Cooperative control method, system and equipment for hybrid energy storage system

Abstract

The invention provides a cooperative control method, system and device for a hybrid energy storage system, and relates to the technical field of electric energy storage, and the method comprises the steps: executing the real-time state collection of an electrochemical energy storage unit and a gravity energy storage unit through a distributed sensor network; reading time sequence external environment data and power grid load demand data, and performing prediction analysis; establishing a collaborative scheduling objective function; performing multi-target scheduling optimization based on the time sequence energy storage demand data and the real-time state data set; and dynamically adjusting a cooperative charging and discharging strategy of the electrochemical energy storage unit and the gravity energy storage unit based on a multi-target scheduling optimization result. Through the method and the device, the technical problem that the overall performance of the hybrid energy storage system is affected due to the fact that the cooperative scheduling of the hybrid energy storage cannot be optimized in real time in a dynamically changing power grid environment due to the adoption of a static and isolated control rule in the prior art is solved, and the overall performance of the hybrid energy storage system is improved through the cooperative scheduling decision-making mode of multi-target dynamic optimization. And the overall performance of the hybrid energy storage system is optimized.

Description

Technical Field

[0001] This application relates to the field of energy storage technology, specifically to collaborative control methods, systems, and equipment for hybrid energy storage systems. Background Technology

[0002] Current control strategies for hybrid energy storage systems largely rely on static rules based on fixed thresholds or independent control of each energy storage unit. These strategies exhibit significant limitations in adaptability to dynamic fluctuations in grid load and renewable energy output. They struggle to provide refined management based on the real-time health status of each energy storage unit, lack comprehensive consideration of overall dynamic characteristics, and fail to achieve optimal energy allocation coordination in complex operating environments with frequent grid fluctuations. Consequently, the overall performance of hybrid energy storage systems falls short of optimal standards. Furthermore, static and isolated control rules not only fail to achieve optimal dynamic power allocation among heterogeneous energy storage units but may also accelerate the aging of critical equipment (such as electrochemical batteries) due to frequent switching operations and inappropriate power commands, thereby increasing the operation and maintenance costs of hybrid energy storage systems.

[0003] In summary, existing technologies suffer from the technical problem that the use of static and isolated control rules makes it impossible to optimize the coordinated scheduling of hybrid energy storage in real time in a dynamically changing power grid environment, thereby affecting the overall performance of the hybrid energy storage system. Summary of the Invention

[0004] The purpose of this application is to provide a collaborative control method, system, and device for hybrid energy storage systems, in order to solve the technical problem in the prior art that the use of static and isolated control rules makes it impossible to optimize the collaborative scheduling of hybrid energy storage in real time in a dynamically changing power grid environment, thereby affecting the overall performance of the hybrid energy storage system.

[0005] To achieve the above objectives, this application provides a collaborative control method, system, and device for hybrid energy storage systems.

[0006] In a first aspect, this application provides a collaborative control method for a hybrid energy storage system. This collaborative control method is implemented through a collaborative control system for the hybrid energy storage system. The method includes: utilizing a distributed sensor network to collect real-time status data of electrochemical energy storage units and gravity energy storage units, establishing a real-time status dataset; reading time-series external environmental data and grid load demand data, and performing predictive analysis using a predictive model based on the time-series external environmental data and grid load demand data to establish time-series energy storage demand data; establishing a collaborative scheduling objective function, which includes a demand adaptation term, a health impact term for energy storage units, an energy storage cost adaptation term, and an energy storage switching penalty term; using the collaborative scheduling objective function to perform multi-objective scheduling optimization based on the time-series energy storage demand data and the real-time status dataset, establishing a multi-objective scheduling optimization result; and dynamically adjusting the collaborative charging and discharging strategies of the electrochemical energy storage units and gravity energy storage units based on the multi-objective scheduling optimization result.

[0007] Optionally, the output analysis layer in the prediction model is used to evaluate the stable power supply through time-series external environmental data and establish a time-series stable output; according to the supply and demand comparison layer, the grid load demand data and the time-series stable output are compared in time series, and time-series energy storage demand data are established based on the time series residuals. The time-series energy storage demand data represents the energy storage supply compensation demand of the hybrid energy storage system with time nodes.

[0008] Optionally, the demand adaptation term of the cooperative scheduling objective function includes power demand adaptation and response rate adaptation, and the calculation formula for the demand adaptation term is as follows: ;in, Characteristic requirements adaptation items, The total time step, Represents the current time point, Characterizing grid load demand at time nodes Power requirements, Characterizing hybrid energy storage systems at time nodes 'output power' Characterizing hybrid energy storage systems at time nodes 'output power' , These are the weighting factors for power demand adaptation and response rate adaptation, respectively.

[0009] Optionally, the calculation formulas for the health impact item and energy storage switching penalty item of the energy storage unit are as follows: in, Characterizing health impact items, Representation at time nodes The health impact values ​​mapped by the electrochemical charge-discharge depth, Representation at time nodes Electrochemical output power, Characterizing the rated discharge power of an electrochemical energy storage unit, Representation at time nodes The health impact values ​​mapped by gravity charge and discharge depth. To be at the time node Gravity output power, Characterizing the rated discharge power of the gravity energy storage unit, and Weighting factors characterizing the charge / discharge depth terms for electrochemical and gravitational forces, respectively. and Weighting factors characterizing the discharge power terms of electrochemical and gravitational forces, respectively; ;in, As a penalty for switching energy storage, For switching indicator functions, To switch the penalty coefficient.

[0010] Optionally, the following steps are taken: acquiring pre-control data; establishing an initial population based on the pre-control data, where each solution in the initial population represents a cooperative control strategy; using the time-series energy storage demand data and the real-time state dataset as constraints; evaluating the fit of the initial population based on the cooperative scheduling objective function; and establishing a mapping fit value; using the mapping fit value to perform population iterative analysis of the initial population and executing iterative optimization management to complete multi-objective scheduling optimization.

[0011] Optionally, a periodic selection strategy for population iteration is established, which includes a random search period, an elite retention period, and a convergent search period; the current iteration round is obtained, and the periodic selection strategy is invoked according to the iteration round to generate a retention threshold and a random factor for elite individuals; a selection operation is performed based on the retention threshold, the random factor, and the mapping fit value, and crossover iteration is performed using the selection results to update the population.

[0012] Optionally, the electrochemical energy storage unit and the gravity energy storage unit are controlled and monitored to establish a real control response; the real control response is used to evaluate the deviation of the coordinated charging and discharging strategy and establish a deviation evaluation feedback; a dynamic compensation strategy is established based on the deviation evaluation feedback, and the dynamic compensation strategy is used to dynamically compensate the coordinated charging and discharging strategy.

[0013] Optionally, the real-time status dataset is used to perform a health assessment of the hybrid energy storage system and establish a health early warning signal; control risk factors are established based on the health early warning signal, and the control risk factors are used as collaborative control risk penalty items for strategy compensation management of collaborative charging and discharging strategies.

[0014] Secondly, this application also provides a collaborative control system for a hybrid energy storage system, used to execute the collaborative control method for a hybrid energy storage system as described in the first aspect, wherein the collaborative control system for the hybrid energy storage system includes: a real-time status acquisition module, used to perform real-time status acquisition of electrochemical energy storage units and gravity energy storage units using a distributed sensor network, and establish a real-time status dataset; a predictive analysis module, used to read time-series external environmental data and grid load demand data, and perform predictive analysis of a predictive model based on the time-series external environmental data and grid load demand data, and establish time-series energy storage demand data; an objective function establishment module, used to establish a collaborative scheduling objective function, the collaborative scheduling objective function including a demand adaptation term, a health impact term of energy storage units, an energy storage cost adaptation term, and an energy storage switching penalty term; a multi-objective scheduling optimization module, used to perform multi-objective scheduling optimization based on the time-series energy storage demand data and the real-time status dataset using the collaborative scheduling objective function, and establish a multi-objective scheduling optimization result; and a strategy adjustment module, used to dynamically adjust the collaborative charging and discharging strategy of the electrochemical energy storage unit and the gravity energy storage unit based on the multi-objective scheduling optimization result.

[0015] Thirdly, this application also provides an electronic device, comprising: at least one processor; and a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of the cooperative control method for a hybrid energy storage system described in any of the first aspects above.

[0016] One or more technical solutions provided in this application have at least the following technical effects or advantages: By utilizing a distributed sensor network to collect real-time status data of electrochemical and gravity energy storage units, a real-time status dataset is established. Time-series external environmental data and grid load demand data are read, and predictive analysis is performed using a predictive model based on these data to establish time-series energy storage demand data. A collaborative scheduling objective function is established, including demand adaptation, energy storage unit health impact, energy storage cost adaptation, and energy storage switching penalty. Multi-objective scheduling optimization is performed based on the time-series energy storage demand data and the real-time status dataset using the collaborative scheduling objective function, establishing the multi-objective scheduling optimization result. The collaborative charging and discharging strategies of the electrochemical and gravity energy storage units are dynamically adjusted based on the multi-objective scheduling optimization result. In other words, by integrating a distributed sensor network to collect real-time status data of electrochemical and gravity energy storage units, and using a predictive model combined with time-series external environmental data and grid load demand data to predict energy storage demand, a multi-objective dynamic optimization mechanism based on real-time status and time-series prediction is established, achieving real-time collaborative control of electrochemical and gravity energy storage, thereby significantly improving the overall performance of the hybrid energy storage system.

[0017] The above description is merely an overview of the technical solution of this application. To better understand the technical means of this application and to facilitate its implementation according to the description, and to make the above and other objects, features, and advantages of this application more apparent, specific embodiments of this application are described below. It should be understood that the content described in this section is not intended to identify key or important features of the embodiments of this application, nor is it intended to limit the scope of this application. Other features of this application will become readily apparent through the following description. Attached Figure Description

[0018] To more clearly illustrate the technical solutions in 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 merely exemplary. For those skilled in the art, other drawings can be obtained based on the provided drawings without creative effort.

[0019] Figure 1 This is a flowchart illustrating the collaborative control method for hybrid energy storage systems used in this application.

[0020] Figure 2 This is a schematic diagram of the collaborative control system used in the hybrid energy storage system of this application.

[0021] Figure 3 This is a schematic diagram of the structure of an exemplary electronic device of this application.

[0022] Figure labeling: Real-time status acquisition module 11, Predictive analysis module 12, Objective function establishment module 13, Multi-objective scheduling optimization module 14, Strategy adjustment module 15. Detailed Implementation

[0023] This application addresses the technical problem in existing technologies where static, isolated control rules prevent real-time optimization of the coordinated scheduling of hybrid energy storage in dynamically changing grid environments, thus affecting the overall performance of the hybrid energy storage system. By providing a coordinated control method, system, and equipment for hybrid energy storage systems, this application solves the problem. Real-time status acquisition of electrochemical and gravity energy storage units is achieved through an integrated distributed sensor network. Energy storage demand is predicted using a predictive model combined with time-series external environmental data and grid load demand data. A multi-objective dynamic optimization mechanism based on real-time status and time-series prediction is established, enabling real-time coordinated control of electrochemical and gravity energy storage, thereby significantly improving the overall performance of the hybrid energy storage system.

[0024] The technical solutions of this application will now be clearly and completely described with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of them. It should be understood that this application is not limited to the exemplary embodiments described herein. 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. It should also be noted that, for ease of description, only the parts related to this application are shown in the accompanying drawings, not all of them.

[0025] Example 1, please refer to the appendix. Figure 1 This application provides a collaborative control method for a hybrid energy storage system, wherein the collaborative control method for a hybrid energy storage system is executed by a collaborative control system for a hybrid energy storage system, and the collaborative control method for a hybrid energy storage system specifically includes the following steps: A distributed sensor network is used to perform real-time status acquisition of electrochemical energy storage units and gravity energy storage units, and a real-time status dataset is established.

[0026] Specifically, sensors are deployed across the electrochemical energy storage subsystem and the gravity energy storage subsystem according to equipment zones, constructing a distributed sensor network. This distributed sensor network is an aggregate of various sensor types dispersed across the electrochemical energy storage unit, the gravity energy storage unit, and key power electronic interfaces. Electrochemical energy storage units typically refer to devices such as lithium-ion batteries and flow batteries that store and release electrical energy through chemical reactions. Gravity energy storage units are mechanical energy storage systems that store energy by electrically lifting a heavy object to a higher position and then releasing the energy by lowering the object to drive a generator.

[0027] The real-time status of electrochemical and gravity energy storage units is continuously measured and acquired at extremely high frequencies through a distributed sensor network, ensuring timestamp synchronization, preliminary cleaning, and real-time reporting. This data characterizes the operating condition of the equipment at any given moment, resulting in a real-time status dataset. The collected real-time status data undergoes clock synchronization, data packet formatting, local preliminary cleaning, and feature extraction to obtain the final real-time status dataset.

[0028] For example, voltage, current, and temperature sensors are deployed at key locations in the electrochemical energy storage unit to monitor its operating status at millisecond-level frequencies, simultaneously calculating key indicators such as state of charge and health status. For instance, the terminal voltage of module A is 735.5V, the charging / discharging current is -680.2A (negative sign indicates discharging), and the cell surface temperature is 28.7℃. Based on this, key status indicators are calculated and reported: SOC is 65.2%, SOH is 92.5%, and the current maximum allowable charging / discharging power is limited to 85MW for charging and 95MW for discharging. In the gravity energy storage unit, the position, mass, and motor torque of the load are accurately captured at second-level frequencies using a height encoder, weight sensor, and torque sensor, thereby calculating its energy storage and output status. For example, the current height of the load is 75 meters, the total height is 150 meters, the total mass of the load is a constant value of 2,000,000 kg, and the motor torque is +125 kN·m (positive sign indicates lifting). Based on this data, the gravity energy storage controller calculates the current potential energy storage of 400 MWh. (75m / 150m) = 200MWh, current actual output / input power is +60MW (indicating charging / lifting heavy objects). Real-time status acquisition is performed through a distributed sensor network to obtain accurate and real-time energy storage unit status information.

[0029] Read time-series external environment data and power grid load demand data, perform predictive analysis of the prediction model based on the time-series external environment data and power grid load demand data, and establish time-series energy storage demand data.

[0030] Furthermore, this application also includes the following steps: using the output analysis layer in the prediction model to evaluate stable power supply through time-series external environmental data, and establishing a time-series stable output; performing a time-series comparison of grid load demand data and time-series stable output according to the supply-demand comparison layer, and establishing time-series energy storage demand data based on the time-series residual, wherein the time-series energy storage demand data characterizes the energy storage supply compensation demand of a hybrid energy storage system with time nodes.

[0031] Specifically, time-series external environmental data and grid load demand data are obtained from meteorological service centers and power grid dispatch centers. The time-series external environmental data is a sequence of external factors affecting renewable energy generation output, arranged chronologically, including irradiance, ambient temperature, and wind speed. The grid load demand data represents the total power demand of all users in the grid, also a time-varying sequence, including total system load, peak-valley variation rate, load forecast curve, and its actual measurement curve.

[0032] The prediction model inputs time-series external environmental data and invokes the output analysis layer. The prediction model is an algorithmic model trained on historical data to predict future data, including an input layer, a time-series feature extraction layer, an output analysis layer, and a supply-demand comparison layer.

[0033] Temporal features are extracted using an LSTM network for stability assessment. The LSTM model has three hidden layers, each with 64 neurons. The input dimension is (T=96, number of features=3), representing 96 time points per day, each including three features: irradiance, wind speed, and temperature. The output is a stable power supply. The stable power supply and grid load demand data are input to the supply-demand comparison layer. The predicted stable power supply output is compared with the load forecast curve corresponding to the grid load demand data, subtracting each time point to calculate the time-series residual, thus depicting the real-time gap between the grid's own generation capacity and electricity demand. A positive time-series residual indicates a power shortage, requiring energy storage discharge or other power source compensation; a negative residual indicates a power surplus, which can be used for energy storage charging or discarded.

[0034] The time-series energy storage demand data is established based on the time-series residuals. This involves arranging the time-series residuals over time to form a time-series energy storage demand dataset. This dataset clearly defines the power and duration that the energy storage system needs to charge or discharge at each time point, thus determining when charging (absorbing excess power) and discharging (making up for power shortages) is required, along with precise instructions on the specific power values. For example, reading forecast data for the next 24 hours at 15-minute intervals, if the predicted irradiance at 12:00 noon tomorrow is 800 W / m², the wind speed is 6.5 m / s, and the grid load demand is 500 MW, the forecast model, based on the time-series external environmental data, calls the built-in photovoltaic and wind power prediction algorithms to calculate that at 12:00, the expected output of the photovoltaic power station is 180 MW, and the expected output of the wind farm is 120 MW. The time-series stable output at this point is 180 MW + 120 MW = 300 MW. This outputs a time-series stable output curve for the next 24 hours, with points every 15 minutes. At 12:00, a supply-demand comparison is performed, specifically the difference between the grid load demand data and the time-series stable output: 500MW - 300MW = +200MW. This means that at 12:00, the grid has a power deficit of 200MW. The generated time-series energy storage demand data for 12:00 will show a value of +200MW, explicitly instructing the hybrid energy storage system to provide 200MW of discharge power at this time to balance the grid. Conversely, if at 2:00 AM, the load is 300MW and the renewable energy output is 400MW, the residual is -100MW. The time-series energy storage demand data at this time will be -100MW, instructing the energy storage system to provide 100MW of charging power.

[0035] By reading and analyzing time-series external environmental data and grid load demand data, future grid load demand can be predicted, and the impact of the external environment on power supply stability can be assessed. The establishment of supply-demand comparisons and time-series energy storage demand data enables hybrid energy storage systems to accurately identify the points in time when additional energy storage supply is needed for compensation. This optimizes the scheduling and operation of energy storage units, helps improve grid stability and reliability, reduces power outages and voltage fluctuations, and simultaneously improves the utilization efficiency of the energy storage system.

[0036] A collaborative scheduling objective function is established, which includes a demand adaptation term, a health impact term for energy storage units, an energy storage cost adaptation term, and an energy storage switching penalty term.

[0037] Furthermore, this application also includes the following steps: the demand adaptation term of the cooperative scheduling objective function includes power demand adaptation and response rate adaptation, and the calculation formula of the demand adaptation term is as follows: ;in, Characteristic requirements adaptation items, The total time step, Represents the current time point, Characterizing grid load demand at time nodes Power requirements, Characterizing hybrid energy storage systems at time nodes 'output power' Characterizing hybrid energy storage systems at time nodes 'output power' , These are the weighting factors for power demand adaptation and response rate adaptation, respectively.

[0038] Furthermore, this application also includes the following steps: the calculation formulas for the health impact item and energy storage switching penalty item of the energy storage unit are as follows: in, Characterizing health impact items, Representation at time nodes The health impact values ​​mapped by the electrochemical charge-discharge depth, Representation at time nodes Electrochemical output power, Characterizing the rated discharge power of an electrochemical energy storage unit, Representation at time nodes The health impact values ​​mapped by gravity charge and discharge depth. To be at the time node Gravity output power, Characterizing the rated discharge power of the gravity energy storage unit, and Weighting factors characterizing the charge / discharge depth terms for electrochemical and gravitational forces, respectively. and Weighting factors characterize the discharge power terms of electrochemical and gravitational forces, respectively. in, As a penalty for switching energy storage, For switching indicator functions, The switching penalty coefficient is defined as follows. Specifically, a cooperative scheduling objective function is established to evaluate the cooperative control strategy. The cooperative scheduling objective function includes a demand adaptation term, a health impact term for energy storage units, an energy storage cost adaptation term, and an energy storage switching penalty term. The demand adaptation term measures the degree to which the actual output power of the hybrid energy storage system meets the grid demand; the health impact term for energy storage units quantifies the impact of the control strategy on the lifespan loss of electrochemical and gravity energy storage units, proactively extending equipment lifespan; the energy storage cost adaptation term incorporates operating costs into the optimization objective; and the energy storage switching penalty term reduces the frequent switching between charging / discharging and standby states of the hybrid energy storage system, improving operational stability and reducing equipment wear.

[0039] Demand adaptation includes power demand adaptation and response rate adaptation. The calculation formula for demand adaptation is as follows: ;in, The demand adaptation term is the time average of power demand adaptation and response rate adaptation over all time steps. The goal is to minimize it to balance tracking accuracy and output smoothness. The total time step, Represents the current time point, Characterizing grid load demand at time nodes Power requirements, Characterizing hybrid energy storage systems at time nodes 'output power' Characterizing hybrid energy storage systems at time nodes 'output power' , These are the weighting factors for power demand adaptation and response rate adaptation, respectively. The larger the value, the more likely it is to reduce power deviation; The larger the value, the more likely it is to allow the hybrid energy storage system to output smoothly, reducing sharp increases or decreases in power.

[0040] This represents the relative power error at time point t, calculated as the relative error between the actual output power of the energy storage system and the power demand of the grid at a given moment. Taking the absolute value is to penalize both under-discharge and over-discharge. Dividing by P... load (t) is used for normalization so that the deviation rate is a dimensionless value, making it easier to combine with other terms. Used to calculate the power change of an energy storage system between two adjacent moments, also using Normalization is performed to make the power output curve of the energy storage system as smooth as possible, avoiding impacts on its own equipment and the power grid. The average adaptation cost for the entire cycle is obtained by averaging the power matching deviation rate and power change stability rate at all time points within a scheduling cycle.

[0041] For example, suppose we are analyzing a 4-hour scheduling cycle, i.e., T=4, and set the weighting factor to be... The grid load demand at t=1 is +50MW, and the hybrid energy storage system output power is +48MW. The hybrid energy storage system output power at the previous time step was 0 (initial state). At t=2, the grid load demand is +150MW, and the hybrid energy storage system output power is +155MW. The hybrid energy storage system output power at the previous time step was +48MW. At t=3, the grid load demand is +80MW, and the hybrid energy storage system output power is +80MW. The hybrid energy storage system output power at the previous time step was +155MW. At t=4, the grid load demand is +120MW, and the hybrid energy storage system output power is +118MW. The hybrid energy storage system output power at the previous time step was +80MW. The demand adaptation term at t=1 is 0.7. 0.04 + 0.3 0.96 = 0.316, the demand adaptation item for t=2 = 0.7 0.033 + 0.3 0.713 = 0.237, so the demand adaptation factor for t=3 is 0.7. 0+0.3 0.938 = 0.281; the demand adaptation term at t=4 is 0.107, and the final calculated total demand adaptation term is 0.235. The cost is highest at t=1, mainly because starting from a standstill, the power change is very drastic, even though the power matching is good. At t=3, although the power matching is perfect, the power was very high in the previous time step, and the power is significantly reduced in this time step, resulting in a higher stability penalty. This value of 0.235 will be added to the calculation results of the health impact term, cost term, etc., as the basis for evaluating the overall performance of the scheduling scheme. The optimization algorithm will attempt to find another scheme that makes this sum smaller.

[0042] The formula for calculating the health impact of an energy storage unit is as follows:

[0043] in, The smaller the value of the health impact term, the less the scheduling strategy will damage the lifespan of the energy storage unit; Representation at time nodes The health impact value mapped by the electrochemical charge-discharge depth is a number between 0 and 1, which is usually related to the state of charge (SOC). DoD is approximately equal to 1-SOC. The lower the SOC (i.e., the higher the DoD), the larger this value is, indicating greater damage to the battery's health. Representation at time nodes The electrochemical output power is positive when discharging and negative when charging; The rated discharge power characterizing the electrochemical energy storage unit is a fixed value used to normalize the actual power. Representation at time nodes The health impact values ​​mapped by gravity charge and discharge depth. To be at the time node Gravity output power, The rated discharge power of the gravity energy storage unit is a fixed value used to normalize the actual power. and Weighting factors characterizing the charge / discharge depth terms for electrochemical and gravitational forces, respectively. and Weighting factors that characterize the discharge power terms of electrochemical and gravitational forces, respectively. and The larger the size, the more inclined the battery is to operate in a shallow charge and discharge state. and The larger the value, the more inclined the battery is to avoid discharging at high power rates. The first part of the formula is used to calculate the health impact of electrochemical energy storage units, and the second part is used to calculate the health impact of gravity-based energy storage units.

[0044] The formula for calculating the energy storage switching penalty is as follows: ;in, This is a penalty term for energy storage switching. The smaller this value, the fewer state switches the energy storage unit will undergo under this scheduling strategy, and the more stable the operation will be. The toggle indicator function is a binary function (0 or 1) that is applied at a specific time point. When the control state of an energy storage unit undergoes a qualitative change, its value is 1; otherwise, it is 0. A qualitative change usually refers to switching from a charging state to a discharging state, or vice versa; switching from a shutdown / standby state to a charging or discharging state; or it may be considered a switch when the power command changes beyond a certain large threshold. The switching penalty coefficient is the fixed cost incurred each time a state switch occurs. The larger the value, the more sensitive the user is to the switching action and the more likely they are to avoid switching. For example, setting the switching penalty coefficient to 0.5, and analyzing the same 4-hour scheduling scheme, focusing on the command state changes of the electrochemical energy storage unit, from t=0 to t=1, switching from standby to discharge +30MW is one switch, switch(1)=1; from t=1 to t=2, continuous discharge increases the power from 30MW to 75MW, but no qualitative change occurs, switch(2)=0; from t=2 to t=3, continuous discharge decreases the power from 75MW to 10MW, but no qualitative change occurs, switch(3)=0; from t=3 to t=4, continuous discharge increases the power from 10MW to 50MW, but no qualitative change occurs, switch(4)=0. In this scheme, there is only one state switch, and the final calculated energy storage switching penalty term is 0.5. 1+038 0+0.5 0+0.5 0 = 0.5. If an alternative causes the electrochemical energy storage to switch back and forth between charging and discharging 3 times within 4 hours, then its energy storage switching penalty is 1.5.

[0045] The energy storage cost adaptation term is used to evaluate the cost of strategy implementation, typically related to electricity prices, aiming to achieve low charging and high discharging, saving electricity costs or generating revenue. Specific mathematical formulas and weighting coefficients are defined for each sub-item to reflect the relative importance of different objectives. The collaborative scheduling objective function = demand adaptation term + energy storage unit health impact term + energy storage cost adaptation term + energy storage switching penalty term. For example, if the demand adaptation term is calculated to be 0.235, the health impact term to be 0.28, the cost adaptation term to be -10.56 (negative values ​​represent revenue), and the switching penalty term to be 0.5, then the total objective function value is calculated to be -9.545. This strategy is negative, mainly because the discharging revenue is very high. The optimization algorithm compares the J_total of different strategies and finds the solution with the smaller value. A strategy that reduces battery output at t=4 and increases gravity energy storage output might slightly increase the health impact factor. However, if it can increase the negative value of the cost adaptation factor (making it more economical) or reduce the switching penalty factor, the new overall objective function value might be lower, thus becoming a better strategy. By simultaneously optimizing demand adaptation, health impact, cost, and switching penalty factors, and balancing the trade-offs between accuracy, system health, economy, and frequent switching based on actual conditions, a comprehensively optimized scheduling strategy is obtained.

[0046] The collaborative scheduling objective function is used to perform multi-objective scheduling optimization based on time-series energy storage demand data and real-time status dataset, and the multi-objective scheduling optimization result is established.

[0047] Furthermore, this application also includes the following steps: acquiring pre-control data; establishing an initial population based on the pre-control data, wherein each solution in the initial population represents a cooperative control strategy; using the time-series energy storage demand data and the real-time state dataset as constraints, performing an adaptation evaluation of the initial population based on the cooperative scheduling objective function, and establishing a mapping adaptation value; using the mapping adaptation value to perform population iterative analysis of the initial population, and performing iterative optimization management to complete multi-objective scheduling optimization.

[0048] Furthermore, this application also includes the following steps: establishing a periodic selection strategy for population iteration, the periodic selection strategy including a random search period, an elite retention period, and a convergent search period; obtaining the current iteration round, calling the periodic selection strategy according to the iteration round, generating a retention threshold and a random factor for elite individuals; performing a selection operation according to the retention threshold, the random factor, and the mapping fit value, and using the selection results to perform crossover iteration to update the population.

[0049] Specifically, pre-control data refers to preliminary information obtained before optimized scheduling, used to provide initial conditions for optimizing the coordinated control strategy. This includes, but is not limited to, grid load demand, the current state of the energy storage system, external environmental data, and the health status of system equipment. Based on the pre-control data, an initial population is generated. This initial population, containing dozens of candidate strategies, can be generated by introducing random perturbations. Each solution in the initial population represents a complete coordinated control strategy, that is, a detailed sequence of power output commands for the electrochemical and gravity energy storage units over a future period.

[0050] Time-series energy storage demand data and real-time status datasets are used as constraints, i.e., inviolable hard constraints. For example, the total output power of the hybrid energy storage system must meet the time-series energy storage demand; the output of each energy storage unit cannot exceed its real-time maximum allowable power; and the State of Charge (SOC) must be within a safe range. A comprehensive adaptation evaluation of each strategy in the population is performed using the established cooperative scheduling objective function, calculating the mapping adaptation value corresponding to each strategy, thus completing the adaptation evaluation of the initial population. The mapping adaptation value is calculated by substituting each solution (i.e., control strategy) in the initial population into the cooperative scheduling objective function. This function value quantifies and maps the overall performance of the strategy; the smaller the value, the better the strategy and the more it meets the expected goal.

[0051] A periodic selection strategy for population iteration is established to determine the time period for selecting different population update strategies during the optimization process. Different search focuses are adopted at different optimization stages, including a random search period, an elite retention period, and a convergent search period. The random search period focuses on exploration, broadly searching for new solutions globally to avoid getting trapped in local optima; the elite retention period focuses on utilization, strongly preserving the currently best individuals in the population to ensure that excellent genes are not lost; the convergent search period focuses on refined search, performing local mining in the vicinity of the current best solution to seek a more accurate optimal solution.

[0052] The selection strategy is dynamically invoked based on the current iteration round. In the early stages, a random search round is used to broadly explore the solution space; in the middle stages, an elite retention round is used to consolidate excellent results; and in the later stages, a convergent search round is used for local fine-tuning. Based on this, the retention threshold and random factor for each round are dynamically generated. The retention threshold is a proportional parameter used in the selection operation to determine which solutions are elite solutions, i.e., excellent candidate solutions, and should be retained for use in subsequent iterations. The random factor is a probability parameter controlling randomness, used to determine whether an individual is directly retained, replaced by crossover, or replaced by a random new solution. Selection operations are performed based on the retention threshold and mapping fitness value to select parent individuals, and genetic operators such as crossover iteration are used to generate a new generation of the population. In other words, in each iteration, the population is optimized through crossover iteration to generate new individuals, and the retention threshold and random factor are used to determine whether to retain the current solution. The crossover iteration operation simulates the natural selection process, and the generated new solution may be better than the original solution. The selection result is evaluated by the mapping fitness value, selecting individuals with better fitness to enter the next generation. In each iteration, the population is optimized through crossover iteration, generating new individuals. A retention threshold and a random factor are used to determine whether to retain the current solution. The crossover iteration operation simulates the natural selection process; the generated new solution may be better than the original solution, until the convergence condition is met, ultimately outputting a multi-objective scheduling optimization result with optimal overall performance. Iterative optimization management refers to the mechanism for monitoring and controlling the entire population iteration process, including determining whether termination conditions have been met, such as the maximum number of iterations or the solution quality no longer significantly improving, and finally outputting the optimization result.

[0053] For example, based on prior control data, 50 strategies are randomly generated. The fitness of the initial population is evaluated using a cooperative scheduling objective function, establishing a mapped fitness value. From these 50 individuals, the top 10 with the best fitness values ​​are directly copied to the next generation. The positions of the remaining 40 new individuals are generated through crossover iterations. For instance, the 1st and 3rd ranked individuals are selected as parents, and their power commands are exchanged at a random time point to generate offspring C. Based on a random factor, 4 positions are replaced by completely random new strategies to maintain diversity. The new generation population consists of 10 elites + 36 crossover offspring + 4 random new solutions, totaling 50 individuals. The algorithm repeats the above process. By the 95th round, a convergence search cycle begins, the retention threshold is increased to 40%, and the random factor is reduced to 0.02 for fine-tuning. When the fitness value of the best individuals improves by less than 0.1% over 10 consecutive generations, the iterative optimization management determines convergence and stops the computation. Finally, the individual with the smallest fitness value in the 100th generation is output as the multi-objective scheduling optimization result.

[0054] Through iterative optimization of the initial population, the final control strategy can accurately match the grid load demand, reduce power deviation, and improve grid stability. Through multi-objective optimization, the scheduling strategy accurately matches the grid load demand, reduces power fluctuations, and improves grid stability. By introducing a periodic selection strategy and an elite retention mechanism, the optimization process ensures the stability and feasibility of the hybrid energy storage system under different operating conditions.

[0055] Based on the multi-objective scheduling optimization results, the coordinated charging and discharging strategy of the electrochemical energy storage unit and the gravity energy storage unit is dynamically adjusted.

[0056] Furthermore, this application also includes the following steps: controlling and monitoring the electrochemical energy storage unit and the gravity energy storage unit to establish a real control response; using the real control response to evaluate the deviation of the coordinated charging and discharging strategy and establishing deviation evaluation feedback; establishing a dynamic compensation strategy based on the deviation evaluation feedback and using the dynamic compensation strategy to dynamically compensate the coordinated charging and discharging strategy.

[0057] Specifically, the multi-objective scheduling optimization results are distributed to the local controllers of the electrochemical and gravity energy storage units as the baseline plan for their execution. Control monitoring is activated to continuously collect the actual output power and status of the electrochemical and gravity energy storage units, forming a real control response dataset. The real control response refers to the actual response exhibited by the electrochemical and gravity energy storage units during actual operation, based on the execution of control commands. This includes the actual charging and discharging power, delay, and excessive fluctuations of the energy storage units, reflecting the actual operating status of the hybrid energy storage system.

[0058] Deviation evaluation of the coordinated charging and discharging strategy is performed using the actual control response. This involves comparing the actual data with the planned values ​​in the optimization results in real time, executing the deviation evaluation, and generating deviation evaluation feedback. The deviation evaluation feedback is a quantitative output of the deviation evaluation result, typically a power deviation value or a state deviation indicator.

[0059] A dynamic compensation strategy is established based on deviation evaluation feedback, typically following a certain logic. This includes prioritizing the use of faster-responding electrochemical energy storage for compensation, ensuring the compensation amount is within its real-time capacity, and generating compensation commands. For example, the dynamic compensation strategy rule might be set to prioritize the use of the fastest-responding electrochemical energy storage for power deviation compensation, but the total power after compensation must not exceed its current maximum safe output limit (set at 100MW). Through dynamic compensation operations, compensation commands are superimposed on the original planned commands, correcting the behavior of the energy storage units in real time, thereby ensuring that the final output of the entire hybrid energy storage system closely matches the initial optimization target. For example, if the energy storage system switches too frequently, the magnitude of these changes is reduced, lowering the stress on batteries and mechanical components. Real-time monitoring and rapid compensation effectively overcome the disconnect between planning and execution caused by factors such as equipment response delays, model errors, or internal losses, ensuring that grid demand is accurately met.

[0060] Furthermore, this application also includes the following steps: using the real-time status dataset to perform a health evaluation of the hybrid energy storage system and establish a health early warning signal; establishing a control risk factor based on the health early warning signal, and using the control risk factor as a collaborative control risk penalty item for strategy compensation management of the collaborative charging and discharging strategy.

[0061] Specifically, a health assessment of the hybrid energy storage system is performed using a real-time status dataset. When a health indicator of a unit exceeds a preset threshold, a corresponding health warning signal is generated. The health warning signal is the output of the health assessment and is a qualitative or quantitative alert. It can be a simple binary flag (0 for normal, 1 for warning) or a level indicating the degree of health deterioration, such as normal, caution, warning, danger, etc.

[0062] Based on health warning signals, control risk factors are established. These risk factors are quantifiable values, typically between 0 and 1, and are mapped from the health warning signals. They characterize the degree of risk that continuing the original high-intensity control strategy might lead to malfunctions or accelerated aging under the current health condition. A higher risk factor indicates a higher risk.

[0063] The risk factor is introduced as a dynamic coefficient of the collaborative control risk penalty term into the original collaborative scheduling objective function. It is typically the product of the risk factor and the planned output of unhealthy units. This causes any high-load command assigned to that unhealthy unit to result in a sharp increase in the objective function value. When the energy storage unit is in poor health, this term significantly increases the objective function value, thereby penalizing control strategies that impose high load demands on that unhealthy unit.

[0064] By incorporating a risk penalty term for collaborative control, the collaborative charging and discharging strategy is adjusted and compensated, enabling it to automatically avoid high-risk operations. This results in the final collaborative charging and discharging strategy, dynamically adjusting the collaborative charging and discharging strategies of the electrochemical and gravity energy storage units. By directly feeding back the health status to the control core, the system proactively reduces operational stress in the early stages of equipment degradation, effectively delaying performance decline, preventing potential failures, achieving predictive maintenance, and significantly extending the overall lifespan of the hybrid energy storage system. Actively managing equipment health significantly reduces the high replacement and maintenance costs caused by premature equipment failure or unexpected downtime. Even with performance degradation in some units, the system can maintain normal overall function through strategy compensation, improving the resilience of the hybrid energy storage system to internal disturbances such as equipment aging.

[0065] In summary, the collaborative control method for hybrid energy storage systems provided in this application has the following technical effects: It establishes a real-time status dataset by utilizing a distributed sensor network to collect the real-time status of electrochemical and gravity energy storage units; it reads time-series external environmental data and grid load demand data, and performs predictive analysis using a predictive model based on the time-series external environmental data and grid load demand data to establish time-series energy storage demand data; it establishes a collaborative scheduling objective function, which includes a demand adaptation term, a health impact term for energy storage units, an energy storage cost adaptation term, and an energy storage switching penalty term; it uses the collaborative scheduling objective function to perform multi-objective scheduling optimization based on the time-series energy storage demand data and the real-time status dataset to establish multi-objective scheduling optimization results; and it dynamically adjusts the collaborative charging and discharging strategies of the electrochemical and gravity energy storage units based on the multi-objective scheduling optimization results. In other words, by integrating a distributed sensor network, real-time status acquisition of electrochemical energy storage units and gravity energy storage units is achieved. Energy storage demand is predicted by using a prediction model combined with time-series external environmental data and grid load demand data. By establishing a multi-objective dynamic optimization mechanism based on real-time status and time-series prediction, real-time coordinated control of electrochemical energy storage and gravity energy storage is realized, thereby significantly improving the overall performance of the hybrid energy storage system.

[0066] Example 2: Based on the same inventive concept as the cooperative control method for hybrid energy storage systems in Example 1, this application also provides a cooperative control system for hybrid energy storage systems. Please refer to the appendix. Figure 2 The collaborative control system for the hybrid energy storage system includes: The system comprises the following modules: a real-time status acquisition module 11, which uses a distributed sensor network to acquire the real-time status of the electrochemical energy storage unit and the gravity energy storage unit, and establishes a real-time status dataset; a prediction and analysis module 12, which reads time-series external environmental data and grid load demand data, performs prediction analysis based on the time-series external environmental data and grid load demand data, and establishes time-series energy storage demand data; an objective function establishment module 13, which establishes a collaborative scheduling objective function, including a demand adaptation term, a health impact term for the energy storage unit, an energy storage cost adaptation term, and an energy storage switching penalty term; a multi-objective scheduling optimization module 14, which uses the collaborative scheduling objective function to perform multi-objective scheduling optimization based on the time-series energy storage demand data and the real-time status dataset, and establishes a multi-objective scheduling optimization result; and a strategy adjustment module 15, which dynamically adjusts the collaborative charging and discharging strategy of the electrochemical energy storage unit and the gravity energy storage unit based on the multi-objective scheduling optimization result.

[0067] Furthermore, the predictive analysis module 12 in the collaborative control system for the hybrid energy storage system is also used to: use the output analysis layer in the predictive model to evaluate the stable power supply through time-series external environmental data and establish a time-series stable output; perform time-series comparison of grid load demand data and time-series stable output according to the supply-demand comparison layer, and establish time-series energy storage demand data based on the time-series residual, wherein the time-series energy storage demand data characterizes the energy storage supply compensation demand of the hybrid energy storage system with time nodes.

[0068] Furthermore, the objective function establishment module 13 in the collaborative control system for the hybrid energy storage system is also used for: the demand adaptation item of the collaborative scheduling objective function includes power demand adaptation and response rate adaptation, and the calculation formula of the demand adaptation item is as follows: ;in, Characteristic requirements adaptation items, The total time step, Represents the current time point, Characterizing grid load demand at time nodes Power requirements, Characterizing hybrid energy storage systems at time nodes Output power Characterizing hybrid energy storage systems at time nodes 'output power' , These are the weighting factors for power demand adaptation and response rate adaptation, respectively.

[0069] Furthermore, the objective function establishment module 13 in the collaborative control system for the hybrid energy storage system is also used for: the calculation formulas for the health impact term and energy storage switching penalty term of the energy storage unit are as follows:

[0070] in, Characterizing health impact items, Representation at time nodes The health impact values ​​mapped by the electrochemical charge-discharge depth, Representation at time nodes Electrochemical output power, Characterizing the rated discharge power of an electrochemical energy storage unit, Representation at time nodes The health impact values ​​mapped by gravity charge and discharge depth. To be at the time node Gravity output power, Characterizing the rated discharge power of the gravity energy storage unit, and Weighting factors characterizing the charge / discharge depth terms for electrochemical and gravitational forces, respectively. and Weighting factors characterize the discharge power terms of electrochemical and gravitational forces, respectively. in, As a penalty for switching energy storage, For switching indicator functions, To switch the penalty coefficient.

[0071] Furthermore, the multi-objective scheduling optimization module 14 in the collaborative control system for the hybrid energy storage system is also used for: acquiring pre-control data; establishing an initial population based on the pre-control data, wherein each solution in the initial population represents a collaborative control strategy; using the time-series energy storage demand data and the real-time state dataset as constraints, performing an adaptation evaluation of the initial population based on the collaborative scheduling objective function, and establishing a mapping adaptation value; and using the mapping adaptation value to perform population iterative analysis of the initial population and perform iterative optimization management to complete the multi-objective scheduling optimization.

[0072] Furthermore, the multi-objective scheduling optimization module 14 in the collaborative control system for the hybrid energy storage system is also used to: establish a periodic selection strategy for population iteration, the periodic selection strategy including a random search period, an elite retention period, and a convergent search period; obtain the current iteration round, call the periodic selection strategy according to the iteration round, generate a retention threshold and a random factor for elite individuals; perform a selection operation according to the retention threshold, the random factor, and the mapping fit value, and perform cross-iteration using the selection results to update the population.

[0073] Furthermore, the strategy adjustment module 15 in the collaborative control system for the hybrid energy storage system is also used for: controlling and monitoring the electrochemical energy storage unit and the gravity energy storage unit to establish a real control response; using the real control response to evaluate the deviation of the collaborative charging and discharging strategy and establish deviation evaluation feedback; establishing a dynamic compensation strategy based on the deviation evaluation feedback and using the dynamic compensation strategy to dynamically compensate the collaborative charging and discharging strategy.

[0074] Furthermore, the strategy adjustment module 15 in the collaborative control system for the hybrid energy storage system is also used to: perform a health assessment of the hybrid energy storage system using the real-time status dataset and establish a health early warning signal; establish a control risk factor based on the health early warning signal, and use the control risk factor as a collaborative control risk penalty item for strategy compensation management of the collaborative charging and discharging strategy.

[0075] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on its differences from other embodiments. Figure 1 The collaborative control method and specific examples for hybrid energy storage systems in Example 1 are also applicable to the collaborative control system for hybrid energy storage systems in this embodiment. Through the foregoing detailed description of the collaborative control method for hybrid energy storage systems, those skilled in the art can clearly understand the collaborative control system for hybrid energy storage systems in this embodiment. Therefore, for the sake of brevity, it will not be described in detail here.

[0076] Example 3: Based on the same inventive concept as the collaborative control method for hybrid energy storage systems in Example 1, this application also provides an electronic device, including: at least one processor; a memory communicatively connected to the at least one processor; wherein the memory stores instructions executable by the at least one processor, the instructions being executed by the at least one processor to enable the at least one processor to perform the steps of the collaborative control method for hybrid energy storage systems described in any one of Examples 1.

[0077] Appendix Figure 3 This is a schematic diagram of the structure of an exemplary electronic device of this application. Figure 3In this document, the bus architecture is represented by bus 300. Bus 300 may include any number of interconnected buses and bridges, and bus 300 connects various circuits including one or more processors represented by processor 302 and memory represented by memory 304. Bus 300 may also connect various other circuits such as peripheral devices, voltage regulators, and power management circuits, which are well known in the art and therefore will not be described further herein. Bus interface 305 provides an interface between bus 300 and receiver 301 and transmitter 303. Receiver 301 and transmitter 303 may be the same element, i.e., a transceiver, providing a unit for communicating with various other devices over a transmission medium. Processor 302 is responsible for managing bus 300 and general processing, while memory 304 can be used to store data used by processor 302 during operation.

[0078] The above description of the disclosed embodiments enables those skilled in the art to make or use this application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of this application. Therefore, this application is not to be limited to the embodiments shown herein, but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

[0079] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of this application and its equivalents, this application also intends to include such modifications and variations.

Claims

1. A collaborative control method for hybrid energy storage systems, characterized in that, include: A distributed sensor network is used to perform real-time status acquisition of electrochemical energy storage units and gravity energy storage units, and a real-time status dataset is established. Read time-series external environment data and power grid load demand data, perform predictive analysis of the prediction model based on the time-series external environment data and power grid load demand data, and establish time-series energy storage demand data; A collaborative scheduling objective function is established, which includes a demand adaptation term, a health impact term for energy storage units, an energy storage cost adaptation term, and an energy storage switching penalty term. The collaborative scheduling objective function is used to perform multi-objective scheduling optimization based on time-series energy storage demand data and real-time status dataset, and the multi-objective scheduling optimization result is established. Based on the multi-objective scheduling optimization results, the coordinated charging and discharging strategy of the electrochemical energy storage unit and the gravity energy storage unit is dynamically adjusted.

2. The collaborative control method for a hybrid energy storage system as described in claim 1, characterized in that, The demand adaptation term of the cooperative scheduling objective function includes power demand adaptation and response rate adaptation, and the calculation formula of the demand adaptation term is as follows: ; in, Characteristic requirements adaptation items, The total time step, Represents the current time point, Characterizing grid load demand at time nodes Power requirements, Characterizing hybrid energy storage systems at time nodes 'output power' Characterizing hybrid energy storage systems at time nodes 'output power' , These are the weighting factors for power demand adaptation and response rate adaptation, respectively.

3. The collaborative control method for a hybrid energy storage system as described in claim 2, characterized in that, The calculation formulas for the health impact item and energy storage switching penalty item of the energy storage unit are as follows: in, Characterizing health impact items, Representation at time nodes The health impact values ​​mapped by the electrochemical charge-discharge depth, Representation at time nodes Electrochemical output power, Characterizing the rated discharge power of an electrochemical energy storage unit, Representation at time nodes The health impact values ​​mapped by gravity charge and discharge depth. To be at the time node Gravity output power, Characterizing the rated discharge power of the gravity energy storage unit, and Weighting factors characterizing the charge / discharge depth terms for electrochemical and gravitational forces, respectively. and Weighting factors characterizing the discharge power terms of electrochemical and gravitational forces, respectively; in, As a penalty for switching energy storage, For switching indicator functions, To switch the penalty coefficient.

4. The collaborative control method for a hybrid energy storage system as described in claim 1, characterized in that, The collaborative scheduling objective function is used to perform multi-objective scheduling optimization based on time-series energy storage demand data and real-time status datasets, including: Acquire pre-control data, establish an initial population based on the pre-control data, and each solution in the initial population represents a cooperative control strategy; Using the time-series energy storage demand data and the real-time status dataset as constraints, an initial population adaptation evaluation is performed based on the cooperative scheduling objective function, and a mapping adaptation value is established. The initial population is analyzed using the mapping adaptation values, and iterative optimization management is performed to complete multi-objective scheduling optimization.

5. The collaborative control method for a hybrid energy storage system as described in claim 4, characterized in that, Population iteration analysis of the initial population using the aforementioned mapping fitness values ​​includes: A periodic selection strategy for population iteration is established, which includes a random search period, an elite retention period, and a convergent search period. Obtain the current iteration round, select a strategy based on the iteration round call cycle, and generate the retention threshold and random factor for elite individuals; A selection operation is performed based on the retention threshold, the random factor, and the mapping fit value. The selection results are then used to perform crossover iterations to update the population.

6. The collaborative control method for a hybrid energy storage system as described in claim 1, characterized in that, Based on the multi-objective scheduling optimization results, the coordinated charging and discharging strategy of the electrochemical energy storage unit and the gravity energy storage unit is dynamically adjusted, including: The electrochemical energy storage unit and the gravity energy storage unit are controlled and monitored to establish a real control response. The deviation evaluation of the coordinated charging and discharging strategy is performed using the actual control response, and a deviation evaluation feedback is established. A dynamic compensation strategy is established based on the deviation evaluation feedback, and the dynamic compensation strategy is used to perform dynamic compensation for the coordinated charging and discharging strategy.

7. The collaborative control method for a hybrid energy storage system as described in claim 1, characterized in that, Based on the aforementioned time-series external environmental data and grid load demand data, a predictive analysis of the predictive model is performed to establish time-series energy storage demand data, including: The output analysis layer in the prediction model is used to evaluate the stable power supply through time-series external environmental data, and a time-series stable output is established. Based on the time-series comparison of grid load demand data and time-series stable output performed by the supply and demand comparison layer, time-series energy storage demand data is established based on the time-series residual. The time-series energy storage demand data characterizes the energy storage supply compensation demand of the hybrid energy storage system with time nodes.

8. The collaborative control method for a hybrid energy storage system as described in claim 1, characterized in that, Based on the multi-objective scheduling optimization results, the coordinated charging and discharging strategy of the electrochemical energy storage unit and the gravity energy storage unit is dynamically adjusted, including: The real-time status dataset is used to perform a health assessment of the hybrid energy storage system and establish a health early warning signal. Based on the health warning signal, a control risk factor is established, and the control risk factor is used as a collaborative control risk penalty item for strategy compensation management of collaborative charging and discharging strategy.

9. A collaborative control system for a hybrid energy storage system, characterized in that, The step of implementing the collaborative control method for a hybrid energy storage system according to any one of claims 1 to 8, wherein the collaborative control system for the hybrid energy storage system comprises: The real-time status acquisition module is used to perform real-time status acquisition of electrochemical energy storage units and gravity energy storage units using a distributed sensor network, and to establish a real-time status dataset. The predictive analysis module is used to read time-series external environment data and power grid load demand data, perform predictive analysis of the predictive model based on the time-series external environment data and power grid load demand data, and establish time-series energy storage demand data. The objective function establishment module is used to establish a collaborative scheduling objective function, which includes a demand adaptation item, a health impact item of the energy storage unit, an energy storage cost adaptation item, and an energy storage switching penalty item. The multi-objective scheduling optimization module is used to perform multi-objective scheduling optimization based on time-series energy storage demand data and real-time status dataset using the cooperative scheduling objective function, and to establish multi-objective scheduling optimization results. The strategy adjustment module is used to dynamically adjust the coordinated charging and discharging strategy of the electrochemical energy storage unit and the gravity energy storage unit based on the multi-objective scheduling optimization results.

10. An electronic device, characterized in that, include: At least one processor; A memory that is communicatively connected to the at least one processor; The memory stores instructions executable by the at least one processor, which are executed by the at least one processor to enable the at least one processor to perform the steps of the cooperative control method for a hybrid energy storage system according to any one of claims 1 to 8.

Images