A zinc-air battery gas-liquid electricity collaborative control method based on multiple physical evidence frames

Abstract

The application discloses a zinc-air battery gas-liquid-electricity collaborative control method based on multiple physical evidence frames and belongs to the technical field of zinc-air battery control. Related parameters of a zinc-air battery or a battery pack are collected to construct a zinc-air state evidence frame. Impedance characteristics, air path pressure drop characteristics, air supply response characteristics, electrolyte concentration proxy characteristics and polarization characteristics are extracted from the zinc-air state evidence frame. Abnormal instability modes are identified according to the above characteristics. Then, air-liquid-electricity coupling constraints and control guard belts are generated in combination with current charging and discharging conditions. Collaborative execution instructions are generated in the control guard belts to adjust the battery state. The application organizes the specific gas, liquid, electricity, heat and electrode degradation state of the zinc-air battery into an interpretable closed-loop control chain, can reduce the operation fluctuation caused by the air supply shortage, flooding, dryness, carbonate blockage, zinc electrode passivation or dendrite risk, and improve the stability and maintainability of the zinc-air battery system.

Description

Technical Field

[0001] This invention relates to the technical field of zinc-air battery control, and in particular to a gas-liquid-electric coordinated control method for zinc-air batteries based on multiple physical evidence frames. Background Technology

[0002] Zinc-air batteries utilize the electrochemical reaction between zinc and air electrodes to output electrical energy. They feature high theoretical energy density, good safety, and relatively abundant material sources, making them suitable for applications such as backup power, energy storage systems, special power supplies, and wearable or portable power sources. For rechargeable zinc-air batteries or zinc-air battery packs, the system typically includes components for air supply, electrolyte circulation, electrolyte replenishment and drainage, air electrode humidity management, charge / discharge control, and three-electrode switching.

[0003] Compared to common rechargeable batteries like lithium-ion batteries, the operating state of zinc-air batteries is not only determined by voltage, current, and temperature, but is also significantly affected by the gas path, electrolyte, and air electrode condition. For example, insufficient air supply during discharge can cause a voltage drop; excessive moisture in the air electrode can lead to flooding, while excessive dryness can increase polarization; carbon dioxide in the air can cause carbonate accumulation and affect the air path or electrolyte condition; and the zinc electrode may also experience passivation, uneven deposition, or dendrite formation during charging and discharging. These issues are often interdependent, and simply looking at the terminal voltage is not enough to determine the true cause.

[0004] Existing battery management solutions commonly monitor voltage, current, and temperature to estimate the state of charge or state of health. Other solutions control the air supply, electrolyte level, or related valves based on pressure, voltage, or capacity thresholds. While these methods are effective for general batteries or simple operating scenarios, they have limitations when used with zinc-air batteries.

[0005] First, the discharge voltage plateau of zinc-air batteries is relatively flat. Simply relying on voltage or impedance to estimate the state of charge makes it difficult to explain specific instability causes such as air path blockage, water flooding of the air electrode, insufficient air supply, and passivation of the zinc electrode.

[0006] Second, there are mutual influences among the gas path, electrolyte, and charge / discharge control. Increasing the air supply may alleviate insufficient air supply, but it may also exacerbate the drying of the air electrode under certain humidity conditions; increasing the electrolyte circulation rate may improve the concentration distribution, but it may also lead to gas-liquid entrainment, liquid level fluctuations, or temperature changes; increasing the charging current may shorten the charging time, but it may also increase the risk of dendrite formation or hydrogen evolution.

[0007] Third, existing threshold controls often only answer "whether a certain pump or valve is open", but cannot fully answer "which zinc-air instability mode is currently in operation", "whether different actuators should move simultaneously", "whether the range of motion is safe", and "whether the system has truly recovered after execution". Summary of the Invention

[0008] The purpose of this invention is to provide a gas-liquid-electro-coordinated control method for zinc-air batteries based on multiple physical evidence frames, which solves the problems in the prior art where the state judgment of zinc-air batteries only depends on a single electrical signal, the control of the gas path and electrolyte is fragmented, the threshold start-stop is difficult to explain the cause of instability, the control action lacks boundary constraints, and the execution effect is difficult to write back.

[0009] To achieve the above objectives, this invention provides a gas-liquid-electric coordinated control method for zinc-air batteries based on multiple physical evidence frames, comprising the following steps: Step S1, in the corresponding sampling window In the process, electrical signals, gas path signals, electrolyte signals, thermal signals, and actuator status of the target zinc-air battery are collected to construct a zinc-air state evidence frame. ; Step S2: Based on the zinc-air state evidence frame, the impedance characteristics, gas path pressure drop characteristics, air supply response characteristics, electrolyte concentration proxy characteristics, and polarization characteristics corresponding to the target zinc-air battery are extracted using the window synchronous feature extraction and evidence quality weight calculation method. Step S3: Generate an instability mode vector based on the impedance characteristics, gas path pressure drop characteristics, air supply response characteristics, electrolyte concentration proxy characteristics, and polarization characteristics. The instability mode vector includes at least two of the following: insufficient air supply, water flooding of the air electrode, dryness of the air electrode, carbonate blockage, passivation of the zinc electrode, and dendrite risk. Step S4, based on the instability mode vector Based on the current charging and discharging conditions, corresponding gas-liquid-electric coupling constraints are generated. ; Step S5: Determine the allowable control range for each actuator based on the gas-liquid-electric coupling constraint, thus forming a control gate zone. ; Step S6, in the control goal zone Within the specified limits, collaborative execution instructions are formed. And then distribute it to the corresponding executor for execution; Step S7: Collect the execution receipt in the observation window after execution. It updates the instability mode identification parameters or control gate based on changes in terminal voltage, impedance, gas path pressure drop, electrolyte parameters or efficiency, and records whether the action is recovered, whether it exceeds the limit, and whether manual intervention is required.

[0010] Preferably, in step S1, the electrical signals include, but are not limited to, cell or module voltage, current, charge / discharge status, estimated open-circuit voltage, and impedance measurement results; the gas path signals include, but are not limited to, air channel differential pressure, air flow rate, carbon dioxide concentration, humidity, dew point, oxygen concentration detection value, or oxygen-enriched air supply status; the electrolyte signals include, but are not limited to, electrolyte temperature, conductivity, pH, liquid level, circulation pump status, replenishment status, draining status, or turbidity; the thermal signals include, but are not limited to, battery cell temperature, module temperature, gas path temperature, or electrolyte temperature, providing evidence of the zinc-air state. Record electrical signals, gas path signals, electrolyte signals, thermal signals, and actuator status within the same sampling window, and record sensor validity markers and stability parameters within the sampling window; corresponding to the sampling window... Internally, the controller constructs zinc void state evidence frames. as follows: ; In the above formula, It represents a set of electrical signals, including voltage, current, charging / discharging state, and impedance information; It represents a set of airway signals, including airway pressure difference, airflow, carbon dioxide concentration, humidity, dew point, and oxygen concentration or oxygen-enriched air supply status based on the battery settings. It represents a set of electrolyte signals, including conductivity, pH, liquid level, temperature, pump status, replenishment / drainage status, and turbidity; It represents a set of thermal signals, including unit temperature, module temperature, gas path temperature, and electrolyte temperature; This represents the set of actuator states, including the states of pumps, valves, fans, air electrode humidity control units, charge / discharge controllers, and three-electrode switching units. Sensor validity markers include missing data markers, drift markers, mutation markers, and contradictory markers.

[0011] Preferably, the process of step S2 is as follows: corresponding to from Engineering features extracted for identifying instability modes include: impedance features, including high-frequency impedance, low-frequency impedance, impedance change rate, and impedance recovery within a rest window; air path pressure drop features, including air channel differential pressure, the relationship between air flow rate and differential pressure, and the trend of differential pressure change over time; air supply response features, including air flow rate, air channel differential pressure, terminal voltage response after load changes, and oxygen concentration or oxygen-enriched air supply status; electrolyte concentration proxy features, including conductivity, pH, liquid level, temperature, and replenishment / drainage status; polarization features, including discharge voltage drop magnitude, charging overpotential, load step response, and temperature rise; actuator response features, including whether the flow rate follows the pump speed change, whether the air path pressure drop or flow rate changes after valve action, and whether the differential pressure recovers after fan action; and the conversion of zinc-air state evidence frames into engineering features. The calculation formula is: ; in, For the synchronous sampling window corresponding to the k-th window, Let m be the window synchronization feature extraction function corresponding to the m-th type of signal source. This includes amplitude, rate of change, recovery, consistency, and anomaly markers; to prevent erroneous actions caused by faulty sensors, evidence quality weights are set. Used to limit The credibility of evidence used in subsequent instability pattern identification corresponds to the generation of evidence quality weights. The formula is as follows: ; In the above formula, m represents a certain type of signal source, including electrical signals, gas signals, electrolyte signals, or thermal signals; This indicates the effectiveness of the sensor from that source; it decreases when there are missing measurements or when the sensor drifts. This indicates the stability of the source within the sampling window, and is reduced when the abrupt change is too large or when it contradicts adjacent sensors; clip indicates that the output result range is set between 0 and 1; when the sensor is missing, drifting, abruptly changing, or contradictory, the weight of the corresponding signal source in the subsequent instability mode recognition is reduced, and the actuator action is prohibited from being directly amplified based solely on the abnormal signal.

[0012] Preferably, in step S3, based on the engineering features extracted in step S2, the instability mode vector is generated as follows: ; In the above formula, This indicates a risk of insufficient air supply. This indicates the risk of water damage to the air electrode. This indicates the risk of the air electrode drying out. This indicates a risk of carbonate blockage or carbonate accumulation. This indicates the risk of zinc electrode passivation. Indicates dendrite risk, Indicating sensor malfunction risk, the above components are based on the engineering features obtained in step S2. and the weight of evidence quality The h-th type of instability mode is identified by weighting, as shown in the following formula: ; In the above formula, h represents one of the following modes: insufficient air supply, flooding, dryness, carbonate blockage, zinc electrode passivation, dendrite formation, or sensor malfunction. It can be one or a combination of rule-based models, state estimation models, lookup table models, and classification models, where state(k) represents the current charging and discharging condition.

[0013] Preferably, in step S3, the identification methods corresponding to different risks are as follows: insufficient air supply risk is identified based on the decrease in terminal voltage after load increase, insufficient air flow, abnormal air path pressure drop, or decrease in oxygen concentration in the oxygen-enriched air supply or closed air channel implementation; air electrode flooding risk is identified based on the increase in air path pressure drop, terminal voltage fluctuation, high-frequency impedance change, and abnormal humidity or dew point; air electrode drying risk is identified based on the decrease in humidity, increase in polarization, abnormal temperature rise, and slow voltage recovery; carbonate blockage risk or carbonate accumulation risk is identified based on carbon dioxide exposure, changes in electrolyte conductivity, changes in air path pressure drop, and historical operating time; zinc electrode passivation risk is identified based on the increase in low-frequency impedance, increase in charging or discharging overpotential, abnormal electrolyte concentration characteristics, and insufficient capacity recovery; dendrite risk is identified based on charging current density, low-frequency impedance change, sudden change in charging overpotential, temperature rise, electrolyte state, and historical cycle count.

[0014] Preferably, in step S4, according to Generates gas-liquid-electric coupling constraint with the current charge-discharge conditions. Gas-liquid-electric coupling constraint At least including: (1) When the risk of insufficient air supply is high, increase the air supply. However, if the risk of the air electrode drying out is also high, it is necessary to limit the increase in the drying airflow or to perform air electrode humidity compensation first. (2) When the risk of water flooding of the air electrode is high, short-term flow increase, drainage, reduction of electrolyte supply or rest and recovery shall be performed, but the amount of air supply required for the reaction on the air electrode side shall not be lower than the lower limit of discharge. (3) When the risk of carbonate blockage is high, carbon dioxide absorption bypass, gas path regeneration, conservative load reduction or maintenance prompts shall be implemented, but the pressure drop shall not be achieved by simply increasing the fan speed for a long period of time. (4) When the risk of zinc electrode passivation is high, the battery should be stopped, electrolyte flow increased, electrolyte replaced or low-rate recovery performed, but the charging current should not be directly increased in an attempt to quickly restore capacity. (5) When the dendrite risk is high, reduce current, pulse charge, pause or electrolyte circulation enhancement shall be performed, but the charging current density and temperature rise boundaries shall not be exceeded. Based on the aforementioned risks, gas-liquid-electric coupling constraint The formula is as follows: In the above formula, It indicates the gas path boundary, including airflow, air channel pressure difference, carbon dioxide, humidity and pressure drop limits, and includes oxygen concentration limits under the oxygen-enriched air supply implementation method; It indicates the electrolyte boundary, including liquid level, conductivity, pH, temperature and pump speed limits; It indicates the electrical control boundary, including limits on charging and discharging current, voltage, rest time, and three-electrode switching time; This indicates the boundaries of manual authorization, protective interlocking, and conservative modes.

[0015] Preferably, in step S5, the process of generating the control goal belt is as follows: Actuators for regulating battery operation include circulation pumps, air valves, fans, air electrode humidity control units, replenishment valves, drain valves, charge / discharge controllers, or three-electrode switching units; in implementations where the battery is in an oxygen-enriched air supply or closed air passage, the actuator also includes an oxygen-enriched air supply actuator. For any actuator... The corresponding formula for controlling the goal belt is as follows: ; In the above formula, and These represent the lower and upper limits of the actuator hardware, respectively; Indicates the execution volume of the previous window; and Indicates the maximum amount of change that adjacent windows are allowed to decrease or increase; and Indicates by The given current chemical, electrical, or gas-liquid boundary; The control gate is set with upper and lower hardware safety limits, maximum change in adjacent control cycles, electrochemical risk boundary, gas path risk boundary, electrolyte risk boundary, and manual authorization boundary. When the instability mode vector shows sensor abnormality, actuator abnormality, severe gas path blockage, or unreliable electrolyte status, the control gate will shrink the corresponding actuator to a conservative control range. For any actuator, the allowable control range is jointly defined by the actuator hardware upper and lower limits, the maximum change in adjacent control cycles, and the current boundary given by the gas-liquid-electric coupling constraint.

[0016] Preferably, in step S6, a cooperative execution instruction is generated. The process is as follows: The coordinated execution instructions include the target actuator, target change direction, target change magnitude, latest execution time, effective observation window, backoff condition, and cause code. Under the constraints of the control gate, the controller selects the control vector corresponding to the coordinated execution instructions based on the improvement cost, side effect cost, and energy consumption or efficiency cost of the current main instability mode. , Indicates the first The actual selected control vector in each control window, u represents a candidate control vector, and the comprehensive cost of the candidate action is calculated as follows: ; In the above formula, This indicates that the comprehensive cost is calculated for the candidate control vector u in the k-th control window; This indicates the cost of improving the current main instability mode; the lower the cost, the more conducive it is to recovery. This indicates the side effects and costs, including the risks of drying up, flooding, dendrite formation, temperature rise, or exceeding the liquid level limit; Indicates energy consumption or efficiency cost; controller from Select the control vector with the lowest overall cost and that meets the conditions for manual verification. The candidate action set is represented as: ; The controller's formula is derived from at least one of the following: pre-set action templates, rule tables, model predictive control candidate sequences, maintenance strategy libraries, or manually authorized actions; ; Select candidate control vectors that meet the conditions of control gate zone and manual confirmation, and then... The package is formed into ACI(k).

[0017] Preferably, step S7 is used to determine the execution effect of the cooperative execution instruction, and the corresponding formula is as follows: ; In the above formula, Indicates the amount of voltage recovery at the terminals; This indicates the amount of abnormal impedance drop; This indicates the abnormal drop in gas pressure. This indicates the degree to which the overall electrolyte condition index has recovered to the target range; The risk of temperature rise, liquid level exceeding the limit, protection triggering, or manual intervention after execution is obtained by weighting the temperature rise exceeding the limit flag, liquid level exceeding the limit flag, protection triggering flag, and manual intervention flag. to Weights are assigned; calculations are performed based on the observation window before and after the action. , , and The formula is as follows: ; ; ; ; ; In the above formula, pH and L represent the normalized state variables corresponding to electrolyte conductivity, acidity / alkalinity, and liquid level, respectively. to Indicates the state weight of the electrolyte; when When the set range is reached, and the gas path, electrolyte, and temperature do not exceed the limits, the controller increases the matching confidence between the current instability mode and the corresponding cooperative execution command; the matching confidence is the h-th type of instability mode within the k-th control window. With corresponding control vector Historical effective matching degree between The updated formula is: ; in and To increase or decrease the step size for confidence level, To restore the evaluation threshold, This is an indicator function that takes 1 if the condition is true and 0 otherwise; when When the match confidence level is low, or when there is a breach of control or manual intervention, the controller lowers the confidence level of the match, shrinks the relevant control gate, and records the reason for the non-fulfillment.

[0018] Preferably, when the target zinc-air battery is a battery pack, a cell or module consistency diagnosis step is set for the corresponding battery pack: the source of cell differences is determined based on the cell voltage, cell temperature, cell impedance, air branch pressure difference, electrolyte branch flow rate and valve status, and coordinated control is performed on the air branch, electrolyte branch, module bypass, load reduction or equalization operation.

[0019] Therefore, the present invention employs the above-described gas-liquid-electric coordinated control method for zinc-air batteries based on multiple physical evidence frames, which has the following advantages: (1). This application unifies electrical signals, gas circuit signals, electrolyte signals, thermal signals and actuator status into a zinc-air state evidence frame to avoid misjudgment caused by looking only at voltage, current or a single impedance index.

[0020] (2). This application distinguishes between insufficient air supply, water flooding of air electrodes, dryness of air electrodes, carbonate blockage, passivation of zinc electrodes and dendrite risk by instability mode vector, so that the control system first judges "what the problem is" and then decides "which actuator should be moved".

[0021] (3). This application uses gas-liquid-electric coupling constraint to process the gas path, electrolyte and charging and discharging action within the same boundary, avoiding new side effects caused by individually increasing the air supply, electrolyte circulation or charging current.

[0022] (4). This application makes the output action more compliant with hardware safety, electrochemical safety and field authorization requirements by controlling the gate to limit the allowable range of pump, valve, fan, air electrode humidity adjustment, replenishment and drainage, charging and discharging current and three-electrode switching.

[0023] (5). This application records the changes in terminal voltage, impedance, gas path pressure drop, electrolyte parameters and efficiency after execution by running a receipt, so that the effectiveness of a single control action can be verified and the identification parameters and control gate can be further corrected.

[0024] (6) The computational processing part of this application does not rely on a fixed black-box prediction structure and can be implemented using rule models, state estimation models, model predictive control, lookup table models, or equivalent methods. The focus is on the unique evidence of zinc-air batteries, instability modes, boundary constraints, and the continuous closed loop between physical actuators. The technical solution of this invention will be further described in detail below with reference to the accompanying drawings and embodiments. Attached Figure Description

[0025] Figure 1 This is the overall flowchart of the zinc-air battery gas-liquid-electro-coordinated control method of this application.

[0026] Figure 2 This is a schematic diagram of the structure of the zinc-air battery gas-liquid-electro-coordinated control system of this application, showing the relationship between the battery pack, air supply unit, electrolyte circulation unit, sensor, controller and actuator.

[0027] Figure 3 This diagram illustrates the relationship between the zinc void state evidence frame, feature extraction, and instability mode vector in this application.

[0028] Figure 4 This is a schematic diagram illustrating the relationship between the gas-liquid-electric coupling constraint, the control gate zone, and the cooperative execution command in this application.

[0029] Figure 5This is a closed-loop diagram illustrating the operation receipt, conservative control, and parameter write-back of this application. Detailed Implementation

[0030] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, not all embodiments. The components of the embodiments of the present invention described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Specific model specifications need to be selected and determined according to the actual specifications of the device, etc. The specific selection calculation method adopts existing technology in the art, and therefore will not be described in detail.

[0031] Example like Figure 1 As shown, this invention provides a gas-liquid-electric coordinated control method for zinc-air batteries based on multiple physical evidence frames, comprising the following steps: like Figure 2 As shown, the zinc-air battery gas-liquid-electro-coordinated control system of this embodiment includes a zinc-air battery or battery pack, an air supply unit, an electrolyte circulation unit, a charge-discharge control unit, a status acquisition module, a controller, and a data storage unit. The zinc-air battery can be a primary zinc-air battery, a rechargeable zinc-air battery, a three-electrode zinc-air battery, a zinc-air flow battery, or a zinc-air battery pack composed of multiple individual cells.

[0032] An air supply unit may include an air inlet, a filter, a fan, an air valve, an air passage, a carbon dioxide absorption or bypass structure, an air pressure drop sensor, a flow sensor, and a humidity or dew point sensor. For most zinc-air battery systems that use ambient air, the air supply unit focuses on providing a stable air supply and controlling the effects of carbon dioxide and air electrode humidity on flooding, drying, and carbonate accumulation; in oxygen-enriched air supply or closed-loop air passage implementations, the air supply unit may also include an oxygen concentration sensor or an oxygen-enriched air supply actuator.

[0033] The electrolyte circulation unit may include an electrolyte tank, circulation pump, heat exchanger, replenishment valve, drain valve, level sensor, conductivity sensor, pH sensor, temperature sensor, and turbidity sensor. Not all electrolyte signals are required; some sensors can be selected based on battery type and cost considerations during implementation.

[0034] The charge / discharge control unit may include a charger, a discharge load interface, a current controller, a voltage sampling circuit, an impedance measurement circuit, a protection circuit, and an optional three-electrode switching relay. Impedance measurement can be an online small-signal measurement or a simplified measurement performed within a rest or low-disturbance window; continuous acquisition of the complete impedance spectrum is not required in all scenarios.

[0035] The controller executes the method steps of this application and issues coordinated execution instructions to the air supply unit, electrolyte circulation unit, air electrode humidity adjustment unit, replenishment and drainage unit, charge and discharge control unit, and three-electrode switching unit. The data storage unit is used to save historical ZAF(k), RMV(k), CGB(k), ACI(k), and FBR(k) for subsequent maintenance and parameter correction.

[0036] Step S1: Constructing the zinc void state evidence frame: as follows Figure 3 As shown, anomalies in zinc-air batteries are often not explainable by a single sensor. A drop in terminal voltage may originate from increased load, insufficient air supply, flooding, drying out, air path blockage, or changes in electrolyte state. Therefore, the multi-source signals within the sampling window are first organized into a single piece of evidence. Within the sampling window k, the formula for the controller to construct the zinc-air state evidence frame is as follows: ; In the above formula, It represents a set of electrical signals, including voltage, current, charging / discharging state, and impedance information; It represents a set of airway signals, including airway pressure difference, airflow, carbon dioxide concentration, humidity, dew point, and oxygen concentration or oxygen-enriched air supply status based on the battery settings. It represents a set of electrolyte signals, including conductivity, pH, liquid level, temperature, pump status, replenishment / drainage status, and turbidity; It represents a set of thermal signals, including unit temperature, module temperature, gas path temperature, and electrolyte temperature; This represents the set of actuator states, including the states of pumps, valves, fans, air electrode humidity control units, charge / discharge controllers, and three-electrode switching units. Indicates sensor validity.

[0037] By grouping together "evidence seen at the same moment," we prevent the voltage, gas path pressure drop, electrolyte, and actuator status from being judged separately. For example, a drop in terminal voltage and an increase in air path pressure difference have different implications than a drop in terminal voltage but a normal gas path pressure difference and an increase in charging overpotential.

[0038] Sensor validity marking includes missing data marking, drift marking, mutation marking, and contradictory marking. When a sensor is judged to be unreliable, its data is not directly deleted. Instead, its weight is reduced in subsequent identification processes, conservative control is made an optional action, and actions are prohibited from being performed solely based on that sensor.

[0039] Step S2: Feature extraction and evidence quality assessment. Engineering features that can be used to identify instability modes are extracted from the zinc-empty state evidence frame ZAF(k). The extracted features include: 1. Impedance characteristics: including high-frequency impedance, low-frequency impedance, rate of change of impedance, and impedance recovery within a rest window; 2. Air pressure drop characteristics: including air channel pressure difference, the relationship between air flow rate and pressure difference, and the trend of pressure difference change over time; 3. Air supply response characteristics: including air flow rate, air channel pressure difference, terminal voltage response after load change, and selectable oxygen concentration or oxygen-enriched air supply status; 4. Electrolyte concentration characteristics: including conductivity, pH, liquid level, temperature, and replenishment / drainage state; 5. Polarization characteristics: including discharge voltage drop, charging overpotential, load step response, and temperature rise; 6. Actuator response characteristics: including whether the flow rate follows the change in pump speed, whether the pressure drop or flow rate in the air circuit changes after the valve is activated, and whether the differential pressure recovers after the fan is activated.

[0040] From Zinc Void State Evidence Frames to Engineering Features The calculation formula is: ; in, For the synchronous sampling window corresponding to the k-th window, Let m be the window synchronization feature extraction function corresponding to the m-th type of signal source. This includes amplitude, rate of change, recovery, consistency, and anomaly markers; to prevent erroneous actions caused by faulty sensors, evidence quality weights are set. Used to limit The credibility of evidence used in subsequent instability pattern identification corresponds to the generation of evidence quality weights. The formula is as follows: ; In the above formula, m represents a certain type of signal source, including electrical signals, gas signals, electrolyte signals, or thermal signals; This indicates the effectiveness of the sensor from that source; it decreases when there are missing measurements or when the sensor drifts. This indicates the stability of the source within the sampling window, decreasing when there are large abrupt changes or when it contradicts adjacent sensors; clip indicates that the output result range is set between 0 and 1; this evidence quality weight is subsequently used for instability mode recognition to prevent blindly trusting a certain abnormal sensor and performing actions that could lead to battery runaway.

[0041] Step S3: Instability mode vector generation, used to determine which zinc-air battery instability mode the current battery condition resembles most. The controller generates the instability mode vector: ; In the above formula, This indicates a risk of insufficient air supply. This indicates the risk of water damage to the air electrode. This indicates the risk of the air electrode drying out. This indicates a risk of carbonate blockage or carbonate accumulation. This indicates the risk of zinc electrode passivation. Indicates dendrite risk, Indicating sensor malfunction risk, the above components are based on the engineering features obtained in step S2. and the weight of evidence quality The h-th type of instability mode is identified by weighting, as shown in the following formula: ; In the above formula, h represents one of the following modes: insufficient air supply, flooding, dryness, carbonate blockage, zinc electrode passivation, dendrite formation, or sensor malfunction. This can be one or a combination of rule-based models, state estimation models, lookup table models, and classification models, where state(k) represents the current charging / discharging condition. Some examples are provided below corresponding to the above: When the voltage drops significantly after the load increases, and there is insufficient airflow or abnormal air pressure drop, but the air pressure drop does not indicate blockage, It can be increased; when the gas pressure drop increases, humidity or dew point is abnormal, or terminal voltage fluctuates and is accompanied by high-frequency impedance changes, It can rise; when humidity decreases, polarization increases, and abnormal temperature rise occurs, It can increase; when carbon dioxide exposure time is long, conductivity or pH shifts, and gas path pressure drop gradually increases, It can rise.

[0042] (2) During the charging phase, if the low-frequency impedance increases, the charging overpotential increases, and the capacity recovery is insufficient, The temperature can be increased; however, if the charging current density is high, there are sudden overpotential changes, abnormal temperature rise, and an increase in the number of cycles, It can be increased. The above example is used to illustrate the recognition logic and does not require all signals to be present simultaneously.

[0043] The significance of the instability mode vector lies in the fact that the system no longer makes a general judgment of "battery malfunction" or "insufficient power," but instead provides specific candidate causes, thereby better judging the battery's operational state. For the risks in this embodiment, the specific identification techniques are as follows: The identification methods for different risks are as follows: Insufficient air supply risk is identified based on the decrease in terminal voltage after load increase, insufficient air flow, abnormal air path pressure drop, or decrease in oxygen concentration in oxygen-enriched air supply or closed air channel implementations; Air electrode flooding risk is identified based on the increase in air path pressure drop, terminal voltage fluctuation, high-frequency impedance change, and abnormal humidity or dew point; Air electrode drying risk is identified based on the decrease in humidity, increase in polarization, abnormal temperature rise, and slow voltage recovery; Carbonate blockage risk or carbonate accumulation risk is identified based on carbon dioxide exposure, changes in electrolyte conductivity, changes in air path pressure drop, and historical operating time; Zinc electrode passivation risk is identified based on the increase in low-frequency impedance, increase in charging or discharging overpotential, abnormal electrolyte concentration characteristics, and insufficient capacity recovery; Dendrite risk is identified based on charging current density, low-frequency impedance change, sudden change in charging overpotential, temperature rise, electrolyte state, and historical cycle count.

[0044] Step S4: Generate gas-liquid-electric coupling constraints. During the operation of the battery, there are mutual influences between air supply, electrolyte circulation and charging / discharging current. A certain action may seem to improve the current problem, but it may introduce new risks. Therefore, corresponding coupling constraints are set. The controller generates gas-liquid-electric coupling constraints GCC(k) based on RMV(k) and the current charge / discharge conditions. The gas-liquid-electric coupling constraints include at least the following constraints: 1. When the risk of insufficient air supply is high, the air supply can be increased. However, if the risk of the air electrode drying out is also high, it is necessary to limit the increase in the drying airflow or to perform air electrode humidity compensation first. 2. When the risk of water flooding of the air electrode is high, short-term flow increase, drainage, reduction of electrolyte supply or cessation of recovery can be performed, but the amount of air supply required for the reaction on the air electrode side must not be lower than the lower discharge limit. 3. When the risk of carbonate blockage is high, carbon dioxide absorption bypass, gas path regeneration, conservative load reduction or maintenance prompts can be implemented, but it is not advisable to simply increase the fan speed to force the pressure drop for a long period of time. 4. When the risk of zinc electrode passivation is high, pause, electrolyte flow increase, electrolyte replacement or low-rate recovery can be performed, but the charging current must not be directly increased in an attempt to quickly restore capacity. 5. When the risk of dendrite formation is high, current reduction, pulse charging, pause, or electrolyte circulation enhancement can be performed, but the charging current density and temperature rise boundaries must not be exceeded. Risk action relationships can be implemented using gating functions, for example, when Reaching the air supply insufficiency threshold and When the dryness threshold has not been reached, increasing the air supply is permitted; when and When the corresponding threshold is reached, the air supply increment is limited to a smaller value and humidity compensation is triggered first; when When the dendrite risk threshold is reached, the charging current increment is limited and the system enters the safe charging boundary.

[0045] Additional note: To ensure that the entire risk action sequence described above can be executed by the controller, a unified gating function can be set for each type of instability mode. Then, the permitted actions are written into the action set Ah(k). For example, when the flooding gate is triggered, short-term gas path regeneration, liquid drainage, reduction of local electrolyte supply, or rest recovery are allowed, while maintaining the air supply at no less than the discharge lower limit; when the carbonate blockage gate is triggered, carbon dioxide absorption bypass, gas path regeneration, conservative load reduction, or maintenance prompts are allowed, while limiting long-term fan increments; when the passivation gate is triggered, rest, electrolyte flow increase, liquid replacement, or low-rate recovery are allowed, while prohibiting rapid capacity recovery by increasing the charging current.

[0046] When the risk of sensor malfunction is high, the controller will prioritize reducing the amplitude of the action and require confirmation through adjacent sensors, manual inspection, or conservative mode.

[0047] Based on the aforementioned risks, gas-liquid-electric coupling constraint The formula is as follows: in, It indicates the gas path boundary, including air flow rate, air channel pressure difference, carbon dioxide, humidity and pressure drop limits, and may include oxygen concentration limits under oxygen-enriched air supply implementation methods; It indicates the electrolyte boundary, including liquid level, conductivity, pH, temperature and pump speed limits; It indicates the electrical control boundary, including limits on charging and discharging current, voltage, rest time, and three-electrode switching time; This indicates the boundaries of manual authorization, protective interlocking, and conservative modes. This formula is used to describe the source of constraints, and in practice, it can be expressed through rule tables, parameter tables, or controller programs, and is not limited to this formula.

[0048] , and All of these can be expanded into a set of upper and lower limit constraints, such as limits for airflow, humidity, airway pressure drop, carbon dioxide, and selectable oxygen concentration. Liquid level, conductivity, pH, temperature, and pump speed limits constitute... The charging and discharging current, voltage, rest time, and three-electrode switching time constitute the... The corresponding formula is expressed as: In oxygen-enriched air supply systems or enclosed air ducts, the corresponding formula is: : In the above formula, Hgas represents the airflow rate, and Hgas represents the humidity of the air passage or the humidity corresponding to the dew point. This is for the air passage pressure drop. This refers to the concentration of carbon dioxide. Oxygen concentration; , , vpump represents liquid level, conductivity, temperature, and pump speed, respectively; Ichg, Idis, V, trest, and tsw represent charging current, discharging current, voltage, rest time, and three-electrode switching time, respectively; quantities with min or max subscripts are boundary values ​​given by the current instability mode vector and operating conditions. Oxygen concentration boundaries are only enabled when the corresponding detection or oxygen-enriched supply actuator is set.

[0049] Step S5: As Figure 4 As shown, the control gate zone generation process transforms the gas-hydraulic-electric coupling constraint into the current operational range of each actuator. By limiting the allowable range of each actuator within the current window, the safety of battery control is ensured. For any actuator... The formula for controlling the goalkeeping zone is as follows: ; In the above formula, and These represent the lower and upper limits of the actuator hardware, respectively; Indicates the execution volume of the previous window; and Indicates the maximum amount of change that adjacent windows are allowed to decrease or increase; and Indicates by The given current chemical, electrical, or gas-liquid boundary; The above formula stipulates that even if an action might aid recovery, it must simultaneously satisfy hardware limitations, rate-of-change limits, and zinc-air electrochemical boundaries. For example, when the dendrite risk is high, the upper limit of the charging current will be... Lower; when the risk of gas path drying is high, the upper limit of dry gas flow will also be lowered.

[0050] All actuators corresponding After combination, the resulting control gate zone is CGB(k).

[0051] Step S6: The cooperative execution instruction includes the target actuator, target change direction, target change magnitude, latest execution time, effective observation window, backoff condition, and cause code; under the control gate constraint, the controller selects the control vector corresponding to the cooperative execution instruction based on the improvement cost, side effect cost, and energy consumption or efficiency cost of the current main instability mode. , Indicates the first The actual selected control vector in each control window, u represents a candidate control vector, and the comprehensive cost of the candidate action is calculated as follows: ; In the above formula, This indicates that the comprehensive cost is calculated for the candidate control vector u in the k-th control window; This indicates the cost of improving the current main instability mode; the lower the cost, the more conducive it is to recovery. This indicates the side effects and costs, including the risks of drying up, flooding, dendrite formation, temperature rise, or exceeding the liquid level limit; Indicates energy consumption or efficiency cost; controller from Select the control vector with the lowest overall cost and that meets the conditions for manual verification. The candidate action set is represented as: ; The controller's formula is derived from at least one of the following: pre-set action templates, rule tables, model predictive control candidate sequences, maintenance strategy libraries, or manually authorized actions; ; Select candidate control vectors that meet the conditions of control gate zone and manual confirmation, and then... The package is formed into ACI(k).

[0052] For example, when the main instability mode is insufficient air supply and low risk of drying out, ACI(k) can increase the fan speed or air valve opening; when insufficient air supply is accompanied by risk of drying out, ACI(k) can first increase humidity or reduce the increase in dry airflow, and then slightly increase air supply; when the risk of dendrite formation is high, ACI(k) can reduce charging current, insert a rest window, or switch to pulse charging.

[0053] The preferred collaborative execution instructions should include at least: target actuator, target change direction, target change magnitude, latest execution time, effective observation window, rollback conditions, cause code, and whether manual confirmation is required. For actions involving three-electrode switching, significant load reduction, fluid replacement, or maintenance prompts, manual confirmation or authorization boundaries can be set.

[0054] Step S7: As Figure 5As shown, the operation receipt and parameter write-back are used to determine whether the collaborative execution instructions have truly improved the state of the zinc-air battery. The operation receipt FBR(k) preferably includes the amount of terminal voltage recovery, impedance change, gas path pressure drop recovery, electrolyte parameter recovery, temperature change, efficiency change, whether protection was triggered, whether manual intervention was required, and the reason for non-compliance, forming a valid record for the next adjustment. In a preferred embodiment, the recovery evaluation quantity is defined as follows: ; In the above formula, Indicates the amount of voltage recovery at the terminals; This indicates the amount of abnormal impedance drop; This indicates the abnormal drop in gas pressure. This indicates the degree to which the overall electrolyte condition index has recovered to the target range; The risk of temperature rise, liquid level exceeding the limit, protection triggering, or manual intervention after execution is obtained by weighting the temperature rise exceeding the limit flag, liquid level exceeding the limit flag, protection triggering flag, and manual intervention flag. to Weights are assigned; calculations are performed based on the observation window before and after the action. , , and The formula is as follows: ; ; ; ; ; In the above formula, pH and L represent the normalized state variables corresponding to electrolyte conductivity, acidity / alkalinity, and liquid level, respectively. to Indicates the state weight of the electrolyte; , , and Preferably, it is calculated from the average value or steady-state value of the observation window before the action and the observation window after the action; It is not a single liquid level recovery quantity, but a comprehensive evaluation quantity of the recovery of electrolyte state variables such as conductivity, pH, and liquid level to the target range. It can be obtained by weighting the flags such as temperature rise exceeding the limit, liquid level exceeding the limit, protection trigger, and manual intervention.

[0055] when When the set range is reached, and the gas path, electrolyte, and temperature do not exceed the limits, the controller increases the matching confidence between the current instability mode and the corresponding cooperative execution command; the matching confidence is the h-th type of instability mode within the k-th control window. With corresponding control vector The degree of effective historical matching between them The update relationship is represented as: ; in and To increase or decrease the step size for confidence level, To restore the evaluation threshold, This is an indicator function that takes 1 if the condition is true and 0 otherwise; when When the match confidence level is low, or when there is a breach of control or manual intervention, the controller lowers the confidence level of the match, shrinks the relevant control gate, and records the reason for the non-fulfillment.

[0056] Matching confidence Used to record the hth type of instability mode With control vector Historical validity between; when When the set range is reached, and the gas path, electrolyte, and temperature are all within limits, the controller increases the matching confidence between the current instability mode and the corresponding cooperative execution command; when When the match confidence level is low, or when there is a breach of control or manual intervention, the controller lowers the confidence level of the match, shrinks the relevant control gate, and records the reason for the non-fulfillment.

[0057] The significance of a run receipt is that the system does not provide suggestions all at once, but leaves evidence of "what was executed, whether it was restored, and why it failed".

[0058] The following simulation scenarios are provided to demonstrate the effectiveness of the control scheme corresponding to the technical solution of this invention. The simulation scenario is set as follows: taking a single rechargeable zinc-air battery as the object, the effective area of ​​the air electrode is... The electrolyte is 6 mol / L KOH, the ambient temperature is 25±2℃, and the discharge condition is 100mA / The charging condition is 40mA / Each threshold and normalization benchmark can be calibrated according to the specific battery pack capacity, air channel volume and electrolyte volume. The following values ​​are used to illustrate the self-consistent calculation of control logic and acknowledgment quantity.

[0059] (i) In a scenario with insufficient air supply, 120 seconds before the action, the voltage inside the observation window dropped from 1.18V to 1.04V, the air flow rate was 0.42L / min, the air channel pressure drop was 0.8kPa, the outlet relative humidity was 48%, the electrolyte conductivity was 0.61S / cm, and the pH was 13.6. The controller increased the fan duty cycle from 45% to 70% and the air valve opening from 50% to 72%. Within 120 seconds after the action, the air flow rate increased to 0.76L / min, the pressure drop was 1.1kPa, the terminal voltage recovered to 1.15V, and the outlet relative humidity remained between 46% and 50%, without triggering desiccation or exceeding the temperature limit. It is 0.68.

[0060] The explanation is as follows: After the discharge load increases, the controller detects a rapid drop in terminal voltage, indicating insufficient airflow or inadequate air supply, but the air pressure drop does not increase significantly, and the electrolyte condition is not abnormal. At this time, RMV(k) The system determined that the main problem was insufficient air supply.

[0061] If GCC(k) indicates a low risk of air electrode drying out, the controller can increase the fan speed or air valve opening within CGB(k) and set an observation window. If the terminal voltage recovers and the air path does not exceed the limits, the operation receipt confirms the action is effective; if the terminal voltage does not recover, the system will proceed to further check the air passage, air electrode status, or sensor malfunction.

[0062] (ii) In the water-flooded scenario of the air electrode, the pressure drop in the air channel before operation is 3.8 kPa, the air flow rate is 0.36 L / min, the relative humidity at the outlet is 96%, the terminal voltage fluctuates between 1.04 V and 1.10 V, and the high-frequency impedance changes from 22 mΩ. Rise to 34m The controller performs a 25-second short-term gas path regeneration, reducing the electrolyte circulation pump duty cycle from 60% to 45%, and sets a 60-second rest and recovery window. After the action, the pressure drop drops to 1.6 kPa, the air flow rate recovers to 0.68 L / min, the terminal voltage stabilizes between 1.14 V and 1.16 V, and the high-frequency impedance drops back to 25 mΩ. , It is 0.72.

[0063] Data Interpretation: When the gas pressure drop increases, the terminal voltage fluctuates, the high-frequency impedance changes, and is accompanied by abnormal humidity or dew point, the system increases... At this time, the controller should not simply increase the fan speed for a long period of time. Instead, it can choose to regenerate the gas path for a short time, reduce the local electrolyte supply, set a pause recovery or trigger the liquid drainage action, and check whether the gas path pressure drop has recovered in the observation window.

[0064] If the gas pressure drop decreases and the terminal voltage fluctuation decreases, the operation log will record the recovery from flooding; if the gas pressure drop remains high, it may record carbonate blockage, gas path blockage, or maintenance requirements.

[0065] (III) In the scenario where the air electrode is dry, the relative humidity at the outlet before operation is 28%, the battery surface temperature is 38.2℃, the terminal voltage is 1.07V, and the low-frequency impedance is 118mΩ. The air pressure drop is only 0.9 kPa; the controller limits the air supply increment to within 0.08 L / min, activates the humidity control unit to raise the inlet relative humidity to 55%, and reduces the discharge current density from 100 mA / min. Temporarily reduced to 80mA / Within 180 seconds after the operation, the terminal voltage recovered to 1.16V, the temperature dropped to 34.6℃, and the low-frequency impedance fell back to 92mΩ. The relative humidity at exports remained between 50% and 58%. It is 0.64.

[0066] Data Interpretation: When humidity decreases, polarization increases, temperature rises abnormally, and terminal voltage recovery is slow, the system improves... At this time, the controller can reduce the increase in drying airflow, activate the air electrode humidity adjustment unit, reduce the discharge current, or set a short pause. If there is also insufficient air supply, the controller will not increase the drying airflow indefinitely, but will slightly increase the air supply after air electrode humidity compensation.

[0067] (iv) Carbonate blockage scenario: The simulation was set with a carbon dioxide concentration of 800 ppm and continuous exposure for 48 hours. Before the action, the electrolyte conductivity decreased from 0.62 S / cm to 0.49 S / cm, the pH decreased from 13.7 to 13.25, and the air channel pressure drop slowly increased to 2.9 kPa. Increasing the fan duty cycle to 80% alone did not reduce the pressure drop to 2.7 kPa. The controller switched to the carbon dioxide absorption bypass, performed 60 seconds of gas path regeneration, reduced the discharge load to 80% of the rated value, and simultaneously triggered 5% electrolyte replenishment. Ten minutes after the action, the pressure drop decreased to 1.8 kPa, the conductivity recovered to 0.55 S / cm, the pH recovered to 13.43, and the terminal voltage recovered from 1.08 V to 1.13 V. It is 0.58.

[0068] Data Interpretation: When carbon dioxide exposure is high, electrolyte conductivity or pH continues to deviate, gas path pressure drop increases slowly and recovery after fan operation is insufficient, the system will increase... At this time, the coordinated execution commands may include switching the carbon dioxide absorption bypass, conservative load reduction, gas path regeneration, maintenance prompts, or fluid replenishment / drainage actions. In the case of severe blockage, the system can enter a conservative control range to avoid prolonged high fan speed operation.

[0069] (V) Dendrite Risk Scenario: In the dendrite risk scenario, before the action, the charging current density is 40 mA / cm², the charging voltage increases from 2.05V to 2.26V, the low-frequency impedance increases from 86mΩ to 132mΩ, the battery surface temperature rises to 37.5℃, and the historical cycle count is 80. The controller reduces the charging current density to 20 mA / cm², adopts a pulse strategy of 30s charging / 15s rest, and increases the electrolyte circulation pump duty cycle from 55% to 70%. Five minutes after the action, the charging voltage drops to 2.09V, the low-frequency impedance falls back to 106mΩ, and the temperature drops to 34.8℃. Short circuit protection or manual intervention is not triggered. It is 0.61.

[0070] Data Interpretation: During the charging phase, if the charging current density is high, there are low-frequency impedance changes, sudden overpotential changes, increased temperature rise, and a large number of historical cycles, the system performance will improve. The controller can reduce the charging current, insert a rest window, perform pulse charging, or enhance electrolyte circulation; if a three-electrode switching unit is configured, protective switching can be performed after the voltage, current, and electrolyte conditions are met.

[0071] The technical solution for controlling the battery pack is as follows: When this application is applied to a zinc-air battery pack, the differences between individual cells may arise from electrochemical aging, uneven distribution of air in the air branch, uneven flow of electrolyte in the electrolyte branch, local flooding, local drying, or sensor malfunction. The system preferably establishes local [control measures] for each individual cell or module. The individual cell voltage, individual cell temperature, individual cell impedance, air branch pressure difference, electrolyte branch flow rate, and valve status were compared.

[0072] If a module's voltage is low and the air branch voltage difference is abnormal, prioritize checking for insufficient air supply or air path blockage. If a module's temperature rise and charging overpotential are abnormal, prioritize checking for dendrite or passivation risks. If multiple modules are abnormal simultaneously, prioritize checking the overall air supply, electrolyte circulation, or sensor reference. Depending on the source of the difference, the system can adjust the air branch valves, electrolyte branch valves, module bypass, load reduction, or balanced operation.

[0073] In the process of controlling the battery, this invention adopts a conservative control mode. When there is a high risk of sensor malfunction, abnormal actuator feedback, severe gas path blockage, unreliable electrolyte state, temperature exceeding the limit, or protection interlock triggering, the controller no longer increases the action range, but instead reduces CGB(k) and performs conservative load reduction, pause, maintenance cycle, alarm, or manual confirmation.

[0074] For actions such as electrolyte replacement, forced gas path regeneration, three-electrode switching, significant load reduction, significant current increase, or maintenance, the system can require manual confirmation. This additional manual confirmation helps determine the specific status of the battery and avoids machine oversights.

[0075] The technical solution of this invention can be implemented in a dedicated battery management controller, or in an energy storage system controller, host computer, edge computing device, or cloud-edge collaborative control system. Instability mode recognition can employ rule-based models, lookup table models, state estimation models, model predictive control, lightweight classification models, or combinations thereof. Regardless of the implementation method used, it should revolve around the object chain ZAF(k), RMV(k), GCC(k), CGB(k), ACI(k), and FBR(k).

[0076] Therefore, this invention employs a gas-liquid-electro-coordinated control method for zinc-air batteries based on multiple physical evidence frames, which solves the problems in the prior art where the state judgment of zinc-air batteries only considers a single electrical signal, the control of the gas path and electrolyte is fragmented, the threshold start-stop is difficult to explain the cause of instability, the control actions lack boundary constraints, and the execution effect is difficult to write back.

[0077] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to preferred embodiments, those skilled in the art should understand that modifications or equivalent substitutions can still be made to the technical solutions of the present invention, and these modifications or equivalent substitutions cannot cause the modified technical solutions to deviate from the spirit and scope of the technical solutions of the present invention.

Claims

1. A gas-liquid-electric coordinated control method for zinc-air batteries based on multiple physical evidence frames, characterized in that: Includes the following steps: Step S1, in the corresponding sampling window In the process, electrical signals, gas path signals, electrolyte signals, thermal signals, and actuator status of the target zinc-air battery are collected to construct a zinc-air state evidence frame. ; Step S2: Based on the zinc-air state evidence frame, the impedance characteristics, gas path pressure drop characteristics, air supply response characteristics, electrolyte concentration proxy characteristics, and polarization characteristics corresponding to the target zinc-air battery are extracted using the window synchronous feature extraction and evidence quality weight calculation method. Step S3: Generate an instability mode vector based on the impedance characteristics, gas path pressure drop characteristics, air supply response characteristics, electrolyte concentration proxy characteristics, and polarization characteristics. The instability mode vector includes at least two of the following: insufficient air supply, water flooding of the air electrode, dryness of the air electrode, carbonate blockage, passivation of the zinc electrode, and dendrite risk. Step S4, based on the instability mode vector Based on the current charging and discharging conditions, corresponding gas-liquid-electric coupling constraints are generated. ; Step S5: Determine the allowable control range for each actuator based on the gas-liquid-electric coupling constraint, thus forming a control gate zone. ; Step S6, in the control goal zone Within the specified limits, collaborative execution instructions are formed. And then distribute it to the corresponding executor for execution; Step S7: Collect the execution receipt in the observation window after execution. It updates the instability mode identification parameters or control gate based on changes in terminal voltage, impedance, gas path pressure drop, electrolyte parameters or efficiency, and records whether the action is recovered, whether it exceeds the limit, and whether manual intervention is required.

2. The zinc-air battery gas-liquid-electric coordinated control method based on multiple physical evidence frames according to claim 1, characterized in that: In step S1, the electrical signals include, but are not limited to, individual cell or module voltage, current, charge / discharge status, estimated open-circuit voltage, and impedance measurement results; the gas path signals include, but are not limited to, air channel differential pressure, air flow rate, carbon dioxide concentration, humidity, dew point, oxygen concentration detection value, or oxygen-enriched air supply status; the electrolyte signals include, but are not limited to, electrolyte temperature, conductivity, pH, liquid level, circulation pump status, replenishment status, draining status, or turbidity; the thermal signals include, but are not limited to, battery cell temperature, module temperature, gas path temperature, or electrolyte temperature, providing evidence of the zinc-air state. Record electrical signals, gas path signals, electrolyte signals, thermal signals and actuator status within the same sampling window, and record sensor validity markers and stability parameters within the sampling window; Corresponding to the sampling window Internally, the controller constructs zinc void state evidence frames. as follows: ； In the above formula, It represents a set of electrical signals, including voltage, current, charging / discharging state, and impedance information; It represents a set of airway signals, including airway pressure difference, airflow, carbon dioxide concentration, humidity, dew point, and oxygen concentration or oxygen-enriched air supply status based on the battery settings. It represents a set of electrolyte signals, including conductivity, pH, liquid level, temperature, pump status, replenishment / drainage status, and turbidity; It represents a set of thermal signals, including unit temperature, module temperature, gas path temperature, and electrolyte temperature; This represents the set of actuator states, including the states of pumps, valves, fans, air electrode humidity control units, charge / discharge controllers, and three-electrode switching units. Sensor validity markers include missing data markers, drift markers, mutation markers, and contradictory markers.

3. The zinc-air battery gas-liquid-electric coordinated control method based on multiple physical evidence frames according to claim 2, characterized in that: The process of step S2 is as follows: (corresponding to...) Engineering features extracted for identifying instability modes include: impedance features, including high-frequency impedance, low-frequency impedance, impedance change rate, and impedance recovery within a rest window; air path pressure drop features, including air channel differential pressure, the relationship between air flow rate and differential pressure, and the trend of differential pressure change over time; air supply response features, including air flow rate, air channel differential pressure, terminal voltage response after load changes, and oxygen concentration or oxygen-enriched air supply status; electrolyte concentration proxy features, including conductivity, pH, liquid level, temperature, and replenishment / drainage status; polarization features, including discharge voltage drop magnitude, charging overpotential, load step response, and temperature rise; actuator response features, including whether the flow rate follows the pump speed change, whether the air path pressure drop or flow rate changes after valve action, and whether the differential pressure recovers after fan action; and the conversion of zinc-air state evidence frames into engineering features. The calculation formula is: ； in, For the synchronous sampling window corresponding to the k-th window, Let m be the window synchronization feature extraction function corresponding to the m-th type of signal source. This includes amplitude, rate of change, recovery, consistency, and anomaly markers; to prevent erroneous actions caused by faulty sensors, evidence quality weights are set. Used to limit The credibility of evidence used in subsequent instability pattern identification corresponds to the generation of evidence quality weights. The formula is as follows: ； In the above formula, m represents a certain type of signal source, including electrical signals, gas signals, electrolyte signals, or thermal signals; This indicates the effectiveness of the sensor from that source; it decreases when there are missing measurements or when the sensor drifts. This indicates the stability of the source within the sampling window, and is reduced when the abrupt change is too large or when it contradicts adjacent sensors; clip indicates that the output result range is set between 0 and 1; when the sensor is missing, drifting, abruptly changing, or contradictory, the weight of the corresponding signal source in the subsequent instability mode recognition is reduced, and the actuator action is prohibited from being directly amplified based solely on the abnormal signal.

4. The zinc-air battery gas-liquid-electric coordinated control method based on multiple physical evidence frames according to claim 3, characterized in that: In step S3, based on the engineering features extracted in step S2, the instability mode vector is generated as follows: ； In the above formula, This indicates a risk of insufficient air supply. This indicates the risk of water damage to the air electrode. This indicates the risk of the air electrode drying out. This indicates a risk of carbonate blockage or carbonate accumulation. This indicates the risk of zinc electrode passivation. Indicates dendrite risk, Indicating sensor malfunction risk, the above components are based on the engineering features obtained in step S2. and the weight of evidence quality The h-th type of instability mode is identified by weighting, as shown in the following formula: ； In the above formula, h represents one of the following modes: insufficient air supply, flooding, dryness, carbonate blockage, zinc electrode passivation, dendrite formation, or sensor malfunction. It can be one or a combination of rule-based models, state estimation models, lookup table models, and classification models, where state(k) represents the current charging and discharging condition.

5. The zinc-air battery gas-liquid-electric coordinated control method based on multiple physical evidence frames according to claim 4, characterized in that: In step S3, the identification methods for different risks are as follows: insufficient air supply risk is identified based on the drop in terminal voltage after the load increases, insufficient air flow, abnormal air pressure drop, or the drop in oxygen concentration in the oxygen-enriched air supply or closed air channel implementation; air electrode flooding risk is identified based on the increase in air pressure drop, terminal voltage fluctuation, high-frequency impedance change, and abnormal humidity or dew point. The risk of air electrode drying out is identified based on decreased humidity, increased polarization, abnormal temperature rise, and slow voltage recovery; the risk of carbonate blockage or carbonate accumulation is identified based on carbon dioxide exposure, changes in electrolyte conductivity, changes in gas path pressure drop, and historical operating time; the risk of zinc electrode passivation is identified based on increased low-frequency impedance, increased charging or discharging overpotential, abnormal electrolyte concentration characteristics, and insufficient capacity recovery; and the risk of dendrite formation is identified based on charging current density, changes in low-frequency impedance, sudden changes in charging overpotential, temperature rise, electrolyte state, and historical cycle count.

6. The zinc-air battery gas-liquid-electric coordinated control method based on multiple physical evidence frames according to claim 5, characterized in that: In step S4, according to Generates gas-liquid-electric coupling constraint with the current charge-discharge conditions. Gas-liquid-electric coupling constraint At least including: (1) When the risk of insufficient air supply is high, increase the air supply. However, if the risk of the air electrode drying out is also high, it is necessary to limit the increase in the drying airflow or to perform air electrode humidity compensation first. (2) When the risk of water flooding of the air electrode is high, short-term flow increase, drainage, reduction of electrolyte supply or rest and recovery shall be performed, but the amount of air supply required for the reaction on the air electrode side shall not be lower than the lower limit of discharge. (3) When the risk of carbonate blockage is high, carbon dioxide absorption bypass, gas path regeneration, conservative load reduction or maintenance prompts shall be implemented, but the pressure drop shall not be achieved by simply increasing the fan speed for a long period of time. (4) When the risk of zinc electrode passivation is high, the battery should be stopped, electrolyte flow increased, electrolyte replaced or low-rate recovery performed, but the charging current should not be directly increased in an attempt to quickly restore capacity. (5) When the dendrite risk is high, reduce current, pulse charge, pause or electrolyte circulation enhancement shall be performed, but the charging current density and temperature rise boundaries shall not be exceeded. Based on the aforementioned risks, gas-liquid-electric coupling constraint The formula is as follows: In the above formula, It indicates the gas path boundary, including airflow, air channel pressure difference, carbon dioxide, humidity and pressure drop limits, and includes oxygen concentration limits under the oxygen-enriched air supply implementation method; It indicates the electrolyte boundary, including liquid level, conductivity, pH, temperature and pump speed limits; It indicates the electrical control boundary, including limits on charging and discharging current, voltage, rest time, and three-electrode switching time; This indicates the boundaries of manual authorization, protective interlocking, and conservative modes.

7. The zinc-air battery gas-liquid-electric coordinated control method based on multiple physical evidence frames according to claim 6, characterized in that: In step S5, the process of generating the control gate zone is as follows: Actuators for regulating battery operation include circulation pumps, air valves, fans, air electrode humidity control units, replenishment valves, drain valves, charge / discharge controllers, or three-electrode switching units; in implementations where the battery is in an oxygen-enriched air supply or closed air passage, the actuator also includes an oxygen-enriched air supply actuator. For any actuator... The corresponding formula for controlling the goal belt is as follows: ； In the above formula, and These represent the lower and upper limits of the actuator hardware, respectively; Indicates the execution volume of the previous window; and Indicates the maximum amount of change that adjacent windows are allowed to decrease or increase; and Indicates by The given current chemical, electrical, or gas-liquid boundary; The control gate is set with hardware safety upper and lower limits, maximum change between adjacent control cycles, electrochemical risk boundary, gas path risk boundary, electrolyte risk boundary, and manual authorization boundary. When the instability mode vector shows sensor abnormality, actuator abnormality, severe gas path blockage, or unreliable electrolyte status, the control gate will shrink the corresponding actuator to the conservative control range. For any actuator, the permissible control range is jointly defined by the upper and lower limits of the actuator hardware, the maximum change in adjacent control cycles, and the current boundary given by the gas-liquid-electric coupling constraint.

8. The zinc-air battery gas-liquid-electric coordinated control method based on multiple physical evidence frames according to claim 7, characterized in that: In step S6, a cooperative execution instruction is generated. The process is as follows: The coordinated execution instructions include the target actuator, target change direction, target change magnitude, latest execution time, effective observation window, backoff condition, and cause code. Under the constraints of the control gate, the controller selects the control vector corresponding to the coordinated execution instructions based on the improvement cost, side effect cost, and energy consumption or efficiency cost of the current main instability mode. , Indicates the first The actual selected control vector in each control window, u represents a candidate control vector, and the comprehensive cost of the candidate action is calculated as follows: ； In the above formula, This indicates that the comprehensive cost is calculated for the candidate control vector u in the k-th control window; This indicates the cost of improving the current main instability mode; the lower the cost, the more conducive it is to recovery. This indicates the side effects and costs, including the risks of drying up, flooding, dendrite formation, temperature rise, or exceeding the liquid level limit; Indicates energy consumption or efficiency cost; controller from Select the control vector with the lowest overall cost and that meets the conditions for manual verification. The candidate action set is represented as: ； The controller's formula is derived from at least one of the following: pre-set action templates, rule tables, model predictive control candidate sequences, maintenance strategy libraries, or manually authorized actions; ； Select candidate control vectors that meet the conditions of control gate zone and manual confirmation, and then... The package is formed into ACI(k).

9. The zinc-air battery gas-liquid-electric coordinated control method based on multiple physical evidence frames according to claim 8, characterized in that: Step S7 is used to determine the execution effect of the cooperative execution instruction, and the corresponding formula is as follows: ； In the above formula, Indicates the amount of voltage recovery at the terminals; This indicates the amount of abnormal impedance drop; This indicates the abnormal drop in gas pressure. This indicates the degree to which the overall electrolyte condition index has recovered to the target range; The risk of temperature rise, liquid level exceeding the limit, protection triggering, or manual intervention after execution is obtained by weighting the temperature rise exceeding the limit flag, liquid level exceeding the limit flag, protection triggering flag, and manual intervention flag. to Weights are assigned; calculations are performed based on the observation window before and after the action. , , and The formula is as follows: ； ； ； ； ； In the above formula, pH and L represent the normalized state variables corresponding to electrolyte conductivity, acidity / alkalinity, and liquid level, respectively. to Indicates the state weight of the electrolyte; when When the set range is reached, and the gas path, electrolyte, and temperature do not exceed the limits, the controller increases the matching confidence between the current instability mode and the corresponding cooperative execution command; the matching confidence is the h-th type of instability mode within the k-th control window. With corresponding control vector Historical effective matching degree between The updated formula is: ； in and To increase or decrease the step size for confidence level, To restore the evaluation threshold, This is an indicator function that takes 1 if the condition is true and 0 otherwise; when When the match confidence level is low, or when there is a breach of control or manual intervention, the controller lowers the confidence level of the match, shrinks the relevant control gate, and records the reason for the non-fulfillment.

10. The zinc-air battery gas-liquid-electric coordinated control method based on multiple physical evidence frames according to claim 9, characterized in that: When the target zinc-air battery is a battery pack, the corresponding battery pack sets up a single cell or module consistency diagnosis step: determine the source of the single cell difference based on the single cell voltage, single cell temperature, single cell impedance, air branch pressure difference, electrolyte branch flow rate and valve status, and perform coordinated control on the air branch, electrolyte branch, module bypass, load reduction or equalization operation.

Images