Energy storage device energy management method and system based on cloud collaboration

Abstract

The application relates to the technical field of energy storage collaborative control, in particular to an energy management method and system for energy storage equipment based on cloud collaboration. The method comprises the following steps: acquiring multi-source operation data flow and cloud weather forecast information, analyzing a future period photovoltaic power fluctuation curve and liquid battery energy storage heat load estimation information based on the multi-source operation data flow and in combination with the cloud weather forecast information to obtain a photovoltaic-heat coupling situation set, analyzing heat variation characteristics of a liquid battery storage tank caused by a heat generation process of a liquid flow battery stack under the photovoltaic power fluctuation curve based on the photovoltaic-heat coupling situation set to obtain a time-segmented heat inertia information set, generating a buffer zone blending control instruction set suitable for instantaneous power fluctuation, dynamically planning a blending ratio time sequence of a cold buffer zone and a hot buffer zone, and outputting an energy storage equipment energy management log. The method effectively improves the operation safety of the liquid flow battery energy storage when coping with high-fluctuation photovoltaic access.

Description

Technical Field

[0001] This application relates to the field of energy storage collaborative control technology, and in particular to energy management methods and systems for energy storage devices based on cloud collaboration. Background Technology

[0002] In the integration of photovoltaic power generation and liquid battery energy storage, photovoltaic power generation is significantly affected by weather changes, and its output power fluctuates frequently and drastically. This fluctuation can cause uneven heat generation in the supporting energy storage equipment such as flow batteries during charging and discharging, thereby affecting the thermal management and overall operating efficiency of the energy storage system.

[0003] Existing energy management methods for energy storage devices typically involve real-time adjustments based on the current operating status or the use of fixed thermal management strategies. These methods fail to adequately consider the coupling characteristics between photovoltaic power fluctuations and the thermal dynamics of energy storage devices. In particular, under weather conditions such as cloudy skies and short periods of intense sunlight, the rapid rise, plateau maintenance, and sudden drop in photovoltaic power can trigger drastic changes in the heat generated by the flow battery stack. Since the storage tank has significant thermal inertia, its temperature response is delayed. Therefore, there is an urgent need for an energy management method that can integrate photovoltaic power prediction and energy storage thermal dynamics analysis to achieve synergistic optimization control of power fluctuations and thermal inertia. Summary of the Invention

[0004] This application provides a cloud-based collaborative energy management method and system for energy storage devices to solve the above-mentioned problems.

[0005] In a first aspect, this application provides a cloud-based collaborative energy storage device energy management method, the method comprising: The system acquires multi-source operational data streams and cloud-based weather forecast information. Based on these data streams and forecasts, it analyzes the photovoltaic power fluctuation curves and liquid battery energy storage heat load predictions for future cycles to obtain a photovoltaic-thermal coupling situation set. Based on this set, it analyzes the temperature change characteristics of the liquid battery tank caused by the heat generation process of the flow battery stack under the photovoltaic power fluctuation curve, obtaining a time-segmented thermal inertia information set. Based on this set, it generates a buffer mixing control command set adapted to instantaneous power fluctuations, dynamically plans the mixing ratio timing of the cold and hot buffer zones, and outputs the energy management log of the energy storage device.

[0006] Through the above technical solutions, an "electricity-thermal" coupled situation set was proactively constructed, solving the problem of lack of prediction of the correlation between fluctuating renewable energy and thermal disturbances in energy storage systems. Furthermore, through refined modeling, the dynamic time-varying law of thermal inertia of the storage tank was analyzed, generating a time-segmented thermal inertia information set, overcoming the shortcomings of vague understanding and extensive utilization of thermal buffer capacity in existing thermal management. Finally, based on accurate thermal inertia understanding, a model predictive control method was adopted to dynamically generate buffer mixing instructions, realizing proactive, smooth, and energy-saving guided management of the thermal state of the energy storage system, effectively improving the operational safety, economy, and lifespan of the flow battery energy storage system when dealing with highly fluctuating photovoltaic access.

[0007] Optionally, the multi-source operational data stream includes photovoltaic power sequences, photovoltaic device status parameters, and liquid battery device status parameters; the cloud-based weather forecast information includes irradiance prediction sequences and cloud cover prediction data; based on the photovoltaic power sequences and photovoltaic device status parameters, combined with the irradiance prediction sequences, the future periodic photovoltaic power change information is analyzed to obtain a basic power curve; based on the cloud cover prediction data, combined with the basic power curve, the power abrupt change time points and power change amplitudes caused by changes in cloud cover are analyzed to obtain power fluctuation characteristics; based on the power fluctuation characteristics, combined with the liquid battery device status parameters, the heat generation process of the flow battery stack under power changes is analyzed to obtain the liquid battery energy storage heat load prediction information; the heat load prediction information and the power fluctuation curve together constitute the photovoltaic-thermal coupling situation set.

[0008] Optionally, based on the cloud coverage prediction data, a time-series abrupt change detection algorithm is used to analyze the start and end times of the rise and fall of the coverage value to obtain cloud coverage change information; based on the cloud coverage change information, a steady-state power estimation algorithm is applied to analyze the photovoltaic power generation information after the end of the cloud change process to obtain the theoretical photovoltaic output level; the start and end times of cloud coverage change information are mapped to the basic power curve, and the decrease or increase process of the power in the basic power curve within the time period defined by the start and end times of cloud coverage is analyzed to obtain the power abrupt change time point and the power change trend; based on the theoretical photovoltaic output level, the power abrupt change time point, and the power change trend, a description set containing the abrupt change time sequence and change amplitude is constructed as the power fluctuation feature.

[0009] Optionally, based on the power mutation time point and the power change amplitude, the power ramp-up period, high-power plateau period, and power drop period are divided; real-time parameter identification technology is used to analyze the Joule heating process caused by the rapid increase of liquid battery current during the power ramp-up period, obtaining stack heat generation ramp-up information; steady-state thermal balance analysis method is used to analyze the steady-state heat generation process of the liquid battery stack during the high-power plateau period, obtaining stack steady-state heat generation information; heat dissipation dynamic characteristic analysis method is used to analyze the lag process of internal polarization heat dissipation caused by the rapid decrease of current during the power drop period, obtaining stack heat dissipation information; combining the stack heat generation ramp-up information, the stack steady-state heat generation information, and the stack heat dissipation information, the liquid battery energy storage heat load prediction information is constructed.

[0010] Optionally, based on the fuel cell stack heat generation ramp-up information, the delayed process of rapid accumulation of Joule heat in the fuel cell stack and its conduction to the storage tank via the heat exchanger during the power ramp-up period is analyzed to obtain the storage tank temperature rise lag characteristic; based on the fuel cell stack steady-state heat generation information, the accumulation process of continuous input of fuel cell stack heat causing heat diffusion from the edge to the center of the electrolyte in the storage tank during the high-power plateau period is analyzed to obtain the storage tank heat absorption accumulation characteristic; based on the fuel cell stack heat dissipation information, the process of reduced fuel cell stack heat generation but the storage tank maintaining its temperature due to thermal inertia and slowly dissipating heat to the environment through the circulation pipeline during the power drop period is analyzed to obtain the storage tank cooling inertia characteristic; combining the storage tank temperature rise lag characteristic, the storage tank heat absorption accumulation characteristic, and the storage tank cooling inertia characteristic, a set describing the response delay and change law of the storage tank temperature under different power periods is constructed as the time-segmented thermal inertia information set.

[0011] Optionally, based on the heat generation ramp-up information of the fuel cell stack, the instantaneous process of rapid generation and accumulation of Joule heat in the fuel cell stack body during the power ramp-up period is analyzed using internal resistance identification technology to obtain the core heat generation information of the fuel cell stack; based on the core heat generation information of the fuel cell stack, the diffusion process of heat from the core of the fuel cell stack to the circulating electrolyte through the heat exchanger pipeline is analyzed according to the heat transfer parameters of the liquid battery heat exchanger to obtain the pipeline heat delay information; based on the pipeline heat delay information, the diffusion delay information of heat in the tank space during the heat convection process of the electrolyte carrying heat flowing from the pipeline inlet to the tank interior and mixing with the original electrolyte in the tank is analyzed according to the flow rate data of the tank circulation inlet to obtain the temperature rise lag characteristics of the tank.

[0012] Optionally, based on the heat dissipation information of the fuel cell stack and combined with the power drop period, the starting moment of the rapid decrease in heat generation of the fuel cell stack is determined, and the inertial process of the overall temperature drop of the storage tank significantly lagging behind the decrease in heat generation of the fuel cell stack due to the constraints of electrolyte heat capacity and circulation flow rate from the starting moment is analyzed. The inertial process is characterized by a slow redistribution of the temperature gradient inside the storage tank, and the rate of decrease of the average temperature of the storage tank over time is limited by both the heat capacity of the storage tank structure and the circulation flow rate. Based on the inertial process, by calculating the functional relationship between the heat dissipation of the pipeline and the ambient temperature difference, the temperature decay rate of the high-temperature electrolyte in the storage tank during the circulating heat exchange process is determined, and the cooling inertial characteristics of the storage tank describing the slow temperature drop characteristics of the storage tank after the power drop are obtained.

[0013] Optionally, based on the lag characteristics of the tank temperature rise and combined with the photovoltaic power fluctuation curve, the buffer mixing operation required in advance to offset the delayed process of heat transfer from the stack to the tank before the start of the power ramp-up period is analyzed, resulting in a preheating mixing ratio sequence; based on the heat absorption accumulation characteristics of the tank and combined with the high-power plateau period, the buffer mixing ratio that needs to be dynamically adjusted to balance the continuous heat generation of the stack and the heat diffusion of the tank during the high-power plateau period is analyzed, resulting in a plateau period mixing adjustment sequence; based on the cooling inertia characteristics of the tank and combined with the power drop period, the buffer mixing operation that needs to be delayed after the start of the power drop period to utilize the residual heat of the tank and slow down the rate of temperature drop is analyzed, resulting in a lag mixing ratio sequence; combining the preheating mixing ratio sequence, the plateau period mixing adjustment sequence, and the lag mixing ratio sequence, they are arranged and smoothed in chronological order to generate the buffer mixing control instruction set for dynamically planning the timing of the hot and cold buffer mixing ratios.

[0014] Optionally, based on the preheating blending ratio sequence, the analysis is performed as an operation process to compensate for heat transfer delay by increasing the blending ratio of the thermal buffer before the stack heat generation increases, resulting in a preheating control sequence for the power ramp-up period; based on the plateau period blending adjustment sequence, the analysis is performed as an operation process to match the heat diffusion rate inside the storage tank by dynamically balancing the blending ratio of the hot and cold buffers during the stack heat generation steady-state period, resulting in a dynamic adjustment sequence for the high-power plateau period; based on the lag blending ratio sequence, the analysis is performed as an operation process to utilize the residual heat of the storage tank and slow down the temperature drop rate by delaying the reduction of the thermal buffer blending ratio after the stack heat generation decreases, resulting in a lag control sequence for the power drop period; the preheating control sequence, the dynamic adjustment sequence, and the lag control sequence are used as control variable sequences in the prediction time domain, and rolling optimization and smooth transition processing are performed to obtain the blending ratio sequence.

[0015] Secondly, this application provides a cloud-based collaborative energy storage device energy management system, the system comprising: The coupling situation module is used to acquire multi-source operational data streams and cloud-based weather forecast information. Based on the multi-source operational data streams and the cloud-based weather forecast information, it analyzes the photovoltaic power fluctuation curve and liquid battery energy storage heat load prediction information for future periods to obtain a photovoltaic-thermal coupling situation set. The segmented analysis module is used to analyze the temperature change characteristics of the liquid battery tank caused by the heat generation process of the flow battery stack under the photovoltaic power fluctuation curve based on the photovoltaic-thermal coupling situation set to obtain a time-segmented thermal inertia information set. The dynamic command module is used to generate a buffer mixing control command set adapted to instantaneous power fluctuations based on the time-segmented thermal inertia information set, dynamically plan the mixing ratio timing of the cold buffer and the hot buffer, and output the energy management log of the energy storage device. Attached Figure Description

[0016] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0017] Figure 1 This is a schematic diagram of an application scenario provided in an embodiment of this application; Figure 2 A flowchart illustrating a cloud-based collaborative energy storage device energy management method provided in an embodiment of this application; Figure 3 This is a schematic diagram of the structure of a cloud-based collaborative energy storage device energy management system provided in one embodiment of this application. Detailed Implementation

[0018] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without creative effort are within the scope of protection of this application.

[0019] Furthermore, the term "and / or" in this article is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, the character " / " in this article, unless otherwise specified, generally indicates that the preceding and following related objects have an "or" relationship.

[0020] The embodiments of this application will now be described in further detail with reference to the accompanying drawings.

[0021] In cloud-based collaborative energy storage device energy management scenarios, photovoltaic power fluctuations lead to uneven heat generation in flow batteries. Existing methods ignore the coupling characteristics between the two and the thermal inertia of the storage tank, which can easily cause overheating and performance degradation. There is an urgent need for a management solution that can coordinate power prediction and thermal dynamics.

[0022] Based on this, this application provides a cloud-based collaborative energy management method and system for energy storage devices. In response to the thermal disturbance problem of liquid battery energy storage caused by high-fluctuation photovoltaic access, the method constructs an electrothermal coupling situation set, performs refined analysis of the dynamic law of thermal inertia of the storage tank, and dynamically generates buffer mixing instructions based on model predictive control, thereby realizing active and smooth control of the thermal state.

[0023] Figure 1 This application provides an illustration of an application scenario. In a cloud-based collaborative energy storage device energy management scenario, the method provided in this application is applied to accurately analyze the system's thermal buffering capacity through electrothermal coupling situation prediction and thermal inertia dynamic modeling. Based on model predictive control, real-time mixing commands are generated, thereby achieving proactive and smooth regulation of the thermal state of the energy storage system.

[0024] Specifically, the method provided in this application can be applied to any server, which interacts with the sensor array and the cloud-based meteorological service platform to obtain multi-source operational data streams provided by the sensor array and cloud-based weather forecast information provided by the cloud-based meteorological service platform. It analyzes the dynamic time-varying law of the thermal inertia of the storage tank and outputs the energy management log of the energy storage device to the energy storage operation and maintenance management personnel, thereby improving the operational safety, economy and lifespan of the flow battery energy storage system when dealing with high-fluctuation photovoltaic access.

[0025] For specific implementation details, please refer to the following examples.

[0026] Figure 2 This is a flowchart illustrating a cloud-based collaborative energy management method for energy storage devices, provided as an embodiment of this application. The method of this embodiment can be applied to servers in the above scenarios. For example... Figure 2 As shown, the method includes: S201. Obtain multi-source operation data streams and cloud weather forecast information. Based on the multi-source operation data streams and cloud weather forecast information, analyze the photovoltaic power fluctuation curve and liquid battery energy storage heat load prediction information for future cycles to obtain the photovoltaic-thermal coupling situation set.

[0027] Multi-source operational data streams can be a set of parameters reflecting the real-time operating status of energy storage devices, with the sensor arrays built into the energy storage devices serving as the data source. Cloud-based weather forecast information can be a set of meteorological data for a specific future period published by a cloud-based meteorological service platform, with the cloud-based meteorological service platform serving as the data source. Photovoltaic power fluctuation curves can be a quantitative depiction of the fluctuation trend of photovoltaic power generation due to weather changes (such as cloud movement) in the future period (e.g., power drops sharply by 30% from its peak and then quickly recovers). Liquid battery energy storage heat load prediction information can be a prediction of the heat generated by the battery during charging and discharging due to internal resistance and electrochemical reactions in the future period and its impact on the thermal balance of the electrolyte system. Photovoltaic-thermal coupling status set can be a correlated dataset that integrates time-series power prediction and corresponding heat load prediction.

[0028] Specifically, in scenarios where flow battery energy storage systems operate in conjunction with fluctuating photovoltaic power sources, the randomness and intermittency of photovoltaic output (such as a sudden drop in power caused by a passing cloud) not only directly affect the power balance of the power grid, but its rapid fluctuations also cause frequent jumps in the operating point of the battery stack, resulting in drastic changes in the heat generation rate. This "electric-thermal" coupling effect will significantly affect the uniformity of temperature distribution and chemical stability of the electrolyte in the storage tank. Existing thermal management strategies based on static or delayed responses of the current state are difficult to cope with such forward-looking and correlated disturbances, which can easily cause local overheating of the storage tank or excessive overall temperature difference, thereby accelerating electrolyte aging and reducing efficiency and lifespan.

[0029] S202. Based on the photovoltaic-thermal coupling situation set, analyze the temperature change characteristics of the liquid battery tank caused by the heat generation process of the flow battery stack under the photovoltaic power fluctuation curve, and obtain a time-segmented thermal inertia information set. The flow battery stack can be the core component of the flow battery to realize the interconversion of chemical energy and electrical energy. The temperature change characteristics of the liquid battery tank can be the characteristics of the temperature change of the electrolyte solution in the liquid battery tank over time. The time-segmented thermal inertia information set can be the information set reflecting the temperature change inertia of the tank after analyzing the temperature change characteristics of the liquid battery tank at a set time interval (e.g., every 2 hours as a time period).

[0030] Specifically, although flow battery systems possess significant thermal inertia due to their large electrolyte volume, this inertia is not constant. It is influenced by the electrolyte flow rate, initial temperature distribution, environmental heat dissipation conditions, and most critically, the transient heat generation power of the battery stack. Without accurately understanding how this thermal inertia evolves over time under different future power fluctuation scenarios, any thermal management strategy will be either blind or conservative. For example, during periods of stable photovoltaic power, the system generates little heat, and thermal inertia may manifest as slow cooling; while during periods of rapid power increase, heat generation surges, and thermal inertia manifests as a buffer against temperature rise.

[0031] S203. Based on the time-segmented thermal inertia information set, generate a buffer mixing control instruction set that adapts to instantaneous power fluctuations, dynamically plan the mixing ratio timing of the cold buffer and the hot buffer, and output the energy management log of the energy storage device.

[0032] The buffer mixing control instruction set can be a set of instructions used to control the mixing operation of electrolyte solutions in the cold and hot buffers of a flow battery system. The mixing ratio timing of the cold and hot buffers can be a planned arrangement of the mixing ratio of electrolyte solutions in the cold and hot buffers at different time points. The energy management log of the energy storage device can be a document recording the entire energy management process of the energy storage device.

[0033] Specifically, existing flow battery thermal management often employs simple threshold-triggered cooling (such as activating the cooler when the temperature at a certain point exceeds 30°C). This method is slow to respond, abrupt, and energy-intensive. It cannot utilize the system's own thermal inertia and is prone to causing unnecessary temperature gradients within the storage tank, affecting the consistency of stack performance. The solution is to transform passive response into active "guidance," which involves using time-segmented thermal inertia information sets and precise understanding of the evolution of the system's own thermal inertia to proactively design intervention strategies to "shape" the thermal state.

[0034] The method provided in this embodiment proactively constructs an "electric-thermal" coupled situation set, solving the problem of lack of prediction of the correlation between fluctuating renewable energy and thermal disturbances in energy storage systems. Furthermore, through refined modeling, the dynamic time-varying law of the thermal inertia of the storage tank is analyzed, generating a time-segmented thermal inertia information set, overcoming the shortcomings of vague understanding and extensive utilization of thermal buffer capacity in existing thermal management. Finally, based on accurate thermal inertia understanding, a model predictive control method is used to dynamically generate buffer mixing instructions, realizing proactive, smooth, and energy-saving guided management of the thermal state of the energy storage system, effectively improving the operational safety, economy, and lifespan of the flow battery energy storage system when dealing with highly fluctuating photovoltaic access.

[0035] In some embodiments, the multi-source operational data stream includes photovoltaic power sequences, photovoltaic device status parameters, and liquid battery device status parameters; cloud-based weather forecast information includes irradiance prediction sequences and cloud cover prediction data; based on the photovoltaic power sequences and photovoltaic device status parameters, combined with the irradiance prediction sequences, the changes in photovoltaic power over future periods are analyzed to obtain a basic power curve; based on the cloud cover prediction data, combined with the basic power curve, the time points and power change amplitudes of abrupt changes in the basic power curve caused by changes in cloud cover are analyzed to obtain power fluctuation characteristics; based on the power fluctuation characteristics, combined with the liquid battery device status parameters, the heat generation process of the flow battery stack under power changes is analyzed to obtain liquid battery energy storage heat load prediction information; the heat load prediction information and the power fluctuation curve together constitute a photovoltaic-thermal coupling situation set.

[0036] A photovoltaic (PV) power sequence can be a sequence of power values ​​collected by real-time monitoring equipment and arranged in chronological order. PV equipment status parameters can be a set of physical quantities reflecting the current operating status of PV modules. Liquid battery equipment status parameters can be a set of parameters reflecting the operating status of a flow battery system. A solar irradiance prediction sequence can be a time-series prediction of solar irradiance over a specific future period. Cloud cover prediction data can be predictions of the degree to which clouds obstruct sunlight over a specific future period. A baseline power curve can be a smoothed prediction curve of future PV power generation under ideal solar irradiance trends, excluding sudden cloud interference. Power fluctuation characteristics can be a set of descriptions of the periods, directions, and magnitudes of rapid increases and decreases in PV power caused by meteorological factors such as cloud movement.

[0037] Specifically, photovoltaic power generation is characterized by significant intermittency and volatility, especially affected by cloud movement. Its output power can change drastically in a short period of time (e.g., power drops by 70% within minutes). This volatility directly leads to rapid changes in the operating current of the associated flow battery energy storage system. These current changes cause drastic fluctuations in the heat generated by the internal resistance of the battery stack (Joule heating), which in turn affects the temperature of the electrolyte in the storage tank through the heat exchanger. If this chain of fluctuations between light, electricity, and heat cannot be accurately predicted in advance, the thermal management of the energy storage system will be in a passive response state. Delayed temperature control may lead to local overheating of the electrolyte, affecting battery life, or excessive cooling, resulting in energy waste. To address the aforementioned issues: Real-time acquisition of local photovoltaic power sequences (such as the minute-by-minute power output data stream from a photovoltaic array inverter) and photovoltaic equipment status parameters (such as module backsheet temperature) is performed. Simultaneously, a cloud-based meteorological service platform is subscribed to a forecast sequence of irradiance and cloud cover (e.g., coverage will rapidly increase from 30% to 80% in the next 30-40 minutes) via API. First, a combined power prediction model trained based on physical mechanisms and historical data is applied, using the measured photovoltaic power sequence, equipment status, and irradiance prediction sequence as inputs to generate a basic power curve reflecting ideal weather trends. Subsequently, targeting the primary cause of power abrupt changes—cloud variations—a time-series abrupt change detection algorithm (such as a method based on sliding windows and statistical criteria) is used to analyze the cloud cover prediction data provided by the cloud, accurately identifying the precise time points when clouds begin to thicken and dissipate (e.g., identifying two abrupt change points at the 32nd and 48th minutes). Next, these two abrupt change points are mapped onto the aforementioned basic power curve, and a steady-state power estimation algorithm (considering the theoretical irradiance attenuation model under complete cloud cover) is used to calculate the expected decrease in photovoltaic power during cloud cover (e.g., from a theoretical value of 1.15MW to 0.45MW) and the recovery rate, thereby constructing a power fluctuation characteristic description set that includes the specific abrupt change time, direction of change (sudden drop / sudden rise), and magnitude of change. Finally, this power fluctuation characteristic is input into the electrothermal coupling dynamic model of the flow battery. This model uses battery device state parameters (such as the equivalent internal resistance curve at a specific SOC, electrolyte specific heat capacity, and flow rate setpoint) as boundary conditions, uses real-time parameter identification technology to correct model parameters online, and simulates and analyzes in segments: during the power ramp-up period, the rapid accumulation of Joule heat due to increased current; during the high-power plateau period, the quasi-steady-state balance process of heat generation and dissipation; and during the power drop period, the hysteresis process of polarization heat dissipation after the current decreases. Through this segmented and refined simulation, the predicted sequence of heat generation power of the liquid battery stack in the corresponding future time period (i.e., heat load prediction information) is output. Finally, this heat load sequence and the power fluctuation curve are aligned and bundled on the time axis to form a complete "photovoltaic-thermal coupling situation set" to describe the coupling relationship between "electrical energy fluctuation" and "heat load response" in the future.

[0038] The method provided in this embodiment constructs a photovoltaic-thermal coupling situation set that integrates high-precision weather forecasting, power fluctuation analysis, and electrothermal coupling modeling. This provides an accurate and forward-looking data foundation for the advanced thermal inertia analysis and dynamic temperature control decision-making of subsequent energy storage systems. This effectively overcomes the shortcomings of existing methods in responding to photovoltaic power mutations and battery thermal load changes with lag, and improves the operational safety and energy management economy of flow battery energy storage systems in scenarios with strong photovoltaic access fluctuations.

[0039] In some embodiments, based on cloud cover prediction data, a time-series abrupt change detection algorithm is used to analyze the start and end times of rises and falls in cloud cover values ​​to obtain cloud cover change information. Based on cloud cover change information, a steady-state power estimation algorithm is applied to analyze photovoltaic power generation information after the end of the cloud cover change process to obtain the theoretical photovoltaic output level. The start and end times of cloud cover change information are mapped to the basic power curve, and the decrease or increase process of power in the basic power curve within the time period defined by the start and end times of cloud cover is analyzed to obtain the power abrupt change time point and power change trend. Based on the theoretical photovoltaic output level, the power abrupt change time point, and the power change trend, a description set containing the abrupt change time sequence and change amplitude is constructed as a power fluctuation feature.

[0040] A time-series abrupt change detection algorithm can be an analytical algorithm used to identify key nodes in time-series data where numerical values ​​undergo significant increases or decreases. Cloud cover change information can be a set of information on the start and end times of cloud cover value changes, as well as the magnitude of these changes. A steady-state power estimation algorithm can be an algorithm used to analyze the photovoltaic power output level after the system reaches a steady state. The theoretical photovoltaic output level can be the stable power output value that the photovoltaic equipment should achieve after the cloud cover change process ends. The power abrupt change time point can be the specific moment when the base power curve undergoes a significant power change due to cloud cover changes. The power change trend can be the direction of change of the base power curve within the time period defined by the cloud cover change, such as a continuous increase, a continuous decrease, or a fluctuating decrease.

[0041] Specifically, photovoltaic power generation is highly dependent on sunlight. Rapid cloud movement and changes can cause power fluctuations on a minute or even second level. This poses a huge challenge to the energy and thermal management of coupled flow battery energy storage systems. For example, rapid cloud passage may cause photovoltaic power to drop sharply from its peak within minutes, changing the stack's operating current and heat generation rate. If the energy storage management system operates solely based on smoothed prediction curves, it cannot respond to changes in heat load in a timely manner, which may cause localized high temperatures or thermal stress concentration in the storage tank, affecting battery life and system safety. To address the aforementioned issues: First, a time-series abrupt change detection algorithm (such as a sliding window-based statistical test) is applied to analyze the cloud cover percentage sequence point by point, identifying the turning points where the values ​​significantly increase or decrease. For example, the algorithm might detect that the cloud cover starts at a stable 10%, then rapidly increases after a certain time point (e.g., 10:05), reaching 60% at another time point (e.g., 10:20) and then stabilizing. This accurately extracts the "starting time" and "ending time" of this cloud thickening, forming structured "cloud cover change information." Next, for the period after the cloud cover change ends, steady-state analysis is applied... The power estimation algorithm, based on the physical characteristics of the photovoltaic module (such as rated power and conversion efficiency) and the predicted steady-state irradiance at that moment, calculates the "theoretical photovoltaic output level" (e.g., 350kW) that the photovoltaic array is expected to recover to if there is no new cloud interference. This value serves as a key benchmark for assessing the power fluctuation amplitude. Then, a crucial mapping and trend analysis is performed: the cloud "start and end times" obtained in the previous step are mapped onto the time axis of the "baseline power curve," which represents a smooth change trend. The analysis focuses on the power response of the baseline curve within the time window defined by these two times (e.g., 10:05 to 10:20). By analyzing the slope and shape of the power curve within this window, the "power mutation time point" (i.e., the moment when the power begins to deviate significantly from the original trend, possibly slightly delayed from the cloud start time) and the "power change trend" (e.g., a power drop of 200kW within 5 minutes, showing a rapid-then-slow decline) caused by this event can be determined. Ultimately, the system integrates and encapsulates all the quantitative elements of this event—including the time of abrupt change (e.g., 10:07), the description of the trend (a decrease of 200kW in 5 minutes), and the theoretical steady-state value (350kW)—into a complete set of "power fluctuation characteristics" that describes a single or multiple fluctuation events. This transforms the vague descriptions in weather forecasts into precise blueprints of instructions that the energy management system can directly understand and use for forward-looking control.

[0042] The method provided in this embodiment enables refined and structured prediction of specific fluctuation events in photovoltaic power caused by sudden changes in cloud cover, generating high-precision power fluctuation characteristics. This provides a key and reliable input for accurately assessing the thermal behavior of flow batteries under dynamic power and implementing forward-looking thermal management, effectively improving the coordinated control capability and operational safety of energy storage systems in response to intermittent renewable energy power fluctuations.

[0043] In some embodiments, based on the power mutation time point and the power change amplitude, the power ramp-up period, high-power plateau period, and power drop period are divided. Real-time parameter identification technology is used to analyze the Joule heating process caused by the rapid increase in liquid battery current during the power ramp-up period, obtaining stack heat generation ramp-up information. Steady-state thermal balance analysis is used to analyze the steady-state heat generation process of the liquid battery stack during the high-power plateau period, obtaining stack steady-state heat generation information. Dynamic heat dissipation characteristic analysis is used to analyze the lag process of internal polarization heat dissipation caused by the rapid decrease in current during the power drop period, obtaining stack heat dissipation information. By integrating the stack heat generation ramp-up information, stack steady-state heat generation information, and stack heat dissipation information, liquid battery energy storage heat load prediction information is constructed.

[0044] The power ramp-up period can be the time when photovoltaic power rapidly increases from a sudden inflection point to a stable level. The high-power plateau period can be the period when photovoltaic power continues to operate after reaching a stable level. The power drop period can be the period when photovoltaic power rapidly decreases from a stable level. Real-time parameter identification technology can be a technology that dynamically identifies the correlation between equipment operating parameters and heat generation based on real-time operating data. Steady-state thermal balance analysis can be a method to calculate the steady-state heat generation level by analyzing the heat exchange law when the equipment's heat generation and dissipation reach a balanced state. The dynamic heat dissipation characteristic analysis method can be a method used to analyze the hysteresis characteristics of the internal residual heat dissipation process after the equipment's heat generation stops or decreases. Stack heat generation ramp-up information can be information characterizing the change in the heat generation rate and heat generation of the flow battery stack over time during the power ramp-up period. Stack steady-state heat generation information can be information characterizing the stability of the stack's heat generation rate and the accumulation law of heat generation during the high-power plateau period. Stack heat dissipation information can be information characterizing the stack polarization heat dissipation rate and dissipation time during the power drop period.

[0045] Specifically, in cloud-based collaborative energy management scenarios for energy storage devices, photovoltaic power fluctuations can cause drastic changes in the current of flow batteries. The heat generation principles, rates, and heat dissipation patterns vary significantly across different time periods. Using a uniform heat load calculation method can lead to prediction errors, affecting the targeting of thermal inertia analysis and buffer zone mixing control commands, potentially causing problems such as temperature runaway and shortened lifespan of the energy storage device. To address these issues, we begin by analyzing the upstream "power fluctuation characteristics." This characteristic indicates, for example, that the power will begin to climb at, for example, 10:15 AM, with an amplitude of, for example, 150kW. Based on this, we finely divide the future time period into: the power ramp-up period (e.g., 10:15-1...). During the power ramp-up period (e.g., 0:25), the high-power plateau period (e.g., 10:25-13:45), and the power drop period (e.g., 13:45-14:00), we apply real-time parameter identification technology. Specifically, through a recursive least squares algorithm deployed on the high-frequency data acquisition unit, we fit the transient relationship between the stack terminal voltage and current (e.g., sampling one set per second) online, track and output the internal resistance value in real time (e.g., during the process of the current surging from 50A to 200A, the internal resistance increases nonlinearly from 5 milliohms to 7 milliohms), and immediately multiply the square of the current at this moment by the dynamic internal resistance value, thereby calculating the Joule heat generation power ramp-up curve second by second, forming the stack heat generation ramp-up information. After entering the high-power plateau period, a steady-state thermal balance analysis method is employed. Specifically, the steady-state internal resistance value identified at the end of the ramp-up period (e.g., stabilizing at 7.2 milliohms) is used, combined with the constant output current during the plateau period (e.g., maintaining 200A), to calculate the stable Joule thermal power (approximately 2880W). Simultaneously, this thermal power is used as input, substituted into the steady-state thermal network equation, which includes parameters such as the specific heat capacity of the fuel cell stack materials, the heat transfer coefficient of the heat exchanger, and the convective heat dissipation area, to solve for the steady-state temperature at which the fuel cell stack reaches thermal equilibrium under a given ambient temperature (e.g., 25 degrees Celsius). This yields complete information on the fuel cell stack's steady-state heat generation. For the power drop period, a dynamic heat dissipation characteristic analysis method is used, specifically establishing a first-order thermal capacity-thermal resistance lumped parameter... In the numerical model, when a sudden drop in current command is detected in a short period of time (e.g., from 200A to 80A within 5 minutes), the step reduction in input power does not immediately cause the model temperature to drop. We convert the polarization voltage component that persists after the current decrease into a residual heat source, and use the model to simulate and calculate the exponential decay process of this residual heat dissipating to the environment through thermal resistance (mainly depending on the electrolyte flow rate, such as 20 liters per minute). This yields information on the heat dissipation of the stack, characterizing the lag between temperature decrease and power decrease. Finally, we fuse the three time-series heat generation data sequences that are connected end to end and mutually corroborating in mechanism to construct a complete liquid battery energy storage heat load prediction information that covers the next few hours and has both transient peak and dynamic decay characteristics.

[0046] The method provided in this embodiment significantly improves the accuracy and reliability of heat load prediction, provides high-quality data support for thermal inertia analysis and control command generation, effectively copes with the thermal shock caused by photovoltaic power fluctuations, ensures stable operation of energy storage equipment, extends the service life of core components, optimizes the mixing ratio and timing of cold and hot buffer zones, and improves the energy conversion efficiency and overall operational benefits of the energy storage system.

[0047] In some embodiments, based on the fuel cell stack heat generation ramp-up information, the delayed process of rapid accumulation of Joule heat in the fuel cell stack and its conduction to the storage tank via the heat exchanger during the power ramp-up period is analyzed, yielding the storage tank temperature rise lag characteristic. Based on the fuel cell stack steady-state heat generation information, the accumulation process of continuous input of fuel cell stack heat generation leading to heat diffusion from the edge to the center of the electrolyte in the storage tank during the high-power plateau period is analyzed, yielding the storage tank heat absorption accumulation characteristic. Based on the fuel cell stack heat dissipation information, the process of reduced fuel cell stack heat generation but the storage tank maintaining its temperature due to thermal inertia and slowly dissipating heat to the environment through the circulation pipeline during the power drop period is analyzed, yielding the storage tank cooling inertia characteristic. Combining the storage tank temperature rise lag characteristic, storage tank heat absorption accumulation characteristic, and storage tank cooling inertia characteristic, a set describing the response delay and change law of storage tank temperature under different power periods is constructed as a time-segmented thermal inertia information set.

[0048] The temperature rise lag characteristic of the storage tank can be seen as a delayed temperature response due to the accumulation of Joule heat in the fuel cell stack during the power ramp-up period. The heat absorption accumulation characteristic of the storage tank can be seen as the accumulation of heat through diffusion from the edge to the center of the electrolyte within the tank during the high-power plateau period. The cooling inertia characteristic of the storage tank can be seen as the tank maintaining its temperature and slowly dissipating heat due to thermal inertia after the heat generation from the fuel cell stack decreases during the power drop period.

[0049] Specifically, in cloud-based collaborative energy management of energy storage devices, photovoltaic power fluctuations cause phased differences in heat generation from the flow battery stack, and the storage tank temperature response exhibits thermal inertia. Ignoring this characteristic will lead to unrealistic thermal management strategies, resulting in temperature control deviations, energy losses (e.g., 10%-15%), equipment aging, and even safety hazards. Time-segmented thermal inertia information sets are crucial for connecting the photovoltaic-thermal coupling situation with co-mixing control, providing a core basis for precise thermal management and a necessary prerequisite for ensuring the stable and efficient operation of the energy storage system. To address these issues: First, for the power ramp-up period, we employ a thermoelectric coupling dynamic analysis method, combining stack heat generation ramp-up information to analyze the Joule heat generated inside the stack and its transmission through the plate heat exchanger (for example, with a heat transfer coefficient of 1500 W / (m²)). 2The entire delayed process of heat transfer from the electrolyte (e.g., at a flow rate of 10 L / min) to the tank inlet is as follows: By establishing a lumped-parameter or distributed-parameter heat transfer model of this process, key parameters of the tank's temperature rise lag characteristics can be extracted, such as the "temperature response delay time τ1" (e.g., 180 seconds) and the "initial temperature rise rate R1". Secondly, for the high-power plateau period, we apply non-uniform medium thermal diffusion simulation techniques. Based on the steady-state heat generation information of the fuel cell stack, we analyze the dynamics of the electrolyte in the tank under continuous heating. For example, after heat is injected from the inlet pipeline, a local high-temperature zone first forms inside the tank, and then the heat is mainly transferred through natural convection and limited conduction (the electrolyte thermal diffusivity is approximately 1.5 × 10⁻⁶). -7 m 2 The heat slowly diffuses from the tank surface (T_s) towards the center and far ends. This accumulation process is quantified using computational fluid dynamics (CFD) simulation or a simplified three-dimensional thermal network model, yielding characteristics of the tank's heat accumulation that characterize the gradual increase in overall temperature, such as the "heat saturation time T_s" (e.g., 30 minutes) and the "spatial temperature difference ΔT_max". Finally, for the power drop period, we employ a quantitative method for the thermal inertia decay process. Based on the heat dissipation information from the fuel cell stack, we analyze that when the heat source input decreases sharply, the heat stored in the tank mainly diffuses to the ambient air (assuming a convective heat transfer coefficient of 10 W / (m·K)) through the tank wall (assuming the thermal conductivity of the insulation layer is 0.04 W / (m·K)) and the external circulation pipes. 2 The heat dissipation process is a slow process. By establishing a first-order or higher-order decay model of this heat dissipation process, the core parameters of the tank's cooling inertial characteristics are obtained through fitting, such as the "temperature decay time constant τ2" (e.g., 600 seconds). Finally, the characteristic parameters and change law models extracted from these three stages are systematically integrated and formatted to form a structured time-segmented thermal inertial information set, thereby fully mapping the full-cycle response behavior of the tank temperature to fluctuating power.

[0050] The method provided in this embodiment accurately captures the temperature response pattern of the storage tank during different power periods, providing precise data support for the mixing control of the buffer zone. This ensures that the timing planning of the mixing ratio closely matches actual temperature changes, reducing energy loss, improving energy conversion efficiency, stabilizing the temperature environment of the storage tank, protecting the performance of the fuel cell stack and electrolyte, extending the service life of the energy storage equipment, and ensuring the long-term stable and efficient operation of the energy storage system under photovoltaic power fluctuation scenarios.

[0051] In some embodiments, based on the stack heat generation ramp-up information, the instantaneous process of rapid generation and accumulation of Joule heat in the stack body during the power ramp-up period is analyzed using internal resistance identification technology to obtain the stack core heat generation information; based on the stack core heat generation information, the diffusion process of heat from the stack core to the circulating electrolyte through the heat exchanger pipeline is analyzed according to the heat transfer parameters of the liquid battery heat exchanger to obtain the pipeline heat delay information; based on the pipeline heat delay information, the diffusion delay information of heat in the tank space during the heat convection process of the electrolyte carrying heat flowing from the pipeline inlet to the tank interior and mixing with the original electrolyte in the tank is analyzed according to the flow rate data of the tank circulation inlet to obtain the tank temperature rise hysteresis characteristics.

[0052] Internal resistance identification technology is a technique used to analyze the internal resistance characteristics of liquid battery stacks, thereby accurately calculating Joule heat generation. Core heat generation information can be instantaneous data showing the rapid generation and accumulation of Joule heat during the power ramp-up phase of the stack. Heat transfer parameters of the liquid battery heat exchanger are key parameters characterizing the heat transfer capability of the heat exchanger. Piping heat delay information is data related to the time delay in the transfer of heat from the stack core through the heat exchanger piping to the circulating electrolyte. Flow data at the tank circulation inlet is real-time flow information at the inlet of the circulating electrolyte into the tank. Thermal convection can be the process of heat transfer through fluid flow when the heat-carrying electrolyte mixes with the existing electrolyte in the tank.

[0053] Specifically, in the energy management system of cloud-based collaborative energy storage devices, there is a contradiction between the surge in heat generation from the fuel cell stack during the power ramp-up period and the non-instantaneous nature of heat transfer. Heat transfer requires multiple stages and involves delays. If the temperature rise lag characteristics of the storage tank are not constructed, subsequent blending control commands will lack accurate data support, easily leading to local overheating or blending strategy imbalance. It will also cause gaps in the time-segmented thermal inertia information set, reducing the adaptability of the energy storage system to photovoltaic power fluctuations. To address the above issues: First, the internal resistance identification technology module (e.g., using a recursive least squares estimation algorithm) is invoked. This module dynamically processes real-time voltage and current sampling signals (such as voltage sampling values) from the fuel cell stack sensors and performs online fitting. The system calculates the equivalent ohmic internal resistance of the fuel cell stack and accurately determines the transient Joule heat generation power curve during the power ramp-up period, which is determined by the product of the square of the current and the internal resistance. This allows the system to extract and output the "core heat generation information" that characterizes the rapid changes in heat source intensity and the localization of heat within the fuel cell stack itself. Following this core heat generation information, the system calls upon a pre-stored database of "liquid battery heat exchanger heat transfer parameters" (e.g., specific parameters recorded in the heat exchanger design document, such as the total heat transfer area of ​​the heat exchanger tubes, the thermal conductivity of the tube wall material, and the tube wall thickness). Simulation calculations are then performed using a lumped-parameter thermal network model or a one-dimensional unsteady-state heat transfer differential equation to specifically analyze the conditions under given inlet electrolyte flow rate and temperature. The calculated heat generation power, when flowing through this specific heat exchanger structure, experiences a time lag and amplitude attenuation in heat transfer due to the heat capacity of the pipe wall and the convective heat transfer resistance between the fluid and the pipe wall. This generates quantified "pipeline heat delay information," which clearly indicates the time constant and temperature change profile required for heat to transfer from the stack outlet to the tank circulation inlet. Finally, the system integrates real-time monitored "tank circulation inlet flow data" (e.g., instantaneous volumetric flow rate readings from an electromagnetic flowmeter) and, based on the tank's three-dimensional geometric model and inlet / outlet layout, uses the heat pulse obtained in the previous step, which has already reached the inlet, as input boundary conditions. Computational fluid dynamics (CFD) is then employed. The FD (Fluidized Diffusion) principle-simplified convection-diffusion mixing model (or based on experimentally calibrated mixing efficiency coefficients) simulates and analyzes the "thermal convection process" driven by a specific flow rate (such as the flow rate reading) when an electrolyte stream carrying additional heat enters the tank and mixes with the existing fluid in the tank due to velocity shear and density difference (caused by temperature difference). It solves the time required for this process to reach a preset uniformity in the tank space (e.g., the temperature standard deviation drops below a threshold), as well as the evolution gradient of the temperature field in the tank during this period. Finally, it outputs a dynamic feature set that comprehensively describes the overall lag of the tank temperature response behind the heat generation of the fuel cell stack, which integrates pipeline delay and tank mixing delay, i.e., the "tank temperature rise lag characteristic".

[0054] The method provided in this embodiment accurately captures the delay pattern of the entire heat transfer process, provides reliable data support for the preheating and mixing ratio sequence, avoids lag or deviation in mixing control, helps the system predict the temperature rise trend of the storage tank, dynamically balances the heat generation of the fuel cell stack and the temperature of the storage tank, maintains the stability of the electrolyte temperature, improves the adaptability and response speed of the energy storage equipment to photovoltaic power fluctuations, and ensures the safe and stable operation of the system and efficient energy utilization.

[0055] In some embodiments, based on the heat dissipation information of the fuel cell stack and combined with the power drop period, the starting moment of the rapid decrease in the heat generated by the fuel cell stack is determined, and the inertial process of the overall temperature drop of the tank being significantly delayed compared to the decrease in the heat generated by the fuel cell stack due to the constraints of electrolyte heat capacity and circulation flow rate is analyzed. The inertial process is characterized by a slow redistribution of the temperature gradient inside the tank, and the rate at which the average temperature of the tank decreases over time is limited by both the heat capacity of the tank structure and the circulation flow rate. Based on the inertial process, the temperature decay rate of the high-temperature electrolyte in the tank during the circulating heat exchange process is determined by calculating the functional relationship between the heat dissipation of the pipeline and the ambient temperature difference, thus obtaining the tank cooling inertial characteristics that describe the slow temperature drop characteristics of the tank after the power drop.

[0056] The initial moment can be the specific point in time when the heat generated by the fuel cell stack begins to decrease rapidly from a steady state. Electrolyte heat capacity can be the ability of the electrolyte in the liquid battery tank to hold heat. Circulation flow rate constraint can be the limiting condition on the flow velocity of the electrolyte in the liquid battery circulation pipeline. The overall tank temperature lag can be the phenomenon where, after the rapid decrease in heat generated by the fuel cell stack, the tank temperature does not decrease synchronously but exhibits a delayed decrease. Temperature gradient redistribution can be the process of readjusting the temperature differences of the electrolyte in different areas of the tank after changes in fuel cell stack heat generation. Tank structural heat capacity can be the ability of the tank structure itself to hold heat. Circulation flow rate can be the volume of electrolyte passing through the circulation pipeline per unit time. Pipeline heat dissipation can be the heat lost by the circulation pipeline during heat exchange with the environment. Ambient temperature difference can be the difference between the external ambient temperature and the electrolyte temperature inside the circulation pipeline. Functional relationship can be a mathematical correspondence used to quantify the relationship between pipeline heat dissipation and the ambient temperature difference. High-temperature electrolyte can be the electrolyte in the tank that is at a high temperature before the power drop due to continuous heat generation by the fuel cell stack. Temperature decay rate can be the rate at which the temperature of the high-temperature electrolyte decreases over time during the circulation heat exchange process.

[0057] Specifically, in cloud-based collaborative energy management of energy storage devices, photovoltaic power is susceptible to sudden drops due to natural factors. Storage tanks and electrolytes exhibit thermal inertia, meaning the tank temperature cannot synchronously decrease after a sudden drop in heat generation from the fuel cell stack. Without precise characterization of this cooling inertia, buffer mixing control commands may malfunction, leading to heat waste or equipment overheating, severely impacting the stability and energy management accuracy of the energy storage system. To address these issues, a time-series data analysis and inflection point identification algorithm is first used to process the "fuel cell stack heat dissipation information" (i.e., a set of time-series data describing the change in the fuel cell stack heat generation rate during the power drop period) transmitted from upstream. This algorithm calculates the local slope change of the sequence through a sliding window. When the slope value suddenly changes from a relatively flat (e.g., stable at a small negative value close to zero) to a large negative value exceeding a preset threshold (e.g., -10kW / min), it is determined that this moment is the "starting moment of rapid decrease" in fuel cell stack heat generation, thus accurately capturing the physical starting point of heat source weakening. Following this initial moment, the "inertial process" of the storage tank is analyzed in depth using distributed parameter thermodynamic modeling and simulation. This process is achieved by constructing a coupled model that describes the temperature field distribution inside the storage tank and the electrolyte flow. The "electrolyte heat capacity" is concretized as the product of the total mass of the electrolyte in the storage tank and its specific heat capacity (for example, for a storage tank with a capacity of 20 cubic meters, its total heat capacity can reach a huge value). The "circulation flow rate constraint" is quantified as the fixed flow rate setpoint of the circulation pump (such as 100 L / s). Simulation analysis reveals that even if the heat generated by the fuel cell stack decreases rapidly after the initial moment, the overall "average temperature" of the storage tank exhibits a significant exponential decay lag due to the "buffering" effect of its large heat capacity and the constraints of heat transfer efficiency at a limited flow rate. To quantify this gradual lag characteristic, a pipeline heat dissipation calculation model based on Newton's law of cooling is further introduced. The "pipeline heat dissipation" is expressed as a function proportional to the "ambient temperature difference" (i.e., the difference between the average temperature of the electrolyte in the pipeline and the ambient temperature). Taking into account the total heat transfer area and circulation flow rate of the pipeline, this heat dissipation model is substituted into the aforementioned thermodynamic model for iterative solution. This allows the calculation of the "temperature decay rate" (e.g., the number of degrees the average temperature decreases per hour) of the storage tank under given initial temperature and ambient conditions. Finally, the key parameters obtained from the analysis, such as the initial moment, the time constant representing the degree of lag, and the calculated temperature decay rate, are integrated and packaged into a structured dataset, which is the "tank cooling inertia characteristic" that accurately describes the dynamic thermal response of the storage tank after a sudden drop in power.

[0058] The method provided in this embodiment accurately captures the temperature drop pattern of the storage tank after a sudden drop in power, providing reliable data support for the generation of blending control commands, avoiding heat waste and equipment damage, making the blending ratio sequence of the hot and cold buffer zones more consistent with actual temperature changes, improving the energy management accuracy of energy storage equipment, ensuring stable operation of the system under power fluctuations, extending the service life of the equipment, and promoting the efficient consumption of photovoltaic energy.

[0059] In some embodiments, based on the lag characteristics of tank temperature rise and combined with the photovoltaic power fluctuation curve, the buffer mixing operation required in advance to offset the delayed process of heat transfer from the stack to the tank before the start of the power ramp-up period is analyzed, resulting in a preheating mixing ratio sequence; based on the heat absorption accumulation characteristics of the tank and combined with the high-power plateau period, the buffer mixing ratio that needs to be dynamically adjusted to balance the continuous heat generation of the stack and the heat diffusion of the tank during the high-power plateau period is analyzed, resulting in a plateau period mixing adjustment sequence; based on the cooling inertia characteristics of the tank and combined with the power drop period, the buffer mixing operation that needs to be delayed after the start of the power drop period to utilize the residual heat of the tank and slow down the rate of temperature drop is analyzed, resulting in a lag mixing ratio sequence; combining the preheating mixing ratio sequence, the plateau period mixing adjustment sequence, and the lag mixing ratio sequence, they are arranged and smoothed in chronological order to generate a buffer mixing control instruction set for dynamically planning the timing of the hot and cold buffer mixing ratios.

[0060] The preheating blending ratio sequence can be a blending ratio change sequence set before the power ramp-up period to compensate for heat transfer delays. The plateau period blending adjustment sequence can be a blending ratio sequence dynamically adjusted during the high-power plateau period to balance the heat generation of the fuel cell stack and the heat diffusion from the storage tank. The delayed blending ratio sequence can be a blending ratio sequence that is adjusted later after the power drop period to utilize the residual heat of the storage tank and slow down the rate of temperature drop. Smooth transition processing can be a method of achieving continuous and stable command by eliminating abrupt changes in the ratio and optimizing the rate of change when integrating blending ratio sequences from different time periods in chronological order.

[0061] Specifically, in the energy management process of cloud-based collaborative energy storage devices, photovoltaic power is easily affected by cloud cover and other factors, resulting in instantaneous fluctuations. The temperature response of flow battery tanks exhibits lag, accumulation, and inertia characteristics. Fixed buffer mixing modes cannot adapt to dynamic operating conditions, easily leading to temperature imbalance, waste of residual heat, or equipment damage. Furthermore, the lack of smooth transitions can cause system oscillations. To address these issues: Firstly, regarding the power ramp-up period, based on the "tank temperature rise lag characteristic" (for example, quantitative analysis shows that heat transfer from the stack to the tank core may be delayed by several minutes), combined with the photovoltaic power fluctuation curve, time-series alignment and heat compensation analysis techniques are employed to analyze the ramp-up... The starting point is determined, and a time window matching the thermal delay is traced back (e.g., 3 minutes in advance). Within this window, a "preheating blending ratio sequence" is initiated. The logic is to gradually increase the blending ratio of the hot buffer (or decrease the cold buffer) before the actual significant increase in heat generation from the fuel cell stack. This is equivalent to pre-injecting some heat into the tank to compensate for the upcoming transfer delay, allowing the tank temperature to more synchronously follow the power increase. Secondly, for the high-power plateau period, based on the "tank heat absorption accumulation characteristics," dynamic heat balance analysis technology is used to analyze in real time the steady-state heat generation rate of the fuel cell stack and the heat diffusion rate inside the tank (this rate is related to the electrolyte circulation) during this period. The dynamic relationship between flow rate and tank geometry is analyzed, and a "plateau period blending adjustment sequence" is generated accordingly. For example, when an overheating risk is detected in a local area of ​​the tank due to insufficient heat diffusion, the instantaneous blending ratio of the cold buffer zone is dynamically increased to enhance cooling. Conversely, if heat diffusion is uniform, the ratio is appropriately adjusted back to save energy. Finally, for the power drop period, the "tank cooling inertia characteristics" are closely considered, and waste heat utilization and temperature drop rate smoothing analysis techniques are used. After identifying the starting point of the power drop, the blending ratio of the hot buffer zone is not immediately and significantly reduced, but a "hysteresis blending ratio sequence" is initiated. This sequence instructs the control system to adjust the blending ratio in the cold buffer zone. The mixing ratio is slowly and gradually adjusted within a certain delay period (such as within 5 minutes calculated based on inertial characteristics). This is to controllably utilize the residual heat accumulated inside the tank, slow down the rate of temperature drop, and avoid the negative impact of sudden temperature drop on the electrochemical reaction activity. Finally, through instruction sequence fusion and smooth transition technology, the three sequences arranged in timeline are seamlessly connected and smoothed (for example, a first-order filtering algorithm is used at the sequence junction to avoid instruction jumps), generating a set of buffer mixing control instructions that are time-continuous, have clear instructions, and can adapt to the panorama of future power fluctuations, directly driving the precise action of the mixing valve.

[0062] The method provided in this embodiment achieves precise matching between blending control and power fluctuations and the thermal inertia of the storage tank, effectively offsetting heat transfer delays, balancing heat accumulation, recovering and utilizing waste heat, maintaining stable electrolyte temperature, avoiding system oscillations through smooth transition, optimizing energy utilization efficiency, reducing equipment wear, extending the service life of energy storage equipment, and ensuring the continuous and efficient operation of the energy storage system under complex fluctuation scenarios.

[0063] In some embodiments, based on the preheating blending ratio sequence, the analysis is a process of compensating for heat transfer delay by increasing the blending ratio of the thermal buffer before the stack heat generation increases, resulting in a preheating control sequence for the power ramp-up period; based on the plateau period blending adjustment sequence, the analysis is a process of dynamically balancing the blending ratio of the hot and cold buffers to match the heat diffusion rate inside the storage tank during the stack heat generation steady-state period, resulting in a dynamic adjustment sequence for the high-power plateau period; based on the lag blending ratio sequence, the analysis is a process of delaying the reduction of the thermal buffer blending ratio to utilize the residual heat of the storage tank and slow down the temperature drop rate after the stack heat generation decreases, resulting in a lag control sequence for the power drop period; the preheating control sequence, dynamic adjustment sequence, and lag control sequence are used as control variable sequences in the prediction time domain, and rolling optimization and smooth transition processing are performed to obtain the blending ratio sequence.

[0064] The prediction time domain can be the time range for dynamically programming the mixing ratio of the cold and hot buffer zones. The control variable sequence can be a set of variables that characterizes the adjustment pattern of the mixing ratio at different time periods, after integrating the preheating control sequence, dynamic adjustment sequence, and lag control sequence. Rolling optimization can be a method of continuously iteratively optimizing the control variable sequence based on real-time operating conditions (such as actual fluctuations in photovoltaic power and real-time tank temperature) within the prediction time domain. Smooth transition processing can be a gradient adjustment of the mixing ratio at the junction of the three types of control sequences to avoid equipment shock caused by sudden changes in the ratio. The mixing ratio time sequence can be a continuous change curve of the mixing ratio of the cold and hot buffer zones at different time points within the prediction time domain after integration and optimization.

[0065] Specifically, in photovoltaic and flow battery energy storage scenarios, photovoltaic power is susceptible to fluctuations in three stages due to natural factors, while the temperature response of flow battery tanks exhibits time-dependent thermal inertia. Without targeted mixing ratio timing planning, this can lead to delayed heat transfer, waste of residual heat, excessive temperature fluctuations, and potentially equipment shock. Therefore, it is necessary to construct control sequences according to different power periods, and optimize and smooth transitions to form a suitable mixing ratio timing sequence. This is a core requirement for ensuring the efficiency and stability of energy storage equipment. To address these issues: based on the predicted power ramp-up start point (e.g., a sudden power surge at 12:00 noon due to cloud cover relocation), combined with the heat transfer delay time calculated by the tank temperature rise lag model (e.g., 5 minutes), the electrolyte ratio pumped from the thermal buffer into the tank will be gradually increased (e.g., linearly increasing from 20% to 50%) before the actual power ramp-up (e.g., 11:55). This injects "advance" heat to offset subsequent transfer delays. After entering the high-power plateau period, an adaptive fuzzy control method is used to generate dynamic... Timing adjustment: Real-time monitoring of temperature gradients measured by multiple thermocouples inside the storage tank (e.g., when the temperature difference between the center and the edge reaches a certain threshold), and dynamic fine-tuning of the mixing ratio based on a fuzzy rule base (e.g., if the center temperature rises too quickly, the hot buffer ratio is temporarily reduced by a few percentage points, such as from 50% to 47%), to balance the continuous heat generation of the fuel cell stack with the heat dissipation and internal heat diffusion capacity of the storage tank itself. When a power drop period is predicted (e.g., at 14:00 due to cloud cover), hysteresis based on model predictive control (MPC) is initiated. Post-control timing generation: The MPC optimizer uses the tank cooling inertia model in the next time domain (e.g., the next 30 minutes) as a constraint, and takes temperature stability and minimum pumping energy consumption as the objective functions to solve for the optimal decrease trajectory of the hot buffer mixing ratio (e.g., after the power decreases, the command first maintains a 40% ratio for 10 minutes, and then slowly and linearly decreases to 25% over 20 minutes), thereby making full use of the tank's residual heat and slowing down the temperature drop rate. Finally, the above three timing segments are used as the basis and input into a rolling time domain optimization (RHO) framework. This framework uses the latest photovoltaic power prediction data and the actual temperature feedback of the tank at a fixed period (e.g., every 5 minutes) to re-roll and refresh the global mixing ratio timing; and applies a first-order low-pass digital filter to the optimized output command sequence to smooth out possible command jumps between adjacent control cycles (e.g., smoothing the calculated sudden increase ratio command). Finally, it outputs a mixing ratio execution timing that is continuous in time, physically matches the thermal inertia of each time period, and can adapt to real-time fluctuations.

[0066] The method provided in this embodiment achieves precise matching between the mixing operation and photovoltaic power fluctuations and storage tank temperature characteristics, effectively solving various temperature-related problems. It not only ensures electrolyte activity and energy conversion efficiency, but also avoids equipment impact, extends service life, and improves the energy storage system's ability to absorb renewable energy, ensuring stable and efficient energy management.

[0067] Figure 3 A schematic diagram of a cloud-based collaborative energy storage device energy management system provided in one embodiment of this application is shown below. Figure 3 As shown, the cloud-based collaborative energy storage device energy management system 300 of this embodiment includes: a coupling situation module 301, a segmented analysis module 302, and a dynamic instruction module 303.

[0068] The coupling situation module 301 is used to acquire multi-source operating data streams and cloud weather forecast information. Based on the multi-source operating data streams and the cloud weather forecast information, it analyzes the photovoltaic power fluctuation curve and liquid battery energy storage heat load prediction information for future periods to obtain a photovoltaic-thermal coupling situation set. The segmented analysis module 302 is used to analyze the temperature change characteristics of the liquid battery tank caused by the heat generation process of the flow battery stack under the photovoltaic power fluctuation curve based on the photovoltaic-thermal coupling situation set to obtain a time-segmented thermal inertia information set. The dynamic instruction module 303 is used to generate a buffer mixing control instruction set adapted to instantaneous power fluctuations based on the time-segmented thermal inertia information set, dynamically plan the mixing ratio timing of the cold buffer and the hot buffer, and output the energy management log of the energy storage device.

[0069] Optionally, during the construction of the photovoltaic-thermal coupling situation set, the coupling situation module 301 is specifically used for: the multi-source operating data stream including photovoltaic power sequence, photovoltaic equipment status parameters, and liquid battery equipment status parameters; the cloud weather forecast information including light intensity prediction sequence and cloud coverage prediction data; based on the photovoltaic power sequence and the photovoltaic equipment status parameters, combined with the light intensity prediction sequence, analyzing the future periodic photovoltaic power change information to obtain a basic power curve; based on the cloud coverage prediction data, combined with the basic power curve, analyzing the power mutation time point and power change amplitude caused by cloud coverage changes to obtain power fluctuation characteristics; based on the power fluctuation characteristics, combined with the liquid battery equipment status parameters, analyzing the heat generation process of the flow battery stack under power changes to obtain the liquid battery energy storage heat load prediction information; the heat load prediction information and the power fluctuation curve together constitute the photovoltaic-thermal coupling situation set.

[0070] Optionally, the coupling situation module 301, during the construction of the power fluctuation characteristics, is specifically used for: based on the cloud coverage prediction data, employing a time-series abrupt change point detection algorithm to analyze the start and end times of the rise and fall of the coverage value, obtaining cloud coverage change information; based on the cloud coverage change information, applying a steady-state power estimation algorithm to analyze the photovoltaic power generation information after the end of the cloud change process, obtaining the theoretical photovoltaic output level; mapping the start and end times of cloud coverage change information to the basic power curve, analyzing the decrease or increase process of the power in the basic power curve within the time period defined by the start and end times of cloud coverage, obtaining the power abrupt change time point and power change trend; and constructing a description set containing the abrupt change time sequence and change amplitude as the power fluctuation characteristics based on the theoretical photovoltaic output level, the power abrupt change time point, and the power change trend.

[0071] Optionally, the coupling situation module 301, during the construction of the liquid battery energy storage heat load prediction information, is specifically used for: dividing the power ramp-up period, high-power plateau period, and power drop period based on the power mutation time point and the power change amplitude; using real-time parameter identification technology to analyze the internal resistance Joule heat generation process caused by the rapid increase of liquid battery current during the power ramp-up period, obtaining stack heat generation ramp-up information; using steady-state thermal balance analysis method to analyze the steady-state heat generation process of the liquid battery stack during the high-power plateau period, obtaining stack steady-state heat generation information; using heat dissipation dynamic characteristic analysis method to analyze the lag process of internal polarization heat dissipation of the liquid battery stack caused by the rapid decrease of current during the power drop period, obtaining stack heat dissipation information; and combining the stack heat generation ramp-up information, the stack steady-state heat generation information, and the stack heat dissipation information to construct the liquid battery energy storage heat load prediction information.

[0072] Optionally, the segmented analysis module 302, during the construction of the time-segmented thermal inertia information set, is specifically used for: analyzing the delayed process of rapid accumulation of Joule heat in the fuel cell stack and its conduction to the storage tank via the heat exchanger during the power ramp-up period, based on the fuel cell stack heat generation ramp-up information, to obtain the storage tank temperature rise lag characteristic; analyzing the accumulation process of continuous fuel cell stack heat generation during the high-power plateau period, causing the electrolyte in the storage tank to diffuse heat from the edge to the center, based on the fuel cell stack heat generation dissipation information, to analyze the process of reduced fuel cell stack heat generation during the power drop period, but the storage tank maintains its temperature due to thermal inertia and slowly dissipates heat to the environment through the circulation pipeline, based on the fuel cell stack heat generation dissipation information, to obtain the storage tank cooling inertia characteristic; and combining the storage tank temperature rise lag characteristic, the storage tank heat absorption accumulation characteristic, and the storage tank cooling inertia characteristic to construct a set describing the response delay and change law of the storage tank temperature under different power periods, as the time-segmented thermal inertia information set.

[0073] Optionally, the segmented analysis module 302, during the construction of the tank temperature rise lag characteristic, is specifically used for: based on the fuel cell stack heat generation ramp-up information, analyzing the instantaneous process of rapid generation and accumulation of Joule heat in the fuel cell stack body during the power ramp-up period using internal resistance identification technology, to obtain fuel cell core heat generation information; based on the fuel cell core heat generation information, analyzing the diffusion process of heat from the fuel cell stack core to the circulating electrolyte through the heat exchanger pipeline according to the heat transfer parameters of the liquid battery heat exchanger, to obtain pipeline heat delay information; based on the pipeline heat delay information, analyzing the diffusion delay information of heat in the tank space during the heat convection process of the electrolyte carrying heat flowing from the pipeline inlet to the tank interior and mixing with the original electrolyte in the tank according to the flow rate data of the tank circulation inlet, to obtain the tank temperature rise lag characteristic.

[0074] Optionally, the segmented analysis module 302, during the construction of the tank cooling inertial characteristics, is specifically used for: determining the starting moment of the rapid decrease in heat generation of the fuel cell stack based on the heat dissipation information of the fuel cell stack and in conjunction with the power drop period; and analyzing the inertial process from the starting moment, where the overall temperature drop of the tank is significantly lagging behind the decrease in heat generation of the fuel cell stack due to the constraints of electrolyte heat capacity and circulation flow rate; the inertial process is characterized by a slow redistribution of the temperature gradient inside the tank, and the rate at which the average temperature of the tank decreases over time is limited by both the heat capacity of the tank structure and the circulation flow rate; based on the inertial process, by calculating the functional relationship between the heat dissipation of the pipeline and the ambient temperature difference, determining the temperature decay rate of the high-temperature electrolyte in the tank during the circulating heat exchange process, and obtaining the tank cooling inertial characteristics that describe the slow temperature drop characteristics of the tank after the power drop.

[0075] Optionally, when the dynamic instruction module 303 generates a buffer mixing control instruction set adapted to instantaneous power fluctuations based on the time-segmented thermal inertia information set, it is specifically used to: analyze the buffer mixing operation required in advance to offset the delayed process of heat transfer from the stack to the storage tank before the start of the power ramp-up period, based on the storage tank temperature rise lag characteristics and the photovoltaic power fluctuation curve, and obtain a preheating mixing ratio sequence; and analyze the balance between the continuous heat generation of the stack and the heat of the storage tank during the high-power plateau period, based on the storage tank heat absorption accumulation characteristics and the high-power plateau period. The mixing ratio of the buffer zone, which needs to be dynamically adjusted due to diffusion, is used to obtain the mixing adjustment sequence during the plateau period. Based on the cooling inertia characteristics of the storage tank and combined with the power drop period, the buffer mixing operation that needs to be delayed after the start of the power drop period in order to utilize the residual heat of the storage tank and slow down the rate of temperature drop is analyzed, resulting in the lag mixing ratio sequence. The preheating mixing ratio sequence, the plateau period mixing adjustment sequence, and the lag mixing ratio sequence are combined and arranged in chronological order and smoothed to generate the buffer mixing control instruction set for dynamically planning the mixing ratio timing of the hot and cold buffer zones.

[0076] Optionally, when the dynamic instruction module 303 dynamically plans the mixing ratio sequence of the cold buffer and the hot buffer, it is specifically used to: analyze the operation process of increasing the mixing ratio of the hot buffer to compensate for the heat transfer delay before the heat generation of the fuel cell stack increases, based on the preheating mixing ratio sequence, to obtain the preheating control sequence for the power ramp-up period; analyze the operation process of dynamically balancing the mixing ratio of the cold and hot buffers to match the heat diffusion rate inside the storage tank during the steady-state period of the fuel cell stack heat generation, based on the plateau mixing adjustment sequence, to obtain the dynamic adjustment sequence for the high-power plateau period; analyze the operation process of delaying the reduction of the mixing ratio of the hot buffer to utilize the residual heat of the storage tank and slow down the temperature drop rate after the heat generation of the fuel cell stack decreases, based on the lag mixing ratio sequence, to obtain the lag control sequence for the power drop period; and use the preheating control sequence, the dynamic adjustment sequence, and the lag control sequence as the control variable sequence in the prediction time domain, and perform rolling optimization and smooth transition processing to obtain the mixing ratio sequence.

[0077] The system in this embodiment can be used to execute the methods of any of the above embodiments, and its implementation principle and technical effect are similar, so they will not be described again here.

Claims

1. A cloud-based collaborative energy storage device energy management method, characterized in that, include: Acquire multi-source operational data streams and cloud-based weather forecast information. Based on the multi-source operational data streams and the cloud-based weather forecast information, analyze the photovoltaic power fluctuation curve and liquid battery energy storage heat load prediction information for future cycles to obtain a photovoltaic-thermal coupling situation set. Based on the photovoltaic-thermal coupling situation set, the temperature change characteristics of the liquid battery storage tank caused by the heat generation process of the flow battery stack under the photovoltaic power fluctuation curve are analyzed to obtain a time-segmented thermal inertia information set. Based on the time-segmented thermal inertia information set, a buffer mixing control instruction set adapted to instantaneous power fluctuations is generated, the mixing ratio timing of the cold buffer and the hot buffer is dynamically planned, and the energy management log of the energy storage device is output.

2. The method according to claim 1, characterized in that, The process of constructing the photovoltaic-thermal coupling situation set includes: The multi-source operation data stream includes photovoltaic power sequence, photovoltaic equipment status parameters and liquid battery equipment status parameters; The cloud-based weather forecast information includes light intensity prediction sequences and cloud cover prediction data; Based on the photovoltaic power sequence, the photovoltaic equipment status parameters, and the light intensity prediction sequence, the changes in photovoltaic power in future periods are analyzed to obtain the basic power curve; Based on the cloud cover prediction data and the basic power curve, the timing and magnitude of power abrupt changes in the basic power curve caused by changes in cloud cover are analyzed to obtain power fluctuation characteristics. Based on the power fluctuation characteristics and combined with the state parameters of the liquid battery device, the heat generation process of the flow battery stack under power change is analyzed to obtain the predicted heat load information of the liquid battery energy storage. The heat load forecast information and the power fluctuation curve together constitute the photovoltaic-thermal coupling situation set.

3. The method according to claim 2, characterized in that, The process of constructing the power fluctuation characteristics includes: Based on the cloud coverage prediction data, a time-series abrupt change detection algorithm is used to analyze the start and end times of the rise and fall of the coverage value, and obtain cloud coverage change information. Based on the cloud cover change information, a steady-state power estimation algorithm is applied to analyze the photovoltaic power generation information after the cloud change process ends, and the theoretical photovoltaic output level is obtained. The cloud cover change information is mapped to the basic power curve at the start and end times of cloud cover. The decrease or increase of power in the basic power curve within the time period defined by the start and end times of cloud cover is analyzed to obtain the power change time point and power change trend. Based on the theoretical photovoltaic output level, the power mutation time point, and the power change trend, a description set including the mutation time sequence and change amplitude is constructed as the power fluctuation characteristics.

4. The method according to claim 3, characterized in that, The process of constructing the liquid battery energy storage heat load prediction information includes: Based on the power mutation time point and the power change amplitude, the power ramp-up period, high power plateau period and power drop period are divided. By employing real-time parameter identification technology, the internal resistance Joule heating process caused by the rapid increase in liquid battery current during the power ramp-up period is analyzed to obtain stack heat generation ramp-up information. The steady-state thermal balance analysis method was used to analyze the steady-state heat generation process of the liquid battery stack during the high-power plateau period, and the steady-state heat generation information of the stack was obtained. The dynamic heat dissipation characteristic analysis method is used to analyze the hysteresis process of polarization heat dissipation inside the liquid battery stack due to the rapid decrease in current during the power drop period, and obtain the heat generation and dissipation information of the stack. By combining the heat generation ramp-up information of the fuel cell stack, the steady-state heat generation information of the fuel cell stack, and the heat dissipation information of the fuel cell stack, the heat load prediction information of the liquid battery energy storage is constructed.

5. The method according to claim 4, characterized in that, The process of constructing the time-segmented thermal inertia information set includes: Based on the heat generation ramp-up information of the fuel cell stack, the delayed process of rapid accumulation of Joule heat in the fuel cell stack and its conduction to the storage tank via the heat exchanger during the power ramp-up period is analyzed, and the lag characteristic of the storage tank temperature rise is obtained. Based on the steady-state heat generation information of the fuel cell stack, the cumulative process of heat diffusion from the edge to the center of the electrolyte in the storage tank caused by the continuous input of heat generation from the fuel cell stack during the high-power plateau period is analyzed, and the heat absorption accumulation characteristics of the storage tank are obtained. Based on the heat dissipation information of the fuel cell stack, the process of heat generation of the fuel cell stack decreasing during the power drop period is analyzed, but the temperature inside the storage tank is maintained due to thermal inertia and heat is slowly dissipated to the environment through the circulation pipeline, thus obtaining the cooling inertia characteristics of the storage tank. By combining the characteristics of the tank temperature rise lag, the characteristics of the tank heat absorption accumulation, and the characteristics of the tank cooling inertia, a set describing the response delay and variation law of the tank temperature under different power periods is constructed as the time-segmented thermal inertia information set.

6. The method according to claim 5, characterized in that, The process of constructing the temperature rise hysteresis characteristics of the storage tank includes: Based on the heat generation ramp-up information of the fuel cell stack, the instantaneous process of rapid generation and accumulation of Joule heat in the fuel cell stack body during the power ramp-up period is analyzed using internal resistance identification technology to obtain the core heat generation information of the fuel cell stack. Based on the heat generation information of the fuel cell core, and according to the heat transfer parameters of the liquid battery heat exchanger, the diffusion process of heat from the fuel cell core to the circulating electrolyte through the heat exchanger pipeline is analyzed to obtain the pipeline heat delay information. Based on the pipeline heat delay information, and according to the flow rate data at the tank circulation inlet, the diffusion delay information of heat in the tank space during the heat convection process of the electrolyte carrying heat flowing from the pipeline inlet to the tank interior and mixing with the original electrolyte in the tank is analyzed, and the temperature rise lag characteristics of the tank are obtained.

7. The method according to claim 6, characterized in that, The process of constructing the cooling inertial characteristics of the storage tank includes: Based on the heat dissipation information of the fuel cell stack, combined with the power drop period, the starting moment of the rapid decrease in the heat generation of the fuel cell stack is determined, and it is analyzed that from the starting moment, the overall temperature drop of the storage tank is significantly lagging behind the inertial process of the decrease in the heat generation of the fuel cell stack due to the constraints of electrolyte heat capacity and circulation flow rate. The inertial process is characterized by a slow redistribution of the internal temperature gradient of the tank, and the rate at which the average temperature of the tank decreases over time is limited by both the heat capacity of the tank structure and the circulation flow rate. Based on the inertial process, by calculating the functional relationship between the heat dissipation of the pipeline and the ambient temperature difference, the temperature decay rate of the high-temperature electrolyte in the tank during the circulating heat exchange process is determined, and the cooling inertial characteristics of the tank describing the slow temperature drop characteristics after a sudden drop in power are obtained.

8. The method according to claim 7, characterized in that, The step of generating a buffer mixing control command set adapted to instantaneous power fluctuations based on the time-segmented thermal inertia information set includes: Based on the lag characteristics of the storage tank temperature rise and combined with the photovoltaic power fluctuation curve, the buffer mixing operation required in advance to offset the delayed process of heat transfer from the stack to the storage tank before the start of the power ramp-up period is analyzed, and the preheating mixing ratio sequence is obtained. Based on the heat absorption and accumulation characteristics of the storage tank, and combined with the high-power plateau period, the buffer mixing ratio that needs to be dynamically adjusted to balance the continuous heat generation of the fuel cell stack and the heat diffusion of the storage tank during the high-power plateau period is analyzed, and the mixing adjustment sequence of the plateau period is obtained. Based on the cooling inertial characteristics of the storage tank and combined with the power drop period, the buffer mixing operation that needs to be delayed after the start of the power drop period in order to utilize the residual heat of the storage tank and slow down the rate of temperature drop is analyzed, and the hysteretic mixing ratio sequence is obtained. The preheating blending ratio sequence, the plateau period blending adjustment sequence, and the lag blending ratio sequence are combined and arranged in chronological order and smoothed to generate the buffer blending control instruction set for dynamically planning the blending ratio timing of the hot and cold buffer zones.

9. The method according to claim 8, characterized in that, The timing of the mixing ratio of the cold buffer and the hot buffer in the dynamic programming includes: Based on the preheating mixing ratio sequence, it is analyzed as an operation process of compensating for heat transfer delay by increasing the mixing ratio of the thermal buffer before the heat generation of the fuel cell stack increases, thus obtaining the preheating control timing sequence for the power ramp-up period. Based on the aforementioned plateau period mixing and adjustment sequence, it is analyzed as an operation process in which the mixing ratio of cold and hot buffer zones is dynamically balanced to match the heat diffusion rate inside the storage tank during the steady state of heat generation of the fuel cell stack, thus obtaining a dynamic adjustment timing sequence for the high-power plateau period. Based on the aforementioned hysteresis blending ratio sequence, the analysis is as follows: after the heat generation of the fuel cell stack decreases, the operation process of delaying the reduction of the blending ratio in the thermal buffer zone to utilize the residual heat of the storage tank and slow down the rate of temperature drop is obtained, resulting in a hysteresis control timing sequence for the power drop period. The preheating control timing sequence, the dynamic adjustment timing sequence, and the lag control timing sequence are used as a sequence of control variables in the prediction time domain. Rolling optimization and smooth transition processing are performed to obtain the mixing ratio timing sequence.

10. A cloud-based collaborative energy storage device energy management system, characterized in that, The method applied to any one of claims 1-9 includes: The coupling situation module is used to acquire multi-source operation data streams and cloud weather forecast information. Based on the multi-source operation data streams and the cloud weather forecast information, it analyzes the photovoltaic power fluctuation curve and liquid battery energy storage heat load prediction information for future periods to obtain a photovoltaic-thermal coupling situation set. The segmented analysis module is used to analyze the temperature change characteristics of the liquid battery storage tank caused by the heat generation process of the flow battery stack under the photovoltaic power fluctuation curve based on the photovoltaic-thermal coupling situation set, and obtain the time-segmented thermal inertia information set. The dynamic instruction module is used to generate a set of buffer mixing control instructions that adapts to instantaneous power fluctuations based on the time-segmented thermal inertia information set, dynamically plan the mixing ratio timing of the cold buffer and the hot buffer, and output the energy management log of the energy storage device.

Images