Electrolysis cooperative control method and system based on electrochemical state feedback

Abstract

The invention discloses an electrolysis cooperative control method and system based on electrochemical state feedback, and relates to the technical field related to electrolysis control, and the method comprises the steps: monitoring a steady-state operation signal of an electrolysis unit, applying controlled perturbation excitation, synchronously collecting an unsteady-state response signal, and carrying out distortion analysis, transient impedance spectrum extraction and phase drift analysis on the unsteady-state response signal, so as to obtain an unsteady-state response signal; obtaining a dynamic feature set; in combination with the steady-state reference features, the combined features are mapped to a unified projection space for electrochemical state inversion, and a state feedback quantity is generated; and extracting state evolution characteristics and judging an evolution interval, and generating a cooperative control strategy according to the evolution interval to execute electrolysis cooperative control management. The technical problems that in the prior art, electrochemical state sensing is lagged and one-sided, dynamic information utilization is lacked, and multi-unit cooperative control is extensive are solved, and the technical effects that the state sensing real-time performance of the electrolysis process, cooperative control accuracy and foresight and the overall energy efficiency and service life of cooperative operation of multiple electrolysis units are improved are achieved.

Description

Technical Field

[0001] This invention relates to the field of electrolysis control technology, specifically to an electrolysis collaborative control method and system based on electrochemical state feedback. Background Technology

[0002] Electrolysis is a key link in energy conversion and core product production. It is usually composed of a large number of parallel or series electrolysis units (such as electrolyzers) operating in coordination. Its overall efficiency, safety, energy consumption and product purity are directly related to the economy and stability of the entire production system. However, the actual electrochemical states within the electrolysis process, such as the active surface state of the electrodes, electrolyte composition distribution, interfacial mass transfer characteristics, degree of side reactions, and aging or contamination trends, cannot be directly and accurately obtained through conventional macroscopic operating parameters such as voltage, current, temperature, and pressure. Traditional electrolysis control relies on the setting and adjustment of these macroscopic steady-state operating parameters for feedback adjustment, which has many limitations. It is not sensitive to early and subtle changes in key states such as electrode performance degradation, electrolyte composition shifts, or deterioration of local reaction conditions, causing problems to accumulate until they affect yield or safety before they are discovered, resulting in a lag in control response. The dynamic response characteristics of the electrolysis unit under operating condition adjustments, load fluctuations, or disturbances contain rich state information different from steady state information. Existing steady-state control ignores this, resulting in the loss of key feature dimensions that can be used for early diagnosis and fine-tuning. At the same time, in scenarios where multiple electrolysis units operate in tandem, current control strategies cannot achieve "tailored measures for each unit," resulting in suboptimal overall system energy efficiency and accelerating the deterioration of poorly performing units, creating a bottleneck effect and affecting overall operational life and safety.

[0003] Therefore, current related technologies suffer from technical problems such as delayed and incomplete electrochemical state perception, lack of dynamic information utilization, and crude multi-unit collaborative control. Summary of the Invention

[0004] This application provides an electrolysis collaborative control method and system based on electrochemical state feedback, which solves the technical problems of lagging and one-sided electrochemical state perception, lack of dynamic information utilization, and crude multi-unit collaborative control in the prior art. It achieves the technical effect of improving the real-time state perception of the electrolysis process, the accuracy and foresight of collaborative control, and the overall energy efficiency and lifespan of multi-electrolysis unit collaborative operation.

[0005] This application provides an electrolysis collaborative control method based on electrochemical state feedback. The method includes: during the operation of the electrolyte equipment, continuously acquiring steady-state operating signals characterizing the long-term operating conditions of each electrolysis unit; configuring oxygen production constraints based on the steady-state operating signals; applying controlled amplitude and timing perturbation excitations to each electrolysis unit under the constraints; and simultaneously acquiring non-steady-state operating response signals caused by the perturbation excitations; performing response distortion analysis, transient impedance spectrum feature extraction, and non-steady-state waveform phase drift analysis on the non-steady-state operating response signals to obtain a dynamic feature set; extracting steady-state baseline operating features based on the steady-state operating signals, jointly fusing them with the dynamic feature set, and mapping them to a unified feature projection space; performing electrochemical state inversion processing within the feature projection space to generate an electrochemical state feedback quantity; extracting state evolution features characterizing the direction, magnitude, and cumulative trend of electrochemical state changes based on the electrochemical state feedback quantity, and determining the evolution interval of the electrochemical state; and generating a collaborative control strategy driven by the state evolution features based on the evolution interval, and executing electrolysis collaborative control management.

[0006] In a possible implementation, a collaborative control strategy driven by state evolution characteristics is generated based on the evolution interval, including: calculating the state offset of each electrolytic unit according to the state evolution characteristics, wherein the state offset includes the change amplitude, change rate, and accumulation trend; generating individual control gains using a weighted mapping method based on the distribution data of the evolution interval and the state offset, wherein the weighted mapping includes: a) assigning a first enhancement weight to electrolytic units in the restricted enhancement region to enhance positive regulation; b) assigning a second maintenance weight to electrolytic units in the stabilization region to maintain fine-tuning; c) assigning a third inhibitory weight to electrolytic units in the instability accumulation region to mitigate abnormal accumulation; fusing the individual control gains with the unsteady-state operation response signal and the steady-state reference operation characteristics to calculate the control adjustment amount of each electrolytic unit, wherein the control adjustment amount includes the adjustment amplitude, adjustment time, and action sequence; and performing global collaborative analysis based on the control adjustment amount of each electrolytic unit and the state evolution distribution of the group to establish a collaborative control strategy.

[0007] In a possible implementation, the collaborative control strategy driven by the state evolution characteristics generated based on the evolution interval further includes: for the state offset of each electrolysis unit, using a dynamic time window calculation method, generating a comprehensive state index based on short-term perturbation response and long-term steady-state benchmark operating characteristics, and adaptively adjusting the dynamic threshold by calling historical cumulative trends and perturbation response sensitivity; comparing the comprehensive state index with the dynamic threshold to determine the state category of the electrolysis unit, and then calculating the individual control gain using a nonlinear mapping function; and establishing control adjustment quantities based on the individual control gain, the unsteady-state operating response signal, and the steady-state benchmark to generate the collaborative control strategy.

[0008] In a possible implementation, response distortion analysis, transient impedance spectrum feature extraction, and non-steady-state waveform phase drift analysis are performed on the non-steady-state operating response signal to obtain a dynamic feature set. This includes: dividing the non-steady-state operating response signal caused by perturbation excitation into an excitation initial segment, an excitation transition segment, and an excitation decay segment according to the excitation timing; within each timing segment, performing distortion degree calculation based on a reference steady state on the voltage and current response waveforms to obtain response distortion features characterizing the degree of nonlinear polarization enhancement and mass transfer limitation; performing a short-time Fourier transform on the non-steady-state operating response signal to extract transient impedance spectrum features, which include the change in equivalent charge transfer impedance and the slope of diffusion impedance evolution; tracking the phase change of the non-steady-state operating response signal over time, calculating the phase drift rate and phase hysteresis, and constructing phase drift features; and outputting the response distortion features, transient impedance spectrum features, and phase drift features as a dynamic feature set.

[0009] In a possible implementation, after extracting steady-state reference operating features based on the steady-state operating signal, these features are jointly fused with a dynamic feature set and mapped to a unified feature projection space. Electrochemical state inversion processing is then performed within this feature projection space, including: performing time aggregation processing on the steady-state operating signal under a preset steady-state criterion to extract steady-state reference operating features characterizing the steady-state polarization level, energy conversion efficiency, and long-term load characteristics of the electrolysis unit; using these steady-state reference operating features as a reference, performing normalization mapping and relative offset calculation on the response distortion features, transient impedance spectrum features, and phase drift features in the dynamic feature set to construct a steady-state-non-steady-state correlated feature vector; performing feature weighted fusion dimension alignment processing on the steady-state-non-steady-state correlated feature vector and projecting it to a unified feature projection space; and performing electrochemical state inversion operation within this feature projection space based on the steady-state-non-steady-state correlated feature vector to obtain an electrochemical state feedback quantity characterizing the current electrochemical reaction activity, mass transfer limitation, and aging evolution level of the electrolysis unit.

[0010] In a possible implementation, configuring the influence constraints of the oxygen production condition based on the steady-state operating signal includes: determining the steady-state current density of the electrolysis unit under the oxygen production condition based on the steady-state operating signal, and using the steady-state current density as a reference operating point; constructing a current disturbance allowable range with the reference operating point as the center according to a preset relative offset ratio, the current disturbance allowable range including an upper offset boundary and a lower offset boundary; and outputting the current disturbance allowable range as an influence constraint.

[0011] In possible implementations, the electrolysis collaborative control management also includes: taking the starting point of the electrolysis collaborative control as the time zero point, constructing a time evolution sequence of the electrochemical state of the electrolyte device under collaborative control; configuring a collaborative feedback signal according to the time evolution sequence, and performing self-optimization management of the collaborative control strategy.

[0012] This application also provides an electrolysis collaborative control system based on electrochemical state feedback. The system includes: an operation signal acquisition module, used to continuously acquire steady-state operation signals characterizing the long-term operating conditions of each electrolysis unit during the operation of the electrolyte equipment; after configuring the influence constraints of oxygen production conditions based on the steady-state operation signals, apply controlled amplitude and timing perturbation excitations to each electrolysis unit under the influence constraints; and synchronously acquire non-steady-state operation response signals caused by the perturbation excitations. A dynamic feature acquisition module is used to perform response distortion analysis, transient impedance spectrum feature extraction, and non-steady-state waveform phase drift analysis on the non-steady-state operation response signals to acquire dynamic features. The system comprises a feature set and a state feedback quantity generation module, which extracts steady-state baseline operating features based on steady-state operating signals, fuses them with the dynamic feature set, and maps them to a unified feature projection space. Within this feature projection space, electrochemical state inversion processing is performed to generate electrochemical state feedback quantities. A state evolution feature extraction module is used to extract state evolution features characterizing the direction, magnitude, and cumulative trend of electrochemical state changes based on the electrochemical state feedback quantities, and to determine the evolution interval of the electrochemical state. A collaborative control management module is used to generate a collaborative control strategy driven by the state evolution features based on the evolution interval, and to execute electrolytic collaborative control management.

[0013] The proposed electrolysis collaborative control method and system based on electrochemical state feedback involves monitoring the steady-state operating signal of the electrolysis unit and applying controlled perturbation excitation, while simultaneously acquiring the unsteady-state response signal. Distortion analysis, transient impedance spectroscopy extraction, and phase drift analysis are performed on this signal to obtain a dynamic feature set. Combined with steady-state baseline features, the joint features are mapped to a unified projection space for electrochemical state inversion, generating state feedback quantities. State evolution features are extracted, and evolution intervals are determined. Based on these intervals, a collaborative control strategy is generated to implement electrolysis collaborative control management. This approach solves the technical problems of lagging and incomplete electrochemical state perception, lack of dynamic information utilization, and crude multi-unit collaborative control in existing technologies. It achieves the technical effects of improving the real-time performance of state perception in the electrolysis process, the accuracy and foresight of collaborative control, and the overall energy efficiency and lifespan of multi-electrolysis unit collaborative operation. Attached Figure Description

[0014] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings of the embodiments of this disclosure will be briefly described below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0015] Figure 1 This is a schematic diagram of the electrolysis collaborative control method based on electrochemical state feedback provided in an embodiment of this application.

[0016] Figure 2 A schematic diagram of the electrolysis collaborative control system based on electrochemical state feedback provided in this application embodiment.

[0017] Explanation of reference numerals in the attached diagram: 10 for operation signal acquisition module, 20 for dynamic feature acquisition module, 30 for state feedback quantity generation module, 40 for state evolution feature extraction module, and 50 for collaborative control and management module. Detailed Implementation

[0018] To further illustrate the technical means and effects adopted by the present invention in order to achieve the intended purpose, the following detailed description is provided in conjunction with the accompanying drawings and preferred embodiments, based on the specific implementation methods, structures, features and effects of the present invention.

[0019] This application provides an electrolysis collaborative control method based on electrochemical state feedback, such as... Figure 1 As shown, the method includes:

[0020] Step S100: During the operation of the electrolyte equipment, for each electrolysis unit, a steady-state operating signal characterizing the long-term operating condition of the electrolysis unit is continuously collected. After configuring the influence constraint of oxygen production condition according to the steady-state operating signal, a micro-perturbation excitation with controlled amplitude and timing is applied to each electrolysis unit under the influence constraint, and the non-steady-state operating response signal caused by the micro-perturbation excitation is collected simultaneously.

[0021] Preferably, during the normal continuous production period of each electrolysis unit in the electrolyte equipment operation, its operating parameters are continuously measured and recorded at a fixed sampling frequency (e.g., once per second or once per minute). Specifically, the sampling time span is much longer than the process disturbance cycle, usually several hours to several days. Within the sampling period, the parameter fluctuation amplitude is less than the preset threshold, such as voltage fluctuation < ±1%, and directly reflects the main operating conditions of the electrolysis unit, including electrical parameters such as DC working voltage, working current, and current density; process parameters such as electrolyte inlet / outlet temperature, electrolyte circulation flow rate, and electrolyte concentration; output parameters such as gas production per unit time and gas purity; and auxiliary parameters such as separator pressure difference and tank temperature distribution, which constitute a steady-state operating signal characterizing the long-term operating conditions of the electrolysis unit.

[0022] Preferably, during the operation of the water electrolysis hydrogen production equipment, based on the steady-state current density and electrolyte temperature of the currently collected steady-state operating signal, a reference operating point corresponding to the current oxygen production rate is calculated, and perturbation limits are set to ensure that applying perturbations does not trigger safety interlocks or affect product quality. This may include amplitude constraints, i.e., the amplitude of the perturbation current must not exceed ±X% (e.g., ±5%) of the steady-state current, and this percentage is determined based on the design margin of the electrolyzer and the capacity of the gas processing system; timing constraints, i.e., the duration of the perturbation must not exceed Y seconds (e.g., 10 seconds), and the interval between two perturbations must not be less than Z minutes (e.g., 30 minutes) to avoid perturbation accumulation; and boundary constraints, i.e., the cell voltage must not exceed the design upper limit during the perturbation period, and the electrolyte temperature fluctuation must not exceed ±ΔT℃, which together constitute the influence constraints of the oxygen production operating condition.

[0023] Preferably, under the influence constraint, a controlled amplitude and timing perturbation excitation is applied to each electrolysis unit. Specifically, the control system issues a command to superimpose a square wave, sine wave, or pseudo-random sequence current perturbation signal onto the steady-state current reference of the rectifier or power regulator. At the same time as the perturbation is applied, a high-speed data acquisition system is started, with a sampling frequency of 100Hz-10kHz, to synchronously acquire the actual output current perturbation waveform, as well as the transient change waveform of the voltage across the electrolysis unit, the change in the electrolyte temperature sensor reading, the change in the gas pressure sensor reading, etc., and finally determine the non-steady-state operation response signal caused by the perturbation excitation.

[0024] Furthermore, step S100 also includes determining the steady-state current density of the electrolysis unit under oxygen production conditions based on the steady-state operating signal, and using the steady-state current density as a reference operating point; constructing a current disturbance allowable range with the reference operating point as the center according to a preset relative offset ratio, the current disturbance allowable range including an upper offset boundary and a lower offset boundary; and outputting the current disturbance allowable range as an influence constraint.

[0025] Preferably, during the operation of the electrolysis equipment, boundary conditions for perturbation excitation testing are safely set. Specifically, the current signal and the effective electrode area of ​​the electrolysis unit are selected from the steady-state operating signal to calculate and determine the steady-state current density of the electrolysis unit under oxygen production conditions. This current density value fluctuates within a very narrow range, representing the current normal operating intensity of the electrolysis equipment. The steady-state current density is used as the reference operating point, i.e., the reference zero point for perturbation excitation. Then, with the reference operating point as the center, a current perturbation allowable range is constructed according to a preset relative offset ratio. The preset relative offset ratio is a parameter pre-set according to engineering safety criteria, such as ±2%, ±5%, etc. The upper offset boundary of the allowable current disturbance range is the steady-state current density value of the reference operating point × (1 + preset upper offset ratio), and the lower offset boundary is the steady-state current density value of the reference operating point × (1 + preset lower offset ratio). Then, a safe operating window is set, and the allowable current disturbance range is output as an influence constraint to the controller responsible for generating the perturbation excitation. When generating the perturbation current waveform, the controller must ensure that its instantaneous value, after being converted into the current density corresponding to the perturbation test current, is within the allowable current disturbance range throughout the entire process. For example, if a sinusoidal perturbation is applied, the entire sine wave must be limited within these upper and lower boundaries to ensure the continuous and stable operation of the main process.

[0026] Step S200: Perform response distortion analysis, transient impedance spectrum feature extraction, and unsteady waveform phase drift analysis on the unsteady operation response signal to obtain a dynamic feature set.

[0027] Step S200 further includes dividing the unsteady-state operating response signal caused by perturbation excitation into an excitation initial segment, an excitation transition segment, and an excitation decay segment according to the excitation timing; within each timing segment, performing distortion calculations based on a reference steady state on the voltage and current response waveforms to obtain response distortion features characterizing the degree of nonlinear polarization enhancement and mass transfer limitation; performing a short-time Fourier transform on the unsteady-state operating response signal to extract transient impedance spectrum features, which include the change in equivalent charge transfer impedance and the slope of diffusion impedance evolution; tracking the phase change of the unsteady-state operating response signal over time, calculating the phase drift rate and phase hysteresis, and constructing phase drift features; and outputting the response distortion features, transient impedance spectrum features, and phase drift features as a dynamic feature set.

[0028] Preferably, based on the start and end times of the perturbation excitation signal, the resulting high-frequency voltage and current response data streams and other unsteady-state operation response signals are divided into an excitation initial segment, an excitation transition segment, and an excitation decay segment along the excitation action time sequence. This determines the different characteristics of the system at different dynamic stages, thereby enabling more refined feature extraction. Specifically, the excitation initial segment is the period from the start of excitation application to the first peak or significant turning point of the response, reflecting the initial response speed and inertia of the system; the excitation transition segment is the period when the excitation response fluctuates near the peak or turning point and tends to a certain metastable state, reflecting the damping characteristics and internal adjustment process of the system; and the excitation decay segment is the period from the start of excitation removal to the system response recovering to near the original steady state, reflecting the system's recovery capability and memory effect.

[0029] Preferably, response distortion analysis is performed on the non-steady-state operation response signal. Specifically, within each time segment, the measured actual voltage / current transient waveform is compared with the theoretical reference waveform under ideal system conditions. The degree of deviation of the actual waveform from the theoretical waveform is calculated, such as quantifying the waveform distortion rate, harmonic content, or deviation amplitude at a specific time point. The response distortion characteristics characterizing the degree of enhanced nonlinear polarization and limited mass transfer are determined. The theoretical reference waveform is calculated based on the steady-state reference value before the excitation begins. The higher the response distortion characteristic value, the stronger the nonlinearity of the system. During electrolysis, this may indicate enhanced nonlinear polarization, such as changes in the reaction mechanism of the electrode surface, the presence of bubble coverage, or limited mass transfer, such as obstructed ion diffusion.

[0030] Preferably, transient impedance spectrum feature extraction is performed on the non-steady-state operation response signal. Specifically, a short-time Fourier transform is applied to the whole or segmented non-steady-state voltage and current signals to give a spectrum of the signal frequency components changing with time. From this time spectrum, transient impedance spectrum features are extracted, that is, the relationship between impedance and frequency during the perturbation. The transient impedance spectrum features include the change in equivalent charge transfer impedance and the slope of diffusion impedance evolution. The change in equivalent charge transfer impedance is the magnitude of the change in impedance value relative to the steady state in the high-frequency region (reflecting the electrochemical reaction steps). The slope of diffusion impedance evolution is the trend of the slope of the curve of impedance changing with frequency in the low-frequency region (reflecting the material diffusion steps). An increase in equivalent charge transfer impedance indicates a decrease in electrode activity and a slowdown in reaction kinetics. A change in the slope of diffusion impedance is used to reflect changes in electrolyte concentration distribution or boundary layer thickness.

[0031] Preferably, unsteady-state waveform phase drift analysis is performed on the unsteady-state operation response signal. This involves analyzing the relationship between the phase difference between the voltage response signal and the current excitation signal over time, calculating two key indicators: phase drift rate and phase hysteresis, and determining the phase drift characteristics. The phase drift rate refers to how quickly the phase difference changes over time, while the phase hysteresis refers to the degree of lag or incomplete recovery of the phase difference to its original value after the excitation is removed. Phase characteristics are closely related to the system's energy storage and dissipation; abnormal phase drift or hysteresis may indicate changes in the double-layer structure, changes in the interface adsorption process, or the presence of relaxation phenomena. Finally, the response distortion characteristics, transient impedance spectrum characteristics, and phase drift characteristics are combined into a structured dynamic feature set for output, used to comprehensively characterize the dynamic "fingerprint" of the electrolytic unit under perturbations.

[0032] Step S300: After extracting steady-state reference operating features based on steady-state operating signals, the features are jointly fused with dynamic feature sets and mapped to a unified feature projection space. Electrochemical state inversion processing is then performed within the feature projection space to generate electrochemical state feedback quantities.

[0033] Step S300 further includes performing time aggregation processing on the steady-state operating signal under a preset steady-state criterion to extract steady-state benchmark operating features characterizing the steady-state polarization level, energy conversion efficiency, and long-term load characteristics of the electrolysis unit; using the steady-state benchmark operating features as a reference, performing normalization mapping and relative offset calculation on the response distortion features, transient impedance spectrum features, and phase drift features in the dynamic feature set to construct a steady-state-non-steady-state correlated feature vector; performing feature weighted fusion dimension alignment processing on the steady-state-non-steady-state correlated feature vector and projecting it to a unified feature projection space; performing electrochemical state inversion operation based on the steady-state-non-steady-state correlated feature vector in the feature projection space to obtain an electrochemical state feedback quantity characterizing the current electrochemical reaction activity, mass transfer limitation degree, and aging evolution level of the electrolysis unit.

[0034] Preferably, the preset steady-state criterion refers to setting mathematical conditions for the judgment signal to reach steady state, such as "within N consecutive minutes (e.g., 30 minutes), the standard deviation of voltage fluctuation is less than the threshold δ". The steady-state operating signal is subjected to time aggregation processing under the preset steady-state criterion, that is, statistical aggregation is performed on the steady-state operating signal segment that meets the criterion within the time window. For example, the average value, median or specific percentile of parameters such as current density, cell voltage, temperature, and efficiency within this period are calculated, and steady-state benchmark operating characteristics that characterize the steady-state polarization level, energy conversion efficiency and long-term load characteristics of the electrolysis unit are extracted. Among them, the steady-state polarization level is characterized by the average cell voltage under a specific current density or the overpotential value it constitutes, reflecting the basic energy required to overcome resistance and drive reaction; the energy conversion efficiency is characterized by the ratio of actual gas production rate to theoretical gas production rate or the ratio of theoretical decomposition voltage to actual operating voltage, reflecting the effectiveness of energy utilization; the long-term load characteristics are characterized by the long-term average current density or the average energy consumption per unit of gas production, reflecting the basic operating load and energy consumption level of the equipment.

[0035] Preferably, the steady-state reference operating characteristics are used as a reference benchmark. The response distortion characteristics, transient impedance spectrum characteristics, and phase drift characteristics in the dynamic characteristic set are normalized and mapped, that is, divided by their respective design values, rated values, or historical normal ranges, and converted to a dimensionless scale, such as the interval [0, 1] or [-1, 1]. The normalized dynamic characteristic values ​​are subtracted from the normalized values ​​of the corresponding steady-state benchmark or their ratios are calculated to determine the relative offset. For example, the relative distortion degree = (current distortion degree - steady-state benchmark distortion degree) / steady-state benchmark distortion degree. The relative offset of the dynamic characteristics, the normalized steady-state benchmark characteristic values, and the original values ​​of the dynamic characteristics are then combined in sequence to form a structured feature vector to determine the steady-state-non-steady-state correlation feature vector, which is used to describe the dynamic disturbance response mode of the state relative to its own baseline.

[0036] Preferably, based on the physical importance of the features, signal-to-noise ratio, or contribution to state prediction, a weight coefficient is assigned to each component of the steady-state-nonsteady-state correlation feature vector to strengthen key features and suppress noisy features. Then, a feature weighting fusion and dimension alignment process is performed on the steady-state-nonsteady-state correlation feature vector to ensure that feature vectors from different batches and units have the same dimension and physical meaning order. Principal component analysis and linear discriminant analysis are used to project the weighted and aligned high-dimensional feature vectors onto a unified feature projection space with lower dimension and more independent features. Finally, within the feature projection space, an electrochemical state inversion process is performed based on the steady-state-nonsteady-state correlation feature vector; that is, within the feature projection space, the electrochemical state inversion process is called... An inversion model capable of outputting specific state parameters based on the projected high-dimensional feature vectors can perform electrochemical state inversion operations. The inversion model may be a multiple linear regression model, a support vector machine regression model, or a neural network model trained on a large amount of historical / experimental data. It learns the mapping relationship between feature patterns and the true state and outputs electrochemical state feedback quantities that characterize the current electrochemical reaction activity, mass transfer limitation, and aging evolution level of the electrolysis unit, i.e., internal true state variables. For example, electrochemical reaction activity is quantified as exchange current density or activation overpotential, mass transfer limitation is quantified as diffusion layer thickness or mass transfer coefficient, and aging evolution level is quantified as membrane resistance growth rate or electrode active area loss rate.

[0037] Step S400: Based on the electrochemical state feedback quantity, extract state evolution characteristics that characterize the direction, magnitude, and cumulative trend of electrochemical state changes, and determine the evolution interval in which the electrochemical state is located.

[0038] Preferably, for the electrochemical state feedback quantity, the state evolution characteristics characterizing the direction, magnitude, and cumulative trend of electrochemical state change are calculated. The direction of electrochemical state change is calculated by taking the first derivative or difference of the state quantity over the most recent one or several analysis periods, for example, (current value - previous value) / time interval. Its positive / negative value directly indicates whether the state is one of increased activity, decreased activity, or stability. The magnitude of change is calculated by taking the absolute or relative value of the state quantity change to quantify the severity of the change; for example, the membrane resistance increased by 5% in the past 24 hours. The cumulative trend involves trend analysis of the state quantity over a longer time window (e.g., several days or weeks), calculating the moving average, fitting the slope of a linear / nonlinear trend line, or analyzing whether it continuously exceeds the normal fluctuation range, thereby determining whether the change is temporary, periodic, continuous, or unidirectional. For example, fitting reveals that the diffusion layer thickness increased linearly at a slope of 0.1 μm per day over the past week, indicating a continuous and cumulative deterioration trend in mass transfer conditions.

[0039] Preferably, based on process knowledge, safety boundaries, and optimization objectives, several typical evolutionary regions are defined in advance. For example, in the constrained enhancement region, the direction of change of key performance state parameters is positive, and the change amplitude is greater than the positive threshold. The cumulative trend shows a clear and continuous improvement, and the electrolysis unit is in the performance recovery or optimization improvement stage due to the adjustment of process conditions. In the stabilization region, the direction of change of state parameters fluctuates slightly near zero, the change amplitude is lower than the fluctuation threshold, the cumulative trend is gentle, and there is no obvious directional drift. The electrolysis unit operates with high stability, and the internal state is in dynamic equilibrium. In the unstable accumulation region, the key state parameters... The direction of change is negative, the magnitude of change may increase from small to large, and the cumulative trend shows a clear and continuous deterioration. There is a gradual performance degradation or accumulation of adverse processes inside the electrolysis unit, such as increased scaling, catalyst deactivation, and membrane aging. The extracted state evolution characteristics characterizing the direction, magnitude, and cumulative trend of electrochemical state change are compared with the definition rules of each evolution interval, and the classification label or interval index is output as the judgment result. In this way, the complex continuous state change process is segmented and characterized, the evolution interval of the electrochemical state is determined, and it is classified into the mode category corresponding to different control strategies.

[0040] Step S500: Based on the state evolution characteristics generated in the evolution interval, a collaborative control strategy driven by electrolysis collaborative control management is executed.

[0041] Step S500 further includes: calculating the state offset of each electrolytic unit based on the state evolution characteristics, wherein the state offset includes the change amplitude, change rate, and accumulation trend; generating individual control gains using a weighted mapping method based on the distribution data of the evolution interval and the state offset, wherein the weighted mapping includes: a) assigning a first enhancement weight to electrolytic units in the restricted enhancement region to enhance positive regulation; b) assigning a second maintenance weight to electrolytic units in the stabilization region to maintain fine-tuning; c) assigning a third inhibitory weight to electrolytic units in the instability accumulation region to mitigate abnormal accumulation; fusing the individual control gains with the unsteady-state operation response signal and the steady-state reference operation characteristics to calculate the control adjustment amount of each electrolytic unit, wherein the control adjustment amount includes the adjustment amplitude, adjustment time, and action sequence; and performing global collaborative analysis based on the control adjustment amount of each electrolytic unit and the state evolution distribution of the group to establish a collaborative control strategy.

[0042] Preferably, the state evolution characteristics are standardized, and the state offset of each electrolytic unit is calculated, including the magnitude of change, the rate of change, and the cumulative trend. The magnitude of change is quantified as the absolute difference / percentage deviation between the current state parameter value and the set value or health baseline value; the rate of change is quantified as the change per unit time, such as the increase in ohms of membrane resistance per hour; and the cumulative trend is quantified as the absolute value of the long-term fitting slope or the cumulative deviation integral over a period of time. The distribution data of the evolution interval serves as the classification label for the evolution interval corresponding to each electrolytic unit. Then, a weighted mapping is performed using the distribution data of the evolution interval and the state offset as input to generate individual control gains. The weighted mapping includes assigning a first enhancement weight to the electrolytic units in the confined enhancement region to increase... A strong positive adjustment, with a weight greater than 1, is used to amplify the adjustment command when calculating the control quantity, encouraging the system to make more proactive positive adjustments in the direction of performance improvement, such as moderately increasing the current to consolidate and expand the improvement trend. A second maintenance weight is assigned to the electrolytic unit in the stabilization zone to maintain fine-tuning. This weight is approximately 1 and is used for conservative fine-tuning, using only a small control action to offset minor disturbances and maintain the status quo to avoid unnecessary intervention. A third inhibitory weight is assigned to the electrolytic unit in the instability accumulation zone to slow down abnormal accumulation. This weight is a positive coefficient less than 1 or a negative coefficient for reverse adjustment, used to suppress or reverse the control command. For example, when accelerated aging is detected, this weight reduces the current increase command or generates a load reduction command to slow down abnormal accumulation.

[0043] Preferably, the individual control gain is fused with the non-steady-state operation response signal reflecting the immediate dynamics and the steady-state reference operation characteristics reflecting the long-term benchmark. The control adjustment quantity of each electrolysis unit is calculated through adaptive PID, including the adjustment amplitude, adjustment time and action sequence, which respectively represent the absolute value or percentage of the current or voltage change, the duration of the adjustment action, and the time sequence of the adjustment action in multi-unit coordination, to determine the specific control command for each electrolysis unit. Then, global constraints and optimization objectives are preset, such as total load balance, fluid and heat balance, priority stabilization or isolation of units in a severely unstable state, and priority adjustment of highly efficient units under the premise of meeting production needs. Global collaborative analysis is performed based on the control adjustment quantity of each electrolysis unit and the state evolution distribution of the group. The analysis results are obtained, and the individual control adjustment quantities are coordinated, sorted or fine-tuned. Finally, a collaborative control strategy that can ensure the safe, stable and efficient operation of the entire system is established, and the overall efficiency of the collaborative operation of multiple electrolysis units is improved.

[0044] Furthermore, step S500 also includes: for the state offset of each electrolysis unit, using a dynamic time window calculation method, generating a comprehensive state index based on short-term perturbation response and long-term steady-state reference operating characteristics, and adaptively adjusting the dynamic threshold by calling historical cumulative trends and perturbation response sensitivity; comparing the comprehensive state index with the dynamic threshold to determine the state category of the electrolysis unit, and then calculating the individual control gain using a nonlinear mapping function; establishing control adjustment quantities based on the individual control gain, the unsteady-state operating response signal, and the steady-state reference to generate a cooperative control strategy.

[0045] Preferably, a dynamic time window is used to calculate the state offset of each electrolytic unit. The dynamic time window automatically adjusts the length of the analysis data according to the signal characteristics. For example, a short time window of seconds is used to analyze the high-frequency perturbation response, and a long time window of hours is used to analyze the long-term trend. Within their respective dynamic time windows, the short-term perturbation response and the long-term steady-state benchmark operating characteristics are weighted and fused to calculate, and a comprehensive state index is output to characterize the information of short-term dynamic anomalies and long-term performance drift. The fusion weight coefficient is determined based on historical data. Then, the historical cumulative trend and perturbation response sensitivity are used to adaptively adjust the dynamic threshold. If the system is in a slow aging process for a long time, the dynamic threshold for judging "instability" is relaxed accordingly to avoid false alarms for normal aging processes. If the unit's response to perturbations becomes abnormally sensitive, it indicates that the system stability has decreased. At this time, the dynamic threshold for judging "instability" or "stability" is tightened accordingly to improve the detection sensitivity of instability.

[0046] Preferably, the comprehensive state index is compared with a dynamic threshold to determine the state category of the electrolysis unit: a restricted enhancement region, a stabilization region, or an instability accumulation region. Individual control gain is calculated using a nonlinear mapping function. Specifically, this nonlinear mapping function may be an sigmoid function, an exponential function, or a piecewise function. When the exponent approaches the threshold boundary, the gain change is gradual; when the exponent penetrates a certain range, the gain change intensifies. The mapping relationship is determined using the comprehensive state index and state offset as inputs. For example, in the instability accumulation region, the gain decreases nonlinearly with the deterioration of the comprehensive state index, even becoming negative, to achieve more precise and gentle inhibitory regulation, thus reflecting the continuous deterioration or improvement of the state more meticulously. Finally, based on the individual control gain, the unsteady-state operation response signal, and the steady-state reference, adaptive PID control is used to calculate the control adjustment quantity, including the adjustment amplitude, adjustment time, and action sequence, to determine the cooperative control strategy and ensure a more delicate and reasonable correspondence between the control action and the severity of the state, improving the accuracy of electrolysis cooperative control.

[0047] Furthermore, step S500 also includes using the starting point of the electrolytic collaborative control as the time zero point to construct a time evolution sequence of the electrochemical state of the electrolyte device under collaborative control; configuring a collaborative feedback signal according to the time evolution sequence to perform self-optimization management of the collaborative control strategy.

[0048] Preferably, the moment of first execution of the collaborative control strategy is taken as the time zero point. Electrochemical state feedback quantities output by the state inversion model, including the reactivity and mass transfer coefficient of each unit, are continuously recorded at fixed intervals (e.g., every minute or every control cycle). These time-sequentially arranged state data points are used to construct a time-series evolution sequence of the electrochemical state of the electrolyte device under collaborative control. Then, the time-series evolution sequence is analyzed to extract indicators that can evaluate the effectiveness of the control strategy and configure them as collaborative feedback signals. These indicators may include state convergence speed, overall state stability, performance improvement rate, and strategy consistency. Finally, the self-optimization management of the collaborative control strategy is implemented, which involves dynamically adjusting the collaborative control strategy using collaborative feedback signals. For example, if the improvement effect of an electrolysis unit is found to be unsatisfactory, the shape of the first, second, and third weights or nonlinear mapping functions of its respective interval is adjusted. Based on the actual state evolution data, the dynamic thresholds for determining the stable and unstable regions are recalibrated to better match the actual equipment characteristics. Based on the overall efficiency improvement, the optimization objectives of global parameters such as total load distribution and adjustment sequence in the collaborative analysis are adjusted to achieve closed-loop learning and strategy iteration, thereby improving the overall energy efficiency and lifespan of the multi-electrolysis unit collaborative operation.

[0049] In the above text, refer to Figure 1 The electrolysis coordinated control method based on electrochemical state feedback according to embodiments of the present invention is described in detail. Next, reference will be made to... Figure 2 An electrolysis coordinated control system based on electrochemical state feedback according to an embodiment of the present invention is described.

[0050] The electrolysis collaborative control system based on electrochemical state feedback according to embodiments of the present invention addresses the technical problems in the prior art, such as lagging and one-sided electrochemical state perception, lack of dynamic information utilization, and coarse multi-unit collaborative control. It achieves the technical effects of improving the real-time nature of state perception in the electrolysis process, the accuracy and foresight of collaborative control, and the overall energy efficiency and lifespan of multi-electrolysis unit collaborative operation. Figure 2 As shown, the electrolysis collaborative control system based on electrochemical state feedback includes: an operation signal acquisition module 10, a dynamic feature acquisition module 20, a state feedback quantity generation module 30, a state evolution feature extraction module 40, and a collaborative control management module 50.

[0051] The operation signal acquisition module 10 is used to continuously acquire steady-state operation signals characterizing the long-term operating conditions of each electrolysis unit during the operation of the electrolyte equipment. Based on the steady-state operation signals, it configures the influence constraints of oxygen production conditions and applies controlled amplitude and timing perturbation excitations to each electrolysis unit under these constraints, simultaneously acquiring the non-steady-state operation response signals caused by the perturbation excitations. The dynamic feature acquisition module 20 is used to perform response distortion analysis, transient impedance spectrum feature extraction, and non-steady-state waveform phase drift analysis on the non-steady-state operation response signals to obtain a dynamic feature set. The state feedback quantity generation module 30... The system is used to extract steady-state baseline operating features based on steady-state operating signals, then fuse them with dynamic feature sets and map them to a unified feature projection space. Electrochemical state inversion processing is performed within the feature projection space to generate electrochemical state feedback quantities. The state evolution feature extraction module 40 is used to extract state evolution features characterizing the direction, magnitude, and cumulative trend of electrochemical state changes based on the electrochemical state feedback quantities, and to determine the evolution interval of the electrochemical state. The collaborative control management module 50 is used to generate a collaborative control strategy driven by state evolution features based on the evolution interval, and to execute electrolysis collaborative control management.

[0052] The specific configuration of the collaborative control management module 50 will be described in detail below. The collaborative control management module 50 further includes: calculating the state offset of each electrolytic unit based on the state evolution characteristics, wherein the state offset includes the magnitude of change, the rate of change, and the cumulative trend; generating individual control gains using a weighted mapping method based on the distribution data of the evolution interval and the state offset, wherein the weighted mapping includes: a) assigning a first enhancement weight to electrolytic units in the restricted enhancement region to enhance positive regulation; b) assigning a second maintenance weight to electrolytic units in the stabilization region to maintain fine-tuning; c) assigning a third inhibitory weight to electrolytic units in the unstable accumulation region to mitigate abnormal accumulation; fusing the individual control gains with the unsteady-state operation response signal and the steady-state reference operation characteristics to calculate the control adjustment amount of each electrolytic unit, wherein the control adjustment amount includes the adjustment amplitude, the adjustment time, and the action sequence; and performing global collaborative analysis and establishing a collaborative control strategy based on the control adjustment amount of each electrolytic unit and the state evolution distribution of the group.

[0053] The specific configuration of the collaborative control management module 50 will be described in detail below. The collaborative control management module 50 further includes: for the state offset of each electrolysis unit, a dynamic time window calculation method is used to generate a comprehensive state index based on short-term perturbation response and long-term steady-state baseline operating characteristics, and the dynamic threshold is adaptively adjusted by calling historical cumulative trends and perturbation response sensitivity; after comparing the comprehensive state index with the dynamic threshold to determine the state category of the electrolysis unit, an individual control gain is calculated using a nonlinear mapping function; and a control adjustment quantity is established based on the individual control gain, the unsteady-state operating response signal, and the steady-state baseline to generate a collaborative control strategy.

[0054] The specific configuration of the dynamic feature acquisition module 20 will be described in detail below. The dynamic feature acquisition module 20 further includes: dividing the unsteady-state operating response signal caused by perturbation excitation into an excitation initial segment, an excitation transition segment, and an excitation decay segment according to the excitation timing; within each timing segment, performing distortion calculations based on a reference steady state on the voltage and current response waveforms to obtain response distortion features characterizing the degree of nonlinear polarization enhancement and mass transfer limitation; performing a short-time Fourier transform on the unsteady-state operating response signal to extract transient impedance spectrum features, which include the change in equivalent charge transfer impedance and the slope of diffusion impedance evolution; tracking the phase change of the unsteady-state operating response signal over time, calculating the phase drift rate and phase hysteresis, and constructing phase drift features; and outputting the response distortion features, transient impedance spectrum features, and phase drift features as a dynamic feature set.

[0055] The specific configuration of the state feedback generation module 30 will be described in detail below. The state feedback generation module 30 further includes: performing time aggregation processing on the steady-state operating signal under a preset steady-state criterion to extract steady-state reference operating features characterizing the steady-state polarization level, energy conversion efficiency, and long-term load characteristics of the electrolysis unit; using the steady-state reference operating features as a reference, performing normalization mapping and relative offset calculation on the response distortion features, transient impedance spectrum features, and phase drift features in the dynamic feature set to construct a steady-state-non-steady-state correlated feature vector; performing feature weighted fusion dimension alignment processing on the steady-state-non-steady-state correlated feature vector and projecting it to a unified feature projection space; performing electrochemical state inversion operation based on the steady-state-non-steady-state correlated feature vector within the feature projection space to obtain an electrochemical state feedback quantity characterizing the current electrochemical reaction activity, mass transfer limitation, and aging evolution level of the electrolysis unit.

[0056] The specific configuration of the operation signal acquisition module 10 will be described in detail below. The operation signal acquisition module 10 further includes: determining the steady-state current density of the electrolysis unit under oxygen production conditions based on the steady-state operation signal, and using the steady-state current density as a reference operating point; constructing a current disturbance allowable range with the reference operating point as the center according to a preset relative offset ratio, the current disturbance allowable range including an upper offset boundary and a lower offset boundary; and outputting the current disturbance allowable range as an influence constraint.

[0057] The specific configuration of the collaborative control management module 50 will be described in detail below. The collaborative control management module 50 further includes: using the starting point of the electrolytic collaborative control as the time zero point, constructing a time-series evolution sequence of the electrochemical state of the electrolyte device under collaborative control; configuring a collaborative feedback signal according to the time-series evolution sequence, and executing self-optimization management of the collaborative control strategy.

[0058] The electrolysis collaborative control system based on electrochemical state feedback provided in this embodiment of the invention can execute the electrolysis collaborative control method based on electrochemical state feedback provided in this embodiment of the invention, and has the corresponding functional modules and beneficial effects of the execution method.

[0059] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. An electrolysis coordinated control method based on electrochemical state feedback, characterized in that, The method includes: During the operation of the electrolyte equipment, for each electrolysis unit, a steady-state operating signal characterizing the long-term operating condition of the electrolysis unit is continuously collected. After configuring the influence constraints of the oxygen production condition based on the steady-state operating signal, a micro-perturbation excitation with controlled amplitude and timing is applied to each electrolysis unit under the influence constraints, and the non-steady-state operating response signal caused by the micro-perturbation excitation is collected simultaneously. Response distortion analysis, transient impedance spectrum feature extraction, and unsteady waveform phase drift analysis are performed on the unsteady operation response signal to obtain a dynamic feature set; After extracting steady-state benchmark operating features based on steady-state operating signals, they are jointly fused with dynamic feature sets and mapped to a unified feature projection space. Electrochemical state inversion processing is then performed within the feature projection space to generate electrochemical state feedback quantities. Based on the electrochemical state feedback, state evolution features characterizing the direction, magnitude, and cumulative trend of electrochemical state changes are extracted, and the evolution interval of the electrochemical state is determined. Based on the state evolution characteristics generated in the evolution interval, a collaborative control strategy is implemented to perform electrolytic collaborative control management.

2. The electrolysis coordinated control method based on electrochemical state feedback as described in claim 1, characterized in that, A collaborative control strategy driven by state evolution characteristics generated based on the aforementioned evolution interval includes: The state offset of each electrolysis unit is calculated based on the state evolution characteristics, and the state offset includes the magnitude of change, the rate of change, and the cumulative trend. Based on the distribution data and state offsets within the evolutionary interval, an individual control gain is generated using a weighted mapping method. The weighted mapping includes: a: Assign a first enhancement weight to the electrolytic unit in the restricted enhancement region to enhance positive regulation; b: Assign a second maintenance weight to the electrolytic cells in the stabilization region to maintain fine-tuning; c: Assign a third inhibitory weight to the electrolytic unit in the unstable accumulation zone to mitigate abnormal accumulation; The individual control gain is fused with the unsteady-state operation response signal and the steady-state reference operation characteristics to calculate the control adjustment amount of each electrolysis unit. The control adjustment amount includes the adjustment amplitude, adjustment time and action sequence. Based on the control and regulation quantities of each electrolysis unit and the state evolution distribution of the group, a global collaborative analysis is performed to establish a collaborative control strategy.

3. The electrolysis coordinated control method based on electrochemical state feedback as described in claim 2, characterized in that, The collaborative control strategy driven by the state evolution characteristics generated based on the evolution interval also includes: For the state offset of each electrolysis unit, a dynamic time window calculation method is adopted to generate a comprehensive state index based on the short-term perturbation response and long-term steady-state benchmark operation characteristics, and the dynamic threshold is adaptively adjusted by calling the historical cumulative trend and perturbation response sensitivity. After comparing the comprehensive state index with the dynamic threshold to determine the state category of the electrolysis unit, the individual control gain is calculated using a nonlinear mapping function. Based on the individual control gain, the unsteady-state operation response signal, and the steady-state reference, control adjustment quantities are established to generate a collaborative control strategy.

4. The electrolysis coordinated control method based on electrochemical state feedback as described in claim 1, characterized in that, The unsteady-state operating response signal is subjected to response distortion analysis, transient impedance spectrum feature extraction, and unsteady-state waveform phase drift analysis to obtain a dynamic feature set, including: The unsteady-state response signal caused by perturbation excitation is divided into excitation initial segment, excitation transition segment and excitation decay segment according to the excitation timing sequence. Within each time segment, the distortion degree of the voltage and current response waveforms is calculated based on the reference steady state to obtain response distortion characteristics that characterize the degree of nonlinear polarization enhancement and mass transfer limitation. A short-time Fourier transform is performed on the unsteady-state operation response signal to extract transient impedance spectrum features, which include the change in equivalent charge transfer impedance and the slope of diffusion impedance evolution. The phase change of the non-steady-state operation response signal over time is tracked, the phase drift rate and phase hysteresis are calculated, and the phase drift characteristics are constructed. The response distortion characteristics, transient impedance spectrum characteristics, and phase drift characteristics are output as a set of dynamic characteristics.

5. The electrolysis coordinated control method based on electrochemical state feedback as described in claim 1, characterized in that, After extracting steady-state baseline operating features based on steady-state operating signals, these features are jointly fused with dynamic feature sets and mapped to a unified feature projection space. Electrochemical state inversion processing is then performed within this feature projection space, including: The steady-state operating signal is subjected to time aggregation processing under a preset steady-state criterion to extract steady-state reference operating characteristics that characterize the steady-state polarization level, energy conversion efficiency and long-term load characteristics of the electrolysis unit; Using the steady-state benchmark operating characteristics as a reference benchmark, normalization mapping and relative offset calculation are performed on the response distortion characteristics, transient impedance spectrum characteristics and phase drift characteristics in the dynamic feature set to construct a steady-state-nonsteady-state correlated feature vector; The steady-state-nonsteady-state correlation feature vectors are subjected to feature weighting fusion dimension alignment processing and projected onto a unified feature projection space. Electrochemical state inversion operation is performed based on the steady-state-nonsteady-state correlation feature vectors in the feature projection space to obtain the electrochemical state feedback quantity characterizing the current electrochemical reaction activity, mass transfer limitation degree and aging evolution level of the electrolysis unit.

6. The electrolysis coordinated control method based on electrochemical state feedback as described in claim 1, characterized in that, The influence constraints of oxygen production conditions are configured based on the steady-state operating signals, including: The steady-state current density of the electrolysis unit under oxygen production conditions is determined based on the steady-state operating signal, and the steady-state current density is used as the reference operating point. Centered on the reference operating point, a current disturbance allowable range is constructed according to a preset relative offset ratio. The current disturbance allowable range includes an upper offset boundary and a lower offset boundary. The allowable range of current disturbance is output as an influence constraint.

7. The electrolysis coordinated control method based on electrochemical state feedback as described in claim 1, characterized in that, The implementation of electrolysis coordinated control management also includes: By taking the starting point of electrolytic collaborative control as the time zero point, a time evolution sequence of the electrochemical state of the electrolyte device under collaborative control is constructed. Configure collaborative feedback signals according to the time-series evolution sequence, and perform self-optimization management of collaborative control strategies.

8. An electrolysis coordinated control system based on electrochemical state feedback, characterized in that, The system is used to implement the electrolysis collaborative control method based on electrochemical state feedback as described in any one of claims 1 to 7, and the system comprises: The operation signal acquisition module is used to continuously acquire steady-state operation signals characterizing the long-term operating conditions of each electrolysis unit during the operation of the electrolyte equipment. After configuring the influence constraints of oxygen production conditions based on the steady-state operation signals, it applies controlled amplitude and timing perturbation excitations to each electrolysis unit under the influence constraints, and simultaneously acquires the non-steady-state operation response signals caused by the perturbation excitations. The dynamic feature acquisition module is used to perform response distortion analysis, transient impedance spectrum feature extraction and non-steady-state waveform phase drift analysis on the non-steady-state operating response signal to obtain a dynamic feature set. The state feedback quantity generation module is used to extract steady-state reference operating features based on steady-state operating signals, then jointly fuse them with dynamic feature sets and map them to a unified feature projection space. Electrochemical state inversion processing is then performed in the feature projection space to generate electrochemical state feedback quantities. The state evolution feature extraction module is used to extract state evolution features that characterize the direction, magnitude, and cumulative trend of electrochemical state changes based on the electrochemical state feedback quantity, and to determine the evolution interval in which the electrochemical state is located. The collaborative control management module is used to generate a collaborative control strategy driven by the state evolution characteristics based on the evolution interval, and to execute electrolytic collaborative control management.

Images