference setting module and a decision execution module. The pulse voltage source module generates an adjustable pulse voltage sequence and sets a parameter interval; the electrolytic reaction module divides reaction units and analyzes current response to generate feature vectors; the parameter regulation and control module extracts key indexes to establish voltage regulation rules; a time sequence optimization module identifies a time correlation mode dynamic calibration index and calculates a voltage offset; the reference setting module deduces an optimal pulse threshold value and generates a deviation sequence; and the decision execution module analyzes the deviation sequence to generate a regulation and control scheme. The system realizes pulse voltage dynamic optimization and reaction unit refined management through multi-module cooperation, improves the selectivity, stability and energy efficiency of carbon dioxide electrolysis, and is suitable for the field of efficient conversion of carbon dioxide.

Description

Technical Field

[0001] The present invention relates to the technical field of carbon dioxide electrolysis, and specifically to a carbon dioxide electrolysis system based on pulse voltage regulation. Background Art

[0002] Electrolysis has become one of the important ways to achieve carbon cycle utilization because it can convert carbon dioxide into high-value chemicals or fuels (such as carbon monoxide, formic acid, hydrocarbons, etc.). However, traditional carbon dioxide electrolysis technologies face many challenges in practical applications, which limit their large-scale promotion and efficient use.

[0003] From the perspective of reaction mechanism, the carbon dioxide molecular structure is stable, and a relatively high activation energy needs to be overcome during electrolysis, resulting in slow reaction kinetics. The traditional constant voltage electrolysis mode is difficult to accurately regulate the reaction path, easily triggering side reactions and reducing the selectivity of target products and the Faraday efficiency. For example, in different voltage ranges, carbon dioxide may undergo multiple reduction reactions to produce different products, and the constant voltage cannot be dynamically adjusted according to the reaction process, leading to a complex product distribution and difficult separation and purification.

[0004] In terms of system stability, traditional electrolysis systems have a lagging response to parameter fluctuations during the reaction process. When key parameters such as current, temperature, and pressure change, the system cannot adjust in time, which may lead to instability of the electrolysis process and even pose safety hazards. For example, a sudden fluctuation in current may cause a sharp increase in the temperature of the electrode surface, accelerating electrode corrosion and shortening the service life of the equipment.

[0005] From the perspective of energy utilization efficiency analysis, in the constant voltage mode, the energy input during electrolysis cannot be matched with the reaction demand in real time, resulting in energy waste. In the initial stage of the reaction, relatively high energy may be required to activate carbon dioxide molecules, while in the later stage of the reaction, lower energy can maintain the reaction, but the constant voltage cannot achieve this dynamic adjustment, leading to low overall energy efficiency.

[0006] Traditional electrolysis systems lack the ability to conduct refined analysis and regulation of the reaction process. It is impossible to obtain detailed characteristic information of the reaction unit in real time, and it is difficult to establish an accurate reaction model, thus unable to achieve optimized control of the electrolysis process. For example, different reaction stages may require different voltage waveforms and parameter settings, while traditional systems cannot perform personalized regulation according to the characteristics of the reaction unit.

[0007] In the prior art, although there are some studies on voltage regulation, most of them adopt simple pulse voltage or segmented voltage control, lacking dynamic optimization of pulse parameters and refined management of reaction units. For example, some studies only use pulse voltages with fixed frequencies and amplitudes to improve the reaction effect, without considering the differences between different reaction units and the real-time correlation between pulse parameters and reaction characteristics. Therefore, there is an urgent need for a carbon dioxide electrolysis system that can achieve dynamic regulation of pulse voltage, refined management of reaction units, and real-time optimization of parameters according to the reaction process, so as to improve the selectivity, stability, and energy utilization efficiency of the reaction and promote the practical application of carbon dioxide electrolysis technology. Summary of the Invention

[0008] The purpose of the present invention is to provide a carbon dioxide electrolysis system based on pulse voltage regulation to solve the problems raised in the above background technology.

[0009] To achieve the above purpose, the present invention provides the following technical solutions: A carbon dioxide electrolysis system based on pulse voltage regulation, the system includes: A pulse voltage source module, used to generate a controllable pulse voltage sequence and set a pulse parameter interval corresponding to the electrolysis reaction; An electrolysis reaction module, used to divide multiple reaction units within the pulse parameter interval, perform current response analysis on the electrolysis process of each reaction unit, and generate an electrolysis characteristic vector corresponding to the reaction unit; A parameter regulation module, used to extract key reaction indicators from the electrolysis characteristic vector, establish a voltage regulation rule associated with the reaction unit, and obtain the pulse correction parameters corresponding to the rule; A timing optimization module, used to identify the time correlation pattern in the pulse correction parameters, dynamically calibrate the key reaction indicators according to the correlation pattern, and calculate the voltage offset of each reaction unit under different calibration strategies; A benchmark setting module, used to deduce the optimal pulse threshold according to the voltage offset, and generate an electrolysis deviation sequence by comparing the current electrolysis characteristic value with the optimal pulse threshold; A decision execution module, used to analyze the electrolysis deviation sequence, and integrate the electrolysis deviation sequence into an electrolysis regulation execution plan based on the voltage offset trend of the reaction unit.

[0010] Preferably, the implementation method of the electrolysis reaction module includes: constructing an electrolysis characteristic library corresponding to the reaction unit, and the electrolysis characteristic library includes current data and a reaction parameter vector mapped by the electrolysis response; Perform similar response matching on the reaction parameter vector, divide the reaction clustering group of the reaction parameter vector according to the matching result; extract the electrolysis focus point of the current data from the reaction clustering group, and set the focus point as the reaction unit.

[0011] Preferably, the reaction clustering groups for dividing the reaction parameter vectors further include: Extract the voltage distribution, reaction polarity, and response density parameters according to the time attribute and current intensity in the reaction parameter vector, and generate reaction feature labels based on the above parameters; Associate the reaction feature labels with the reaction parameter vectors, and filter the parameter vectors with a similarity higher than the preset response threshold by calculating the response similarity between the feature labels to form reaction clustering groups.

[0012] Preferably, the implementation method for generating the electrolysis feature vectors corresponding to the reaction units includes: For each reaction unit, obtain the current fluctuation data of the unit within a preset window according to the timing position of the unit in the pulse parameter interval, and calculate the current fluctuation coefficient of the unit; When the current fluctuation coefficient exceeds the first response threshold, mark the unit as a high-response unit and extract its current data to form an electrolysis feature vector; when the current fluctuation coefficient is lower than the first response threshold, mark the unit as a steady-state unit, and perform response superposition on the current data of the adjacent units of the unit, and reconstruct the superposed data into an electrolysis feature vector.

[0013] Preferably, the implementation method of the parameter regulation module includes: Separate the positive reaction ratio, negative reaction ratio, and neutral response parameters from the electrolysis feature vector, and generate a voltage regulation rule for the reaction unit based on the above parameters; If the number of reaction units covered by the current voltage regulation rule is less than the preset response threshold, traverse the electrolysis feature vectors of the adjacent reaction units, and add the response indicators not included in the regulation rules of the adjacent units to the current rule.

[0014] Preferably, the implementation method of the timing optimization module includes: obtaining the sequence factor of the time switching frequency and the change factor of the response jump amplitude in the time correlation pattern; Construct a weight distribution matrix associated with the sequence factor and the change factor, and determine the voltage offset under different calibration strategies according to the distribution probability of each element in the matrix.

[0015] Preferably, constructing the weight distribution matrix further includes: Identify the time period characteristics of the sequence factor. If the current period characteristics exactly match the preset time period, set the sequence factor as the starting index of the weight distribution matrix; Calculate the correlation matching degree between the sequence factor and the change factor, and generate the middle index and the ending index of the weight distribution matrix in descending order of the matching degree; Perform path backtracking on the ending index. When the matching degree of the ending index is lower than the preset matching threshold, output it as the final distribution of the weight distribution matrix.

[0016] Preferably, the implementation method for calculating the voltage offset includes: Statistically analyze the sequence factor mean and the range of variation factors for each termination index in the weight distribution matrix, and calculate the global variance of all index factors; Subtract the sequence factor mean of a single termination index from the sequence factor mean of an adjacent index, and divide the result by the global variance to obtain the time series offset coefficient; at the same time, calculate the ratio of the range of variation factors to the global variance, and perform weighted summation of the two as the voltage offset of this index.

[0017] Preferably, the implementation method for deriving the optimal pulse threshold includes: Extract the reaction mode in the historical data that is closest to the current voltage offset, and calculate the cosine similarity between the two in terms of time distribution as the first dynamic reference value; Statistically analyze the difference in the number of response peaks between the current voltage offset and the historical reaction mode, and use the difference quantity as the second dynamic reference value; Based on the linear combination of the first dynamic reference value and the second dynamic reference value, match the optimal pulse threshold in the preset pulse threshold table.

[0018] Preferably, the implementation method of the decision execution module includes: dividing the positive offset interval and the negative offset interval according to the voltage offset direction of each unit in the electrolysis deviation sequence; Extract the convergence rate of the electrolysis deviation in the positive offset interval and the diffusion rate of the electrolysis deviation in the negative offset interval, and perform adversarial fusion of the two according to the time weight of the reaction unit to generate the adjustment parameter of the electrolysis regulation execution plan.

[0019] Compared with the prior art, the beneficial effects of the present invention are: In terms of the refined management of the reaction process, the electrolysis reaction module can divide multiple reaction units within the pulse parameter range, and perform current response analysis on the electrolysis process of each unit to generate corresponding electrolysis feature vectors. This process realizes in-depth analysis and classification of the reaction parameter vectors by constructing an electrolysis feature library, performing similar response matching, and dividing reaction clustering groups. For example, reaction feature tags are generated according to parameters such as time attributes and current intensity, and clustering groups are formed through response similarity screening, enabling the system to clearly identify the characteristics and rules of different reaction units. This refined management method changes the drawback of the traditional system's general processing of the reaction process, and can perform targeted regulation according to the characteristics of each reaction unit, providing a solid foundation for subsequent parameter optimization and voltage adjustment.

[0020] The ability of dynamic optimization of pulse parameters is a major highlight of this system. The parameter regulation module extracts key reaction indicators from the electrolysis feature vectors, establishes voltage regulation rules associated with the reaction units, and obtains pulse correction parameters. By separating the proportion of forward and negative reactions and neutral response parameters, the system can formulate personalized voltage regulation rules according to the actual situation of the reaction units. When the number of reaction units covered by the current rule is insufficient, it can automatically traverse the feature vectors of adjacent units to supplement the response indicators, ensuring the comprehensiveness and effectiveness of the regulation rules. This dynamic optimization mechanism enables the pulse voltage to adapt to the changes in the reaction process in real time, effectively improving the selectivity of the reaction, reducing the occurrence of side reactions, and thus enhancing the yield and purity of the target product.

[0021] The timing optimization and dynamic calibration functions further enhance the stability and adaptability of the system. The timing optimization module constructs a weight distribution matrix by identifying the time correlation patterns in the pulse correction parameters, and determines the voltage offset according to the distribution probability of the matrix elements. This process fully considers the influence of the time switching frequency and the response jump amplitude, and can accurately capture the time dynamic characteristics in the reaction process. For example, by identifying the time period characteristics of the sequence factors, determining the index of the weight distribution matrix, and calculating the correlation matching degree, the system can dynamically calibrate the key reaction indicators according to the time correlation patterns. This dynamic calibration mechanism can respond in a timely manner to the parameter fluctuations in the reaction process, adjust the voltage offsets of each reaction unit, ensure the stability of the electrolysis process, and avoid the instability problems caused by the lagged response of parameters in traditional systems.

[0022] The benchmark setting and deviation analysis provide a scientific basis for system regulation. The benchmark setting module derives the optimal pulse threshold based on the voltage offset, and generates an electrolysis deviation sequence by comparing the current electrolysis feature values with the optimal threshold. This process combines historical data and current reaction information, and can accurately determine the optimal pulse threshold by calculating dynamic reference values such as cosine similarity and response peak differences. The generation of the electrolysis deviation sequence enables the system to clearly understand the gap between the current reaction state and the optimal state, providing a clear regulation direction for the decision-making and execution module. This method of benchmark setting and deviation analysis based on data and models improves the scientificity and accuracy of system regulation, and avoids the blindness of traditional regulation methods.

[0023] The efficiency and pertinence of decision execution are the keys to the system achieving precise regulation. The decision execution module divides the positive and negative offset intervals according to the electrolysis deviation sequence, extracts the convergence rate and diffusion rate, and performs adversarial fusion according to the time weight to generate adjustment parameters. This processing method can target the reaction units with different offset directions and rates for targeted regulation, ensuring the efficiency and effectiveness of the regulation scheme. For example, different adjustment strategies are adopted for the units with a fast convergence rate in the positive offset interval and the units with a slow diffusion rate in the negative offset interval, enabling the system to quickly correct the deviation and return to the optimal reaction state. This efficient decision execution mechanism greatly improves the response speed and regulation accuracy of the system, ensuring the efficient and stable operation of the electrolysis process.

[0024] Through the collaborative work of each module, the system forms a complete closed loop from data collection, analysis to parameter optimization and regulation execution. It realizes the full-process dynamic management and precise regulation of the carbon dioxide electrolysis process, effectively solves the problems of slow reaction kinetics, low selectivity, poor stability and low energy efficiency in traditional electrolysis technologies, provides advanced technical support for the efficient conversion and utilization of carbon dioxide, has significant economic and environmental benefits, and is of great significance for promoting carbon cycle utilization and addressing climate change. Brief Description of the Drawings

[0025] Figure 1 It is the working principle diagram of the carbon dioxide electrolysis system based on pulse voltage regulation described in the present invention; Figure 2 It is the working principle diagram of the electrolysis reaction module; Figure 3 It is the working principle diagram of the reaction clustering group division; Figure 4 It is the working principle diagram of the electrolysis feature vector generation. Detailed Embodiments

[0026] Next, the technical solutions in the embodiments of the present invention will be clearly and completely described in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only a part of the embodiments of the present invention, rather than all the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those of ordinary skill in the art without creative efforts shall fall within the protection scope of the present invention.

[0027] Please refer to Figures 1-4 , the carbon dioxide electrolysis system based on pulse voltage regulation involved in the present invention includes a pulse voltage source module, an electrolysis reaction module, a parameter regulation module, a timing optimization module, a reference setting module and a decision execution module. Each module collaborates to realize the pulse voltage dynamic regulation of the carbon dioxide electrolysis process. The specific implementation is as follows: Pulse voltage source module: Configure a programmable pulse signal generator to generate a controllable pulse voltage sequence. By combining hardware circuit parameters and software algorithms, set the pulse parameter range corresponding to the electrolysis reaction, which includes the voltage amplitude range, pulse frequency range, and duty cycle range, to provide a basic excitation signal for the electrolysis reaction.

[0028] Electrolysis reaction module: Within the pulse parameter range, divide the continuous electrolysis process into multiple discrete reaction units. For each reaction unit, use a current sensor to collect the current response data of the electrolysis process in real time, and use signal processing algorithms such as Fourier transform and wavelet analysis to analyze the current response characteristics, generating an electrolysis feature vector containing parameters such as current peak value, rising edge time, and energy integral to characterize the reaction state of the unit.

[0029] Parameter regulation module: Extract key reaction indicators (such as the proportion of forward reactions, the proportion of reverse reactions, and neutral response parameters) from the electrolysis feature vector, and based on preset fuzzy logic rules or machine learning models, establish voltage regulation rules associated with the reaction unit. For example, when the proportion of forward reactions is lower than the threshold, generate a regulation rule to increase the pulse amplitude and obtain the pulse correction parameters corresponding to the rule (such as voltage increment, frequency adjustment).

[0030] Timing optimization module: Identify the time correlation patterns in the pulse correction parameters through timing analysis algorithms, such as the time interval rules for adjusting pulse parameters in different reaction stages. According to the correlation patterns, dynamically calibrate the key reaction indicators, and use algorithms such as Kalman filter to calculate the voltage offset of each reaction unit under different calibration strategies.

[0031] Reference setting module: Based on the voltage offset, deduce the optimal pulse threshold through historical data matching algorithms. Specifically, compare the current electrolysis characteristic values with the characteristic values of the historical optimal reaction mode to generate an electrolysis deviation sequence containing the voltage deviations of each reaction unit to reflect the difference between the current reaction state and the ideal state.

[0032] Decision execution module: Use the deviation analysis algorithm to analyze the electrolysis deviation sequence, and combine the voltage offset trend of the reaction unit to integrate the discrete deviation data into a continuous electrolysis control execution plan, and use the drive circuit to adjust the output parameters of the pulse voltage source module in real time to achieve closed-loop control of the electrolysis process.

[0033] The present invention will be further described below in conjunction with Embodiments 1 to 5:

[0034] Embodiment 1: In the electrolysis reaction module, constructing an electrolysis feature library corresponding to the reaction unit is the basic link of the implementation method. The electrolysis feature library stores the current data of each reaction unit in real time through a data acquisition system. This current data is a waveform sequence of continuous sampling, and the sampling frequency is synchronized with the output frequency of the pulse voltage source module to ensure the consistency of the data in the time dimension. For example, when the output frequency of the pulse voltage source module is 1000 Hz, the sampling frequency of the current data is set to not less than 2000 Hz to meet the Nyquist sampling theorem and avoid signal aliasing. The reaction parameter vector is generated through multi-sensor fusion technology. In addition to the current data, it also integrates the real-time data collected by devices such as temperature sensors, pressure sensors, and gas composition analyzers, and performs feature mapping through a neural network model to convert multi-dimensional physical quantities into a low-dimensional reaction parameter vector. This vector contains parameters such as temperature gradient, pressure fluctuation coefficient, and product selectivity coefficient that indirectly reflect the reaction state, realizing the multi-dimensional characterization of the electrolysis process.

[0035] When performing similar response matching on the reaction parameter vector, a clustering algorithm in machine learning is used to achieve data grouping. Specifically, first, the reaction parameter vector is preprocessed by standardization to eliminate the influence of different parameter dimensions on the clustering result. The standardization method can adopt Z-score standardization, that is, by calculating the mean and standard deviation of each parameter, the original data is converted into standard normal distribution data with a mean of 0 and a standard deviation of 1. The preprocessed data is input into the clustering algorithm model. Common algorithms include the K-means clustering algorithm, hierarchical clustering algorithm, or DBSCAN density clustering algorithm. Taking the K-means clustering algorithm as an example, first, K clustering centers are randomly initialized, and then the distance between each reaction parameter vector and each clustering center is calculated (the distance metric can adopt Euclidean distance or cosine distance), and the vector is assigned to the cluster to which the nearest clustering center belongs. Then, the centroid of each cluster is recalculated as the new clustering center, and the above assignment and update process are repeated until the clustering center no longer changes significantly or reaches the preset number of iterations, thereby completing the clustering division of the reaction parameter vector and forming different reaction clustering groups.

[0036] The electrolytic focus point for extracting current data from the reaction clustering group is a crucial step in determining the reaction unit. The extraction of the electrolytic focus point is based on the feature analysis of the current data, aiming to identify the key data points that can represent the typical reaction characteristics of the clustering group. In the specific operation, first, the current waveforms within each clustering group are superimposed and averaged to generate the typical current curve of the clustering group. Then, the typical current curve is differentiated to calculate the slope change of the curve, and the characteristic positions such as the rising edge, falling edge, and peak point of the current are identified. At the same time, the energy integral value of the current curve is calculated to determine the extreme point of energy consumption. These characteristic positions and extreme points are used as candidate focus points, and further screened through statistical analysis to select the points with a frequency higher than the preset threshold within the clustering group as the final electrolytic focus point. For example, if the current peak appears at 2 ms after the rising edge of the pulse voltage in 80% of the current waveforms within a certain clustering group, then this time point is determined as the electrolytic focus point of this clustering group.

[0037] When setting the focus point as the reaction unit, it needs to be defined in combination with the time dimension of the pulse parameter interval. The reaction unit can be the time point or time period corresponding to the focus point, depending on the time resolution requirements of the electrolysis process. For example, if the focus point is the time point when the current peak appears, the reaction unit can be defined as a short time period centered on this time point, such as a time window of 1 ms before and after, to ensure that the unit contains the complete peak characteristics. After dividing the reaction unit, each unit corresponds to a specific reaction stage in the electrolysis process, facilitating the subsequent refined analysis of the current response in each stage.

[0038] In the process of constructing the electrolysis feature library, the data storage structure is implemented using a database management system, such as MySQL or MongoDB, to support the efficient storage and rapid query of massive data. The database table structure design includes fields such as timestamp, current data field, reaction parameter vector field, and clustering label field. The data index is established through the timestamp to ensure the real-time access of each module to the data. At the same time, to avoid data redundancy, the reaction parameter vector adopts a serialized storage method, converting the multi-dimensional vector into a binary data stream for storage, reducing the storage space occupation.

[0039] In the process of similar response matching, the parameter settings of the clustering algorithm need to be adjusted according to the actual reaction characteristics. For example, for scenarios with a relatively fast change in the reaction state, the clustering number K value can be appropriately increased to improve the clustering fineness; for scenarios with a relatively stable reaction process, the K value can be reduced to lower the computational complexity. In addition, to avoid the K-means algorithm falling into a local optimal solution, a strategy of multiple random initializations of the clustering center and taking the optimal result can be adopted, or an optimization algorithm such as a genetic algorithm can be combined for the initialization of the clustering center.

[0040] The extraction of electrolysis focusing points can also be manually manipulated by incorporating domain knowledge. For example, based on the chemical reaction mechanism of carbon dioxide electrolysis, the current characteristic ranges corresponding to certain key reaction stages (such as carbon dioxide adsorption, electron transfer, and product desorption) can be predefined to guide the focusing point extraction process to prioritize the characteristic data of these stages. This combination of data-driven and knowledge-driven approaches can improve the scientific and practical nature of reaction unit delineation.

[0041] After the reaction units are divided, the electrolytic feature vectors of each unit need to be checked for consistency to ensure that the reaction units within the same cluster have similar reaction characteristics. This verification method can be used to calculate the intra-cluster distance of the electrolytic feature vectors of each unit within the cluster. If the intra-cluster distance exceeds a preset threshold, the clustering process is reviewed, and clustering parameters are adjusted or focal points are re-extracted until consistency requirements are met.

[0042] Example 2: When dividing the reaction parameter vector into reaction cluster groups, it is necessary to combine the time attribute and current intensity in the reaction parameter vector, further extract the voltage distribution, reaction polarity and response density parameters to generate reaction feature labels, and achieve accurate division of cluster groups by calculating the similarity of the feature labels. The specific implementation method is as follows: The extraction of time attributes is based on the timestamp information corresponding to the reaction parameter vector, analyzing the temporal position of the parameter vector in the electrolysis process, such as the rising edge, falling edge, plateau period, or intermittent period of the pulse voltage. The current intensity is directly taken from the real-time current data in the electrolysis feature library, reflecting the activity level of the electrolysis reaction at that moment. By analyzing the correlation between time attributes and current intensity, the differences in reaction characteristics at different stages can be identified. For example, the plateau period may correspond to a stable electrolysis reaction stage, while the rising or falling edge may be accompanied by violent current fluctuations, corresponding to key processes such as reactant adsorption or product desorption.

[0043] Extracting voltage distribution parameters requires combining the real-time voltage sequence output by the pulse voltage source module to calculate the fluctuation range of parameters such as voltage amplitude, pulse width, and duty cycle at the corresponding time point for each reaction parameter vector. For example, the maximum, minimum, and average voltage amplitudes over multiple consecutive pulse cycles are statistically analyzed to form a characteristic description of the voltage distribution. The reaction polarity parameter is determined by the correlation between current direction and voltage polarity. When the current direction matches the pulse voltage polarity, a positive reaction is considered dominant (e.g., a reduction reaction); otherwise, a negative reaction is considered dominant (e.g., an oxidation reaction). If the current fluctuation is small and lacks clear directionality, a neutral response state is defined. The response density parameter is calculated by counting valid reaction events per unit time. A valid reaction event can be defined as a fluctuation event whose current change amplitude exceeds a preset threshold. This threshold is dynamically adjusted based on the noise level of the electrolysis reaction, for example, 2-3 times the standard deviation of the current baseline is used as the discrimination threshold.

[0044] When generating reaction feature labels, semantic conversion is performed on the above-extracted voltage distribution, reaction polarity, and response density parameters. For example, if the voltage amplitude fluctuation range corresponding to a certain reaction parameter vector is between 30 - 40V (voltage distribution feature), the current direction is consistent with the voltage polarity (dominated by forward reactions), and the number of effective reaction events per unit time is 15 (high-density response), then a feature label of "high voltage fluctuation - forward-dominated - high-density response" is generated. The semantic structure of the label is composed of "attribute - value" pairs to ensure that each label can uniquely represent a class of reaction feature combinations, facilitating subsequent similarity calculation and clustering analysis.

[0045] After associating the reaction feature labels with the reaction parameter vectors, it is necessary to calculate the response similarity between the feature labels to screen the members of the clustering group. The similarity calculation method can adopt a metric based on semantic distance. For example, a semantic network of feature labels is constructed, and each label is decomposed into independent feature dimensions (such as voltage distribution, reaction polarity, response density), and each dimension corresponds to different value ranges (such as "high voltage fluctuation" corresponding to a voltage amplitude fluctuation range > 20V, "medium voltage fluctuation" corresponding to 10 - 20V, "low voltage fluctuation" corresponding to < 10V). For each dimension, if the values of two labels belong to the same interval or adjacent intervals, a higher similarity score (such as 0.8 - 1.0) is assigned; if they belong to non-adjacent intervals, the score is decreased according to the interval distance (such as a score of 0.5 for an interval apart, and a score of 0.2 for two or more intervals apart). Finally, the similarity scores of each dimension are weighted and averaged to obtain the comprehensive similarity value of the two labels. The weighting coefficients are preset according to the influence degree of each dimension on the reaction mechanism. For example, the weight of the reaction polarity dimension can be set to 0.4, and the weights of the voltage distribution and response density dimensions are each set to 0.3.

[0046] When screening parameter vectors with a similarity higher than the preset response threshold (such as 0.7) to form a reaction clustering group, a hierarchical clustering merging strategy is adopted. First, each parameter vector is regarded as an independent clustering unit, and the label similarity between all units is calculated. The unit pair with the highest similarity and exceeding the threshold is merged into a new cluster. Then, the similarity between the new cluster and other units (or clusters) is recalculated, and the merging process is repeated until there are no similarity pairs that meet the threshold condition. During this process, the similarity values and clustering member compositions of each merge need to be recorded to facilitate subsequent interpretability analysis of the clustering results.

[0047] To extract temporal attributes, to avoid high-frequency noise interference, the timestamp sequence can be filtered using a sliding window. The window width is set based on the pulse voltage cycle length. For example, when the pulse frequency is 500 Hz, the window width is set to 5 cycles (i.e., 10 ms). The smoothed temporal features are obtained by calculating the median or mean of the temporal attributes within the window. Current intensity is also processed using filtering algorithms, such as Gaussian filtering or Savitzky-Golay filtering, to eliminate the impact of sampling noise on feature extraction.

[0048] Determining reaction polarity requires considering the electrode configuration of the electrolysis reaction and the mechanism of carbon dioxide reduction. For example, in a typical carbon dioxide electrolysis cell, oxidation reactions occur at the anode (such as oxygen evolution from water) and reduction reactions occur at the cathode (conversion of carbon dioxide to hydrocarbons or carbon monoxide). Therefore, a current flowing from anode to cathode corresponds to a forward reaction (where the reduction reaction dominates), while a reverse current flow may indicate abnormal electrode polarization or the occurrence of a side reaction. By establishing a mapping between current direction and reaction polarity, abstract current data can be converted into physically meaningful reaction characteristics.

[0049] The calculation of the response density parameter requires defining a time interval constraint for valid reaction events to avoid splitting continuous fluctuations within the same reaction process into multiple independent events. For example, if the event interval threshold is set to 0.5ms, two current fluctuation events whose time interval is less than this threshold are considered to be continuations of the same event and are not counted repeatedly. This constraint can be adjusted based on the time constant of the electrolysis reaction to ensure that the count of valid reaction events truly reflects the independent chemical reaction steps.

[0050] When generating reaction feature labels, a unified label dictionary must be established to clearly define the value range and semantic definition of each feature dimension to avoid clustering bias caused by terminology ambiguity. For example, in the voltage distribution dimension, the classification of "high," "medium," and "low" must be linked to the voltage amplitude range of the pulse parameter interval. When the pulse voltage amplitude range is 0-50V, high voltage fluctuations can be defined as >35V, medium voltage fluctuations as 15-35V, and low voltage fluctuations as <15V, ensuring that the label definitions are consistent with the actual operating parameters of the system.

[0051] During the similarity calculation process, for parameter vectors with missing feature dimensions (e.g., missing data for a dimension due to a sensor failure), nearest neighbor interpolation or a feature correlation-based filling algorithm can be used to fill in the gaps using corresponding dimension data from other vectors in the same clustering group to ensure the integrity of the similarity calculation. If a vector is missing multiple key dimensions (e.g., missing both voltage distribution and reaction polarity), it will not be included in the clustering until the data is complete or marked as an anomaly.

[0052] After forming the reaction clustering groups, it is necessary to verify the rationality of the clustering results. The verification methods include: (1) Visual analysis: By using principal component analysis (PCA), the high-dimensional reaction parameter vector is reduced to two or three dimensions, and the distribution scatter plot of the clustering groups is drawn to observe whether each cluster forms an obvious cluster structure; (2) Statistical test: Using the ANOVA variance analysis method, test whether there are significant differences in the means of different clustering groups on each feature dimension (significance level α = 0.05). If the mean difference in a certain dimension is not significant, it is necessary to re-examine the feature extraction process of this dimension or adjust the clustering parameters.

[0053] Example 3: When generating the electrolysis feature vectors corresponding to the reaction units, it is necessary to dynamically analyze the current fluctuation characteristics of each reaction unit at its timing position within the pulse parameter interval and adopt a differentiated feature vector generation strategy. The specific implementation method is as follows: Determine the timing position of the reaction unit within the pulse parameter interval, which is jointly determined by the timestamp corresponding to the unit and the periodic characteristics of the pulse voltage sequence. For example, the pulse voltage sequence consists of continuous rectangular pulses, and each pulse period includes four stages: rising edge, flat top period, falling edge, and intermittent period. The reaction unit may be located in any stage. By synchronously collecting the trigger signal of the pulse voltage source module and the current data, the mapping relationship between the timing position of the reaction unit and the pulse stage is established, providing a time reference for subsequent current fluctuation analysis.

[0054] For each reaction unit, obtain its current fluctuation data within a preset window. The preset window is centered on the timestamp of the reaction unit and extends a certain duration forward and backward. The window length is determined according to the pulse period and the reaction time constant. For example, when the pulse frequency is 1000 Hz (period 1 ms), the preset window can be set to 0.5 ms forward and backward, with a total length of 1 ms, ensuring that the window covers the pulse stage corresponding to the unit and part of the adjacent stages to capture the complete dynamic process of the current response. The current fluctuation data is a sequence of current sampling values within this window, and the sampling frequency needs to meet the Nyquist criterion, usually more than twice the pulse frequency, to avoid frequency aliasing.

[0055] When calculating the current fluctuation coefficient of the unit, statistical methods are used to quantify the degree of current data fluctuation. The calculation formula of the current fluctuation coefficient is:

[0056] Where, is the current fluctuation coefficient, used to characterize the relative amplitude of the current fluctuation; is the number of sampling points within the preset window; is the current value at the th sampling point; is the average value of the current within the preset window. By calculating the coefficient of variation of the current data (the ratio of the standard deviation to the mean), this formula eliminates the influence of the current mean on the fluctuation amplitude, facilitating the horizontal comparison of the fluctuation degrees between different reaction units.

[0057] When the current fluctuation coefficient exceeds the first response threshold , the unit is determined to be a high-response unit. The first response threshold is preset according to the noise level of the electrolysis reaction and historical data. For example, it is taken as twice the coefficient of variation of the current during normal steady-state operation to ensure that the high-response units correspond to the stages of intense reaction or sudden state changes. For high-response units, their current data is directly extracted to form the electrolysis feature vector. The extracted current data includes but is not limited to: peak current , average current , slope of the current rising edge , slope of the falling edge , current duration , etc. These parameters directly reflect the kinetic characteristics of the electrolysis reaction within the unit. For example, the peak current corresponds to the utilization rate of the reaction active sites, and the slope of the rising edge reflects the reaction start-up speed.

[0058] When the current fluctuation coefficient is lower than the first response threshold , the unit is marked as a steady-state unit. The steady-state unit corresponds to the stable stage of the electrolysis reaction, with relatively small current fluctuations. However, the current data of adjacent units may contain gradually changing trend information or weak characteristic changes. At this time, the current data of the adjacent units before and after the steady-state unit are subjected to response superposition. The superposition method uses weighted summation, and the weight coefficient decreases according to the time interval from the current unit. For example, the exponential decay weight is used:

[0059] where is the weight coefficient of the th adjacent unit, is an integer, and ; is the time constant, which is set according to the pulse period and is usually taken as pulse periods. Through this weight assignment, the adjacent units closer to the current unit contribute more to the superposition result, reflecting the time locality characteristic of the current response.

[0060] The current data after response superposition needs to be reconstructed to generate the electrolysis feature vector. The reconstruction process includes: First, the current data of adjacent units are arranged in chronological order to form a length of sequence; then perform smoothing filtering on this sequence to eliminate the high-frequency noise that may be introduced during the superposition process; finally, extract the statistical features and trend features of the filtered data, such as the mean, variance, linear fitting slope, trend change points, etc., as the electrolysis feature vector of the steady-state unit. These features reflect the overall level and slow change trend of the current in the steady state stage, avoiding the loss of valuable information due to too small fluctuations in the data of a single steady-state unit.

[0061] When determining the preset window length, it is necessary to comprehensively consider the pulse width of the pulsed voltage and the generation rate of the reaction products. For example, for a pulse with a pulse width of 0.8 ms, the preset window length can be set to 1.2 ms to ensure that the window covers the entire pulse width and part of the intermittent period to capture the current response of the possible afterglow effect or product desorption process after the pulse ends. For a fast reaction system (such as a reaction involving a highly active catalyst), the window length can be appropriately shortened to focus on the transient response characteristics; for a slow reaction system, the window length can be extended to obtain more steady-state data.

[0062] The first response threshold needs to be set through statistical analysis of historical data. The specific steps are as follows: collect the current data of continuous operation under normal conditions, calculate the current fluctuation coefficient of all reaction units , and draw its probability density distribution curve. Determine the distribution range of during steady-state operation according to the curve. Usually, take the mean plus 2 times the standard deviation as so that the threshold can distinguish more than 95% of the steady-state units from the high-response units. After the threshold is set, it also needs to be dynamically adjusted according to the actual operating conditions. For example, when replacing the catalyst or adjusting the electrolyte composition, re-statistics the data to update the threshold to ensure the adaptability of the classification standard.

[0063] When extracting the current data of high-response units, attention should be paid to the processing of outliers. For sampling points that deviate significantly from the normal fluctuation range (such as current values exceeding ±3 times the standard deviation of the preset range), they are determined as outliers and corrected using interpolation methods (such as linear interpolation or cubic spline interpolation) to avoid the interference of outliers on the feature vector. The detection and correction of outliers are carried out in real time during the data acquisition stage to ensure the accuracy of subsequent analysis.

[0064] The number of adjacent units for response superposition is an adjustable parameter, and its value range is usually 1 - 5 units. When , only 1 adjacent unit before and after is superposed, which is suitable for scenarios where the current trend changes rapidly; when , 5 units before and after are superposed, which is suitable for scenarios where slow trend changes need to be captured. The value of The ability of the eigenvector to characterize the reaction state is selected to make the eigenvector with the highest classification accuracy. value.

[0065] The dimensionality of the electrolysis eigenvector varies depending on the type of reaction unit. For highly responsive units, the eigenvector may contain 5-8 dimensions (such as peak current, average current, and slope parameters). For steady-state units, after response superposition and reconstruction, the eigenvector may contain 3-5 dimensions (such as mean, variance, and trend slope). The dimensionality of the eigenvector must be kept within a reasonable range to avoid excessively high dimensionality, which can lead to increased computational complexity in subsequent parameter control modules, while ensuring sufficient information capacity to distinguish between different reaction states.

[0066] In the time series position analysis, if the reaction unit is located in the rising edge stage of the pulse sequence, its current fluctuation is usually more intense, and the proportion of high-response units is higher. At this time, it is necessary to focus on extracting parameters such as peak current and rising edge slope; if it is located in the plateau stage, the current tends to be stable and the proportion of steady-state units increases. It is necessary to capture the subtle change trend of adjacent units through response superposition; if it is located in the intermittent stage, the current may be close to the baseline level. At this time, the characteristic vector can reflect the recovery process of the electrode or the background current characteristics.

[0067] Example 4: In the parameter control module, the process of separating the positive reaction ratio, negative reaction ratio, and neutral response parameters from the electrolysis feature vector and generating voltage regulation rules requires combining the electrochemical mechanism of the electrolysis reaction with data feature analysis. The following detailed implementation is described using specific examples: Assume that the electrolysis characteristic vector for a reaction unit contains the following parameters: peak current of 2.5A, average current of 1.8A, rising slope of 0.5A / ms, falling slope of -0.3A / ms, and reaction duration of 4ms. Based on the correlation between current direction and pulse voltage polarity (assuming a positive voltage polarity and current flow from cathode to anode is defined as a forward reaction), the current direction of this unit is positive, and the forward reaction is determined to be dominant. Using preset thresholds (e.g., a forward reaction contribution of ≥60% is considered high, 30%-60% is considered medium, and <30% is considered low), combined with current parameters and reaction product analysis (e.g., a high CO generation rate detected by a gas sensor), the forward reaction contribution of this unit is determined to be 70%, the negative reaction contribution is 20%, and the neutral response parameter (e.g., side reaction or background current contribution) is 10%.

[0068] Based on the above parameters, the logic for generating voltage regulation rules is as follows: If the proportion of the forward reaction is higher than the threshold (for example, the preset threshold is 50%), and the proportion of the negative reaction is lower than the threshold (such as 30%), it is determined that the current voltage parameters are appropriate, and the existing pulse voltage sequence is maintained; If the proportion of the forward reaction is lower than the threshold and the proportion of the negative reaction is higher than the threshold, for example, the forward proportion is 25% and the negative proportion is 55%, a regulation rule of "increase the pulse amplitude by 10% and decrease the frequency by 20%" is generated to enhance the driving force of the forward reaction and extend the reaction time. If the neutral response parameter exceeds the preset threshold (such as 20%), the monitoring of the electrolyte composition or the electrode state is triggered. For example, impedance spectroscopy analysis is used to determine whether there is electrode passivation or electrolyte contamination, and then an auxiliary regulation rule (such as starting the electrolyte circulation or adjusting the electrode spacing) is generated.

[0069] When the number of reaction units covered by the current voltage regulation rule is less than the preset response threshold (for example, the total number of reaction units is 100, and the current rule only covers 15, which is lower than the preset 20), it is necessary to traverse the electrolysis feature vectors of adjacent reaction units to expand the rule. Taking the adjacent unit A as an example, its feature vector shows that the proportion of the forward reaction is 35%, the proportion of the negative reaction is 45%, the neutral response parameter is 20%, and it includes parameters not covered by the current rule such as the pulse rise time (2 ms) and the reaction product selectivity (CO selectivity 60%). Through feature difference analysis, it is found that the pulse rise time of this unit is relatively long, which may lead to insufficient adsorption of reactants and thus affect the proportion of the forward reaction. Therefore, the response index of "shorten the pulse rise time to within 1.5 ms" is added to the current rule to form a new combination of regulation rules: "increase the pulse amplitude by 10%, decrease the frequency by 20% and shorten the rise time by 30%".

[0070] In another scenario, if a reaction unit is in the intermittent period of the pulse sequence, its electrolysis feature vector shows that the current is close to the baseline level (0.1 A), the proportion of the forward reaction is 10%, the proportion of the negative reaction is 15%, and the neutral response parameter is 75%. At this time, the parameter control module identifies that this unit is in a low-activity state and generates a regulation rule of "superimpose small-amplitude pulses (amplitude 5 V, frequency 100 Hz) during the intermittent period" to activate the residual reactants on the electrode surface and reduce the proportion of the neutral response. If this rule only covers 5 units in the intermittent period (lower than the preset threshold of 10), then traverse the adjacent flat-top period units and find that there is a negative correlation between the average current and the pulse duty cycle in the flat-top period (the higher the duty cycle, the slower the increase in the average current). Then, the index of "dynamically adjust the duty cycle during the flat-top period (reduce the duty cycle by 5% every 10 ms)" is incorporated into the rule to form a cross-stage collaborative regulation strategy.

[0071] The generation of voltage regulation rules relies on a preset rule library, which is constructed based on the basic theory of electrolytic reactions and historical data. For example, the rule library contains empirical rules such as "when the pulse amplitude increases by 1V, the proportion of the forward reaction increases by about 2%-5% (in the range of 30-40V)", and statistical laws such as "when the frequency decreases by 100Hz, the proportion of the negative reaction decreases by about 3%-6%". The parameter range in the rule library needs to match the adjustable range of the pulse voltage source module. For example, the amplitude adjustment step is 1V and the frequency adjustment step is 50Hz to ensure the executability of the rules.

[0072] When traversing adjacent reaction units, the feature difference analysis uses the principal component analysis (PCA) method to reduce the high-dimensional electrolytic feature vector to 2-3 principal components. By observing the distribution distances of each unit in the principal component space, samples with significant differences from the units covered by the current rule are identified. For example, the units covered by the current rule are mainly distributed in the positive intervals of principal component 1 (reflecting current intensity) and principal component 2 (reflecting time characteristics), while adjacent unit A is in the negative interval of principal component 2, indicating that its time characteristics (such as the pulse rise time) are significantly different from the characteristics concerned by the current rule, and time-related indicators need to be incorporated into the rule.

[0073] For an electrolytic system with a multi-electrode array, the reaction units at different electrode sites may have different characteristic parameters. For example, due to the higher mass transfer efficiency, the proportion of the forward reaction in the reaction units in the edge region of the electrode is generally higher than that in the central region. The parameter regulation module needs to generate differentiated rules for the units in different regions. For example, a lower pulse amplitude is used for the units in the edge region (to avoid side reactions caused by overpotential), and a higher frequency is used for the units in the central region (to enhance mass transfer). At this time, if the number of units covered by the rules in a certain region is insufficient, adjacent units in the same region should be traversed first to avoid the decline in the generalization ability of the rules caused by the mixing of cross-region features.

[0074] The update frequency of the voltage regulation rules is related to the time resolution of the reaction units. When the division granularity of the reaction units is 10ms, the rule update period can be set to conduct a global evaluation every 100ms (i.e., 10 units) to ensure that the regulation actions can respond to reaction changes in a timely manner and avoid system instability caused by frequent adjustments. When updating the rules each time, first count the number of units covered by each current rule. If it is below the threshold, trigger the traversal mechanism; otherwise, maintain the existing rules.

[0075] When generating pulse correction parameters, the response delay of the hardware module needs to be considered. For example, the amplitude adjustment of the pulse voltage source module requires a stabilization time of 0.5 ms. Therefore, the amplitude adjustment amount in the rule needs to compensate for this delay in time to ensure that the correction parameters take effect on the target reaction unit. The value range of the correction parameters also needs to be limited by the hardware capabilities. For example, the amplitude cannot exceed the upper limit of 50 V, and the frequency cannot be lower than the lower limit of 10 Hz, to avoid issuing ineffective or device-damaging instructions.

[0076] The analysis of the neutral response parameters can also be combined with thermodynamic calculations. For example, by calculating the difference (overpotential) between the theoretical decomposition voltage and the actual applied voltage under the current pulse parameters, it can be judged whether the neutral response is caused by insufficient energy. If the overpotential is less than 80% of the theoretical value, a rule of "increasing the amplitude to increase the overpotential" is generated; if the overpotential is sufficient but the neutral response is still high, it may point to a kinetic barrier (such as insufficient catalyst active sites), and other modules (such as the electrode pretreatment module) need to be coordinated for adjustment.

[0077] Example 5: In the timing optimization module, the process of identifying the time correlation pattern in the pulse correction parameters and calculating the voltage offset needs to combine the time characteristics of the pulse sequence and the dynamic correlation of parameter adjustment. The following details the implementation method through specific examples: Suppose the pulse correction parameter sequence in a certain stage contains the following adjustment records: at 10 ms, the pulse amplitude is adjusted by +5% and the frequency is adjusted by -10%; at 20 ms, the amplitude is adjusted by +3% and the frequency is adjusted by -5%; at 30 ms, the amplitude is adjusted by +2% and the frequency is adjusted by -3%. By analyzing the time intervals and parameter change amplitudes of these records, sequence factors of the time switching frequency (such as adjusting once every 10 ms) and change factors of the response jump amplitude (such as the amplitude adjustment amplitude decreasing from 5% to 2% and the frequency adjustment amplitude decreasing from 10% to 3%) can be extracted. The sequence factor reflects the time density of parameter adjustment, and the change factor reflects the amplitude difference between adjacent adjustments. The two together constitute the core characteristics of the time correlation pattern.

[0078] When constructing a weight assignment matrix associated with sequence factors and variation factors, first identify the time - period characteristics of the sequence factors. If the time intervals of the above - mentioned adjustment records are all 10 ms, which exactly match the preset pulse period (10 ms), it is determined that the sequence factor has strict periodicity, and it is set as the starting index of the weight assignment matrix. The starting index corresponds to the first row or the first column of the matrix, which is used to anchor the benchmark of the time dimension. Then calculate the correlation matching degree between the sequence factor and the variation factor. For example, by observing whether the amplitude adjustment amplitude decreases as the time interval lengthens, determine whether there is a negative correlation between the two. If it is statistically found that for every 5 - ms increase in the time interval, the amplitude adjustment amplitude decreases by an average of 2%, the correlation matching degree can be defined as a relatively high level (such as a matching - degree score of 0.8), and the middle index and the ending index of the matrix are generated in descending order of the matching degree. The middle index corresponds to the sequence factors at different time intervals, and the ending index corresponds to the variation - factor amplitudes at each time interval.

[0079] When performing path backtracking on the ending index, it is necessary to check whether its matching degree meets the preset threshold (such as 0.5). For example, for an ending index corresponding to a time interval of 15 ms and an amplitude adjustment amplitude of 1%, after analysis, it is found that this adjustment has no obvious correlation pattern with the previous and subsequent adjustments, and the calculated matching degree is 0.4, which is lower than the threshold. Then it is output as an independent item in the final distribution of the weight assignment matrix to avoid errors caused by forced association. The structure of the weight assignment matrix can be designed as a two - dimensional table, where the rows represent the sequence factors (time intervals), the columns represent the variation factors (adjustment amplitudes), and the cell values are the probabilities or weights of the occurrence of this time - amplitude combination. For example, at a 10 - ms interval, the probability of a 5% amplitude adjustment is 0.6, the probability of a 3% adjustment is 0.3, and the probability of a 2% adjustment is 0.1.

[0080] When calculating the voltage offset, first calculate the mean of the sequence factors and the range of the variation factors for each termination index in the weight distribution matrix. For example, for a time interval sequence corresponding to a certain termination index of 10 ms, 10 ms, 20 ms, the mean is 13.3 ms; the variation factors (amplitude adjustment ranges) are 5%, 3%, 2%, and the range is 3%. At the same time, calculate the global variance of all index factors to reflect the overall fluctuation level of the time interval and the adjustment amplitude. Subtract the mean of the sequence factors of a single termination index from the mean of the adjacent index to obtain the time series offset (for example, the difference between 13.3 ms and the mean of the adjacent index of 10 ms is 3.3 ms), and divide it by the global variance to obtain the time series offset coefficient. This coefficient is dimensionless and is used to standardize the impact of time differences. At the same time, calculate the ratio of the range of the variation factors to the global variance to reflect the relative change degree of the adjustment amplitude. Weighted sum the time series offset coefficient and the amplitude change ratio according to the preset weights (such as a time weight of 0.6 and an amplitude weight of 0.4) to obtain the voltage offset of this index. For example, if the time series offset coefficient is 0.5 and the amplitude change ratio is 0.3, the weighted offset is 0.5×0.6 + 0.3×0.4 = 0.42, corresponding to a voltage adjustment amplitude of +0.42% (or -0.42%, the sign is determined by the offset direction).

[0081] In another scenario, if the time correlation pattern of the pulse correction parameter shows non-periodic characteristics, such as the adjustment intervals are 8 ms, 12 ms, 9 ms in sequence and the sequence factors have no obvious periodic characteristics, then it is not set as the starting index, but instead a matrix is constructed mainly based on the amplitude characteristics of the variation factors. For example, records with an amplitude adjustment amplitude ≥ 4% are classified as the "high amplitude" category, 2% - 4% are classified as the "medium amplitude" category, and < 2% are classified as the "low amplitude" category, and the time interval distributions under each category are statistically analyzed respectively. If the time intervals in the "high amplitude" category are concentrated in 8 - 10 ms, the "medium amplitude" category is concentrated in 10 - 12 ms, and the "low amplitude" category is concentrated in 12 - 14 ms, then the weight distribution matrix has the amplitude categories as rows, the time interval ranges as columns, and the cell values are the occurrence frequencies of the time intervals under that category. For example, the frequency of "high amplitude - 8 - 10 ms" is 0.7, reflecting that high amplitude adjustments are more likely to occur within a shorter time interval.

[0082] When identifying time periodic characteristics, the autocorrelation analysis method can be used to calculate the autocorrelation function of the sequence factors and detect whether there are significant peak points (such as the autocorrelation coefficient is 0.9 when lagging 10 ms) to judge whether periodicity exists. If periodicity exists, further verify whether the cycle length is consistent with the basic cycle of the pulse voltage source module (such as the cycle corresponding to the pulse frequency set by the user) to ensure that the time correlation pattern is synchronized with the inherent rhythm of the system.

[0083] The construction of the weight distribution matrix also needs to consider the stage characteristics of the electrolysis reaction. For example, in the reaction startup stage (the first 100 ms), the adjustment frequency of the pulse correction parameter is relatively high (the sequence factor is small), and the change amplitude is relatively large (the change factor is high). The first few rows in the weight matrix can preferentially assign high weights to combinations with short intervals and high amplitudes. In the stable operation stage, the adjustment frequency decreases and the amplitude decreases, and the weight matrix then turns to combinations with long intervals and low amplitudes. By constructing the matrix in stages, the timing optimization can better fit the evolution process of the reaction kinetics.

[0084] During the path backtracking process, if it is found that the matching degree of a certain termination index is lower than the threshold, in addition to outputting it as an independent item, it can also be marked as "abnormal adjustment" to trigger the backtracking analysis of the reaction unit corresponding to this adjustment. For example, check whether there are abnormal values in the electrolysis feature vector of this unit (such as a sudden increase in the current fluctuation coefficient), and judge whether the adjustment is caused by noise data. If so, delete this record to avoid abnormal data contaminating the weight matrix.

[0085] The calculation result of the voltage offset needs to be mapped to the adjustable parameter range of the pulse voltage source module. For example, if the calculated offset is +0.42% and the amplitude adjustment step of the module is 1%, then the result needs to be rounded. The actual adjustment amplitude is +1%, and the adjustment error is recorded for subsequent optimization. If the offset is negative, such as -0.35%, the amplitude is correspondingly reduced by 0% (due to the step limit), but the error needs to be accumulated until the next adjustment to ensure the continuous regulation ability of the system.

[0086] The identification of the time correlation pattern can also be combined with domain knowledge, such as the mass transfer lag effect in carbon dioxide electrolysis. If it takes a certain time (such as 5 ms) for the generation of reaction products to be reflected in the current data after the pulse parameters are adjusted, then when analyzing the time correlation pattern, a time delay factor needs to be introduced to associate the adjustment record with the characteristics of the reaction unit 5 ms later, so as to avoid misjudgment of the pattern caused by time sequence misalignment.

[0087] It should be noted that in this article, relational terms such as first and second are only used to distinguish one entity or operation from another entity or operation, and do not necessarily require or imply any actual relationship or order between these entities or operations. Moreover, the terms "include", "comprise" or any other variant thereof are intended to cover non-exclusive inclusion, so that a process, method, article or device including a series of elements not only includes those elements, but also includes other elements not expressly listed, or also includes elements inherent to such process, method, article or device.

[0088] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A carbon dioxide electrolysis system based on pulsed voltage regulation, characterized in that, Including: A pulse voltage source module for generating a controllable pulse voltage sequence and setting a pulse parameter range corresponding to the electrolysis reaction; An electrolysis reaction module for dividing a plurality of reaction units within the pulse parameter range, analyzing the current response of the electrolysis process of each reaction unit, and generating an electrolysis feature vector corresponding to the reaction unit; A parameter regulation module for extracting key reaction indicators from the electrolysis feature vector, establishing a voltage regulation rule associated with the reaction unit, and obtaining a pulse correction parameter corresponding to the rule; A timing optimization module for identifying the time correlation pattern in the pulse correction parameter, dynamically calibrating the key reaction indicators according to the correlation pattern, and calculating the voltage offset of each reaction unit under different calibration strategies; A reference setting module for deriving an optimal pulse threshold based on the voltage offset and generating an electrolysis deviation sequence by comparing the current electrolysis feature value with the optimal pulse threshold; A decision execution module for analyzing the electrolysis deviation sequence and integrating the electrolysis deviation sequence into an electrolysis regulation execution plan based on the voltage offset trend of the reaction unit.

2. The carbon dioxide electrolysis system based on pulse voltage regulation according to claim 1, characterized in that, The implementation method of the electrolysis reaction module includes: constructing an electrolysis feature library corresponding to the reaction unit, and the electrolysis feature library contains current data and a reaction parameter vector mapped by the electrolysis response; Performing similar response matching on the reaction parameter vector, dividing the reaction clustering group of the reaction parameter vector according to the matching result; extracting the electrolysis focus point of the current data from the reaction clustering group and setting the focus point as the reaction unit.

3. The carbon dioxide electrolysis system based on pulse voltage regulation according to claim 2, wherein, Dividing the reaction clustering group of the reaction parameter vector further includes: Extracting voltage distribution, reaction polarity, and response density parameters according to the time attribute and current intensity in the reaction parameter vector, and generating a reaction feature label based on the above parameters; Associating the reaction feature label with the reaction parameter vector, and screening the parameter vectors with a response similarity higher than the preset response threshold to form a reaction clustering group by calculating the response similarity between the feature labels.

4. The carbon dioxide electrolysis system based on pulse voltage regulation according to claim 1, characterized in that, The implementation method of generating the electrolysis feature vector corresponding to the reaction unit includes: For each reaction unit, obtaining the current fluctuation data of the unit within a preset window according to the timing position of the unit in the pulse parameter range, and calculating the current fluctuation coefficient of the unit; When the current fluctuation coefficient exceeds the first response threshold, marking the unit as a high-response unit and extracting its current data to form an electrolysis feature vector; when the current fluctuation coefficient is lower than the first response threshold, marking the unit as a steady-state unit and performing response superposition on the current data of the adjacent unit of the unit, and reconstructing the superimposed data into an electrolysis feature vector.

5. The carbon dioxide electrolysis system based on pulse voltage regulation according to claim 1, characterized in that, The implementation method of the parameter regulation module includes: Separating the positive reaction ratio, negative reaction ratio, and neutral response parameters from the electrolysis feature vector, and generating a voltage regulation rule for the reaction unit based on the above parameters; If the number of reaction units covered by the current voltage regulation rule is less than the preset response threshold, traversing the electrolysis feature vectors of adjacent reaction units, and adding the response indicators not included in the regulation rules of the adjacent units to the current rule.

6. The carbon dioxide electrolysis system based on pulse voltage regulation according to claim 1, wherein, The implementation method of the timing optimization module includes: obtaining the sequence factor of the time switching frequency and the change factor of the response jump amplitude in the time correlation pattern; Construct a weight assignment matrix associated with sequence factors and variation factors, and determine the voltage offset under different calibration strategies according to the distribution probability of each element in the matrix.

7. The carbon dioxide electrolysis system based on pulse voltage regulation according to claim 6, characterized in that, Constructing the weight assignment matrix further includes: Identifying the time period characteristics of the sequence factor. If the current period characteristics exactly match the preset time period, set the sequence factor as the starting index of the weight assignment matrix; Calculating the correlation matching degree between the sequence factor and the variation factor, and generating the middle index and the ending index of the weight assignment matrix in descending order of the matching degree; Performing path backtracking on the ending index. When the matching degree of the ending index is lower than the preset matching threshold, output it as the final distribution of the weight assignment matrix.

8. The carbon dioxide electrolysis system based on pulse voltage regulation according to claim 7, wherein, The implementation method of calculating the voltage offset includes: Statistical mean of the sequence factors and the range of variation factors for each ending index in the weight assignment matrix, and calculate the global variance of all index factors; Subtract the mean of the sequence factors of a single ending index from the mean of the sequence factors of the adjacent index, and divide by the global variance to obtain the time series offset coefficient; at the same time, calculate the ratio of the range of variation factors to the global variance, and weight and sum the two as the voltage offset of this index.

9. The carbon dioxide electrolysis system based on pulse voltage regulation according to claim 1, wherein The implementation method of deriving the optimal pulse threshold includes: Extract the reaction mode in the historical data that is closest to the current voltage offset, and calculate the cosine similarity in the time distribution between the two as the first dynamic reference value; Statistical difference in the number of response peaks between the current voltage offset and the historical reaction mode, and use the difference number as the second dynamic reference value; Based on the linear combination of the first dynamic reference value and the second dynamic reference value, match the optimal pulse threshold in the preset pulse threshold table.

10. The carbon dioxide electrolysis system based on pulse voltage regulation according to claim 1, wherein The implementation method of the decision execution module includes: dividing the positive offset interval and the negative offset interval according to the voltage offset direction of each unit in the electrolysis deviation sequence; Extract the convergence rate of the electrolysis deviation in the positive offset interval and the diffusion rate of the electrolysis deviation in the negative offset interval, and perform adversarial fusion on the two according to the time weight of the reaction unit to generate the adjustment parameter of the electrolysis regulation execution plan.

Images