A battery energy storage system for fast frequency regulation and a control method thereof

Abstract

The application relates to the technical field of safety control of electric power energy storage systems, and discloses a battery energy storage system for fast frequency modulation and a control method thereof. The system comprises a distributed power buffer module array for absorbing or releasing high-frequency fluctuation components in power instructions from a power grid dispatching end; a partitioned counter-flow liquid cooling subsystem, wherein cooling circuits of each liquid cooling partition adopt a counter-flow arrangement mode; and a global energy router for selecting part of battery clusters to execute the power instructions and synchronously adjusting the cooling intensity of the corresponding liquid cooling partitions according to the power instructions and the states of the power buffer modules. According to the application, external millisecond-level power impact is converted into smooth thermal load and periodic stress relaxation inside the battery through three defense lines, the problem of local thermal runaway of the electric power energy storage system in the fast frequency modulation scene is solved, and the safety and durability of the electric power energy storage system in long-term and high-intensity fast frequency modulation service are ensured.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of safety control of power energy storage systems, in particular to a battery energy storage system for fast frequency regulation and a control method thereof. BACKGROUND

[0002] With the increasing penetration of renewable energy and the growing demand for fast regulation resources in power systems, large-scale containerized lithium-ion battery energy storage systems have become the core equipment for providing services such as time-of-use price arbitrage, grid peak shaving and frequency regulation, and emergency backup power. For example, such systems can typically be integrated into a standard 20-foot container, which contains a large number of battery clusters, uses thermal insulation design to maintain a stable environment, and relies on efficient liquid cooling air conditioning systems such as three-stage liquid cooling pipelines to accurately control the operating temperature of the battery cells in the ideal range of 20-30℃, to ensure their cycle life and safety.

[0003] In actual operation, to stabilize the grid frequency, energy storage systems need to participate in auxiliary services with high requirements such as primary frequency regulation. For example, when the grid experiences a sudden frequency drop of about 0.2Hz due to the sudden trip of a large generator, the grid dispatching end will issue a power command requiring the energy storage power station to immediately switch from a near-full standby state to discharging at maximum power within hundreds of milliseconds to support the grid frequency. This scenario exposes the fundamental limitations of traditional battery thermal management systems: the heat generation rate of the battery can change dramatically on a millisecond scale, while the liquid cooling system, which relies on a large inertia circulating working medium, has a significant delay in temperature and flow regulation on a second scale. This serious mismatch between fast electrical heating and slow thermal cooling causes the battery to withstand high-frequency alternating thermal stress, and the direct consequence is that the micro-interface of the electrode material and the current collector will peel off due to fatigue, and the thermal interface between the battery cell and the liquid cooling plate will crack due to stress. The former will cause irreversible growth of internal resistance and local overheating, and the latter will directly lead to heat dissipation failure, both of which will cause the risk of local thermal runaway to rise sharply when the battery responds to subsequent power shocks, posing a serious threat to the safety of containerized energy storage systems with high integration.

[0004] The existing technology mainly relies on the post-incident thermal protection strategy of the battery management system, i.e., reducing power or increasing cooling intensity after monitoring that the battery temperature or temperature difference exceeds the standard. However, this mode of damage first and remedy later has inherent defects: first, temperature sensors usually monitor macroscopic average or local point temperature and cannot sense the mechanical fatigue accumulated at the micro-interface of the material due to rapid temperature fluctuations; second, the response has a lag and cannot prevent the occurrence of thermal shock; third, simple global derating will sacrifice the regulation performance and benefits of the system. Although some studies have attempted to use complex thermal models for predictive control, they have large computational loads, rely on accurate parameters, and lack robustness in actual operating environments, making it difficult to implement reliable online applications in the integrated controller of the containerized energy storage system.

[0005] Therefore, an innovative system architecture and control method is urgently needed in the industry to fundamentally decouple the instantaneous power demand of the power grid and the thermal shock on the battery body, ensuring the safety and durability of the power storage system in long-term and high-intensity fast frequency modulation service without sacrificing fast response capability. SUMMARY

[0006] To solve the above problems, the present application provides the following technical solutions:

[0007] In a first aspect, the present application discloses a battery energy storage system for fast frequency modulation, comprising a battery cluster array, a high-voltage tank and a power converter, the system comprising:

[0008] A distributed power buffer module array is arranged one-to-one corresponding to the battery clusters in the battery cluster array and electrically connected between the corresponding battery clusters and the high-voltage tank, for absorbing or releasing high-frequency fluctuation components in the power instruction from the power grid dispatching end;

[0009] A partitioned counter-flow liquid cooling subsystem comprises a liquid cooling partition independently arranged for each battery cluster, and the cooling circuit of each liquid cooling partition adopts a counter-flow arrangement mode with the inlet located at the top of the battery cluster and the outlet located at the bottom of the battery cluster;

[0010] A global energy router is communicatively connected to the distributed power buffer module array and the partitioned counter-flow liquid cooling subsystem, for selecting part of the battery clusters as service units to execute the power instruction and synchronously adjusting the cooling intensity of the corresponding liquid cooling partitions according to the power instruction and the state of each power buffer module.

[0011] Further, the distributed power buffer module is a flywheel-super capacitor composite module, which contains a flywheel energy storage unit and a super capacitor unit connected physically coaxially.

[0012] Further, the global energy router evaluates the real-time available capacity of each power buffer module according to the speed state of the flywheel energy storage unit or the voltage state of the super capacitor unit.

[0013] Further, the cooling circuits of each liquid cooling partition are connected in series through pipelines, so that the liquid cooling partition far from the high-voltage tank has its cooling liquid inlet connected to the cooling liquid outlet of the liquid cooling partition adjacent to the high-voltage tank, thereby forming a cooling liquid temperature gradient field in the arrangement direction of the battery cluster array.

[0014] Further, the battery cluster array is pre-divided into at least two logical groups; the global energy router is configured to control one logical group as a service unit and control the remaining logical groups in a thermal recovery state when responding to the power instruction, and periodically or according to the state to trigger rotation between the logical groups.

[0015] Further, the global energy router preferentially selects a logical group with high available capacity of the corresponding power buffer module and low inlet temperature of the corresponding liquid cooling subarea as the serving unit.

[0016] Further, when triggering logical group rotation, the global energy router dynamically adjusts the heat recovery duration of the logical group to which the battery cluster belongs according to the historical cumulative discharge depth or the highest temperature experienced by the battery cluster.

[0017] Further, the global energy router is configured to: when receiving a power instruction issued by a grid scheduling end, decompose a corresponding frequency segment in the power instruction into a high-frequency component and a low-frequency component, and preferentially allocate the high-frequency component to the buffer by the distributed power buffer module array.

[0018] In a second aspect, the present application discloses a control method for fast frequency modulation, which is used to realize the battery energy storage system for fast frequency modulation, and the method is executed by the global energy router and includes the following steps:

[0019] Receiving real-time power instructions from a grid scheduling end;

[0020] Obtaining the available capacity state of each distributed power buffer module and the cooling state of each liquid cooling subarea;

[0021] Selecting a subset from all battery clusters as a current serving cluster based on the available capacity state and the cooling state;

[0022] Allocating the real-time power instructions to the distributed power buffer module corresponding to the current serving cluster;

[0023] Synchronously adjusting the cooling intensity of the liquid cooling subarea corresponding to the current serving cluster.

[0024] Further, the step of selecting the current serving cluster based on the available capacity state and the cooling state includes:

[0025] Dividing the battery clusters into at least two logical groups;

[0026] Calculating the comprehensive state score of each logical group, which is positively correlated with the available capacity of the power buffer module corresponding to each battery cluster in the group and is negatively correlated with the inlet temperature of the liquid cooling subarea corresponding to each battery cluster;

[0027] Selecting the logical group with the highest comprehensive state score as the current serving cluster.

[0028] Compared with the related art, the present application has the following beneficial effects:

[0029] The application forms the first active defense line on the electrical path by constructing a distributed power buffer module array, decouples the high-frequency components with violent changes from the power instructions from the grid scheduling end and the direct current flowing to the battery cluster, makes the current change rate borne by the battery tend to be gentle through physical buffering, thereby directly inhibiting the battery internal violent heat generation fluctuation caused by the millisecond-level sharp change of the current from the source; The partitioned counter-flow liquid cooling subsystem constitutes a second heat dissipation defense line that is precisely matched, through topology reconstruction, the distribution logic of the cooling medium spontaneously matches the spatial distribution of the system thermal load, the counter-flow design improves the single cluster heat dissipation uniformity, and the temperature gradient field formed in series realizes the on-demand allocation of cooling capacity, ensuring that the heat generated by the battery can be efficiently and uniformly dissipated, avoiding heat accumulation and local overheating caused by delayed or uneven heat dissipation; The global energy router as the brain of the system forms a third intelligent scheduling defense line, which dynamically selects the optimal working unit by real-time fusion of electrical buffering capacity and heat dissipation state information, synchronously strengthens the cooling of the working unit, and arranges a controlled heat recovery period for other units, breaking the vicious cycle of continuous electrical and thermal stress accumulation and fatigue damage from the system operation strategy level. Through the seamless connection and closed-loop operation of the three defense lines, the external millisecond-level power impact is converted into a gentle thermal load and periodic stress relaxation inside the battery, solving the problem of local thermal runaway of the power storage system in the fast frequency modulation scene, and ensuring the safety and durability of the power storage system in long-term and high-intensity fast frequency modulation service.

[0030] The application introduces a power buffer module composed of a flywheel and a super capacitor, the flywheel responds to power shortage or excess lasting for several seconds with its huge moment of inertia, and the super capacitor responds to transient power peaks with millisecond-level response speed, the two are physically coaxially coupled to form a hybrid power system with complementary response characteristics, so that the battery cluster is completely decoupled from the high-frequency power fluctuations on the grid side, and always works in a gentle current working condition, eliminating the physical conditions that cause the peeling force between active material particles and current collectors due to the sharp change of current, and improving the efficiency of full-band decomposition and processing of grid power instructions.

[0031] The application executes a dynamic grouping rotation scheduling strategy based on real-time electrical and thermal state fusion evaluation through a global energy router, divides the battery cluster array into logical groups and gives them alternating states of service and heat recovery, calculates the comprehensive state score of each group in real time and always prioritizes the group with the highest score to undertake power tasks, while ensuring that the remaining groups obtain complete heat recovery time, this mechanism actively creates a periodic thermal stress relaxation window at the system level, so that the battery can effectively recover the microstructure stability after experiencing electrochemical and thermal load, thereby breaking the vicious chain of continuous accumulation and irreversible deterioration of fatigue damage in traditional operation. BRIEF DESCRIPTION OF DRAWINGS

[0032] Figure 1 A working process schematic diagram of a battery energy storage system for fast frequency modulation is provided for the present application;

[0033] Figure 2 A step flow schematic diagram of a control method for fast frequency modulation is provided for the present application. DETAILED DESCRIPTION

[0034] The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor fall within the scope of protection of the present application.

[0035] The present embodiment provides a battery energy storage system for fast frequency modulation, which comprises a battery cluster array, a high-voltage box, a power converter, a distributed power buffer module array, a partitioned counter-flow liquid cooling subsystem, and a global energy router.

[0036] The battery energy storage system can be integrated as a containerized lithium-ion battery energy storage system in a standard 20-foot container, and the inside of the container is designed to be thermally insulated and sealed to provide a stable operating environment for the internal equipment. The core energy storage unit of the system is the battery cluster array, which can be composed of 12 independent battery clusters arranged longitudinally, and each battery cluster is further integrated by 4 battery packs. The battery cell can use lithium iron phosphate material, and the single capacity is 280 ampere-hours to 314 ampere-hours. The total capacity of the system is 5 megawatt-hours.

[0037] A high-voltage box can be centrally arranged at one end of the container, which integrates DC bus, fuse, contactor and other power distribution and protection devices inside, realizes the collection, distribution and electrical isolation of the electric energy of all battery clusters, and this centralized arrangement is conducive to maintenance. Adjacent to the high-voltage box is the power converter, the DC side of which is connected with the high-voltage box, and the AC side is connected with the power grid. The power converter is responsible for bidirectional energy conversion between DC and AC, and executes the charge and discharge power instructions from the grid dispatching end.

[0038] The system can be configured with a liquid cooling air conditioner, which can provide forced cooling for the battery cluster array through a three-stage liquid cooling pipeline, and the target is to maintain the battery operating temperature between 20 degrees Celsius and 30 degrees Celsius.

[0039] In the present embodiment, the distributed power buffer module array is arranged one-to-one corresponding to the battery clusters in the battery cluster array and is electrically connected between the corresponding battery clusters and the high-voltage box, for absorbing or releasing the high-frequency fluctuation components in the power instructions from the grid dispatching end. The distributed power buffer module can be a flywheel-super capacitor composite module, which contains a flywheel energy storage unit and a super capacitor unit connected coaxially.

[0040] Wherein, the flywheel energy storage unit is a mechanical energy storage device, and its core is a composite material rotor suspended by magnetic bearings and rotating at high speed in a vacuum chamber. The energy (kinetic energy) stored by the flywheel energy storage unit is proportional to the square of the rotational speed, and the expression is , wherein is the energy storage capacity (joule), is the rotational inertia of the rotor (kg·m²), is the angular velocity (rad / s). The flywheel energy storage unit has the characteristics of high power density and extremely long cycle life, and is suitable for dealing with power fluctuations lasting for several seconds to tens of seconds. The super capacitor unit, also known as a double-layer capacitor, can store electric charges through the formation of a double layer at the interface between the electrode and the electrolyte. The super capacitor unit has the characteristics of extremely high power density, millisecond-level charge and discharge response time, and more than one million cycles, and is suitable for dealing with high-frequency and transient power spikes.

[0041] In the present embodiment, the motor or generator rotor of the flywheel energy storage unit and the motor and generator rotor of the super capacitor unit can be rigidly connected through the same mechanical shaft to form a unified electromechanical energy converter. This enables seamless cooperation between the two, achieving direct and efficient coupling of mechanical and electrical forms of energy.

[0042] Specifically, an independent flywheel-super capacitor composite power buffer module can be connected in series between the DC positive and negative bus of each battery cluster and the container high-voltage centralized distribution box. The module can be located between the battery cluster and the high-voltage box in the electrical topology to form a series path of battery cluster-buffer module-high-voltage box. During the smooth period of the power grid without severe power instructions, the battery cluster supplies a small DC current to the composite module, i.e., the normal operation (float) stage. This current drives the motor inside the module, allowing the flywheel rotor to maintain a preset rated speed, for example, 80% of its maximum energy storage, and charging the super capacitor to the rated voltage. At this time, the module acts as a stable load and does not draw fluctuating power from the battery cluster.

[0043] Please refer to Figure 1 When the power converter requires power changes dramatically within milliseconds due to the power instruction from the dispatching end of the power grid, the instruction first acts on the composite power buffer module during the dynamic response (buffer) stage. Specifically, when the instruction requires instantaneous discharge, the module controls the motor to operate as a generator, and the large rotational inertia of the flywheel causes its speed to drop gently, releasing the stored kinetic energy. At the same time, the super capacitor provides a transient peak current through rapid discharge. The two work together to respond to the high-frequency fluctuations and initial large current demand in the instruction. When the instruction requires instantaneous charging, the process is reversed, and the module absorbs the instantaneous high-power pulse, converting it into the kinetic energy of the flywheel and the electric charge of the super capacitor.

[0044] In the present embodiment, the global energy router can evaluate the real-time available capacity of each power buffer module according to the rotational speed state of the flywheel energy storage unit or the voltage state of the supercapacitor unit. Specifically, the flywheel speed can be collected by a high-precision optical encoder installed on the flywheel shaft, which is a core parameter for calculating the real-time energy storage capacity of the supercapacitor is the nominal capacitance value, is the terminal voltage; the module input and output currents can be collected by a Hall current sensor for monitoring their instantaneous power throughput and efficiency, and for evaluation.

[0045] The present embodiment realizes complementary advantages in the frequency band by combining flywheels and supercapacitors. Flywheels are good at dealing with low-frequency, large-inertia continuous power components; supercapacitors are good at processing high-frequency, fast transient power spikes. The combination of the two can cover the full frequency band fluctuations from sub-second to minutes in the grid power instruction, ensuring the smoothness of the battery current in any situation. And the battery cluster works in a smooth DC working condition, and the internal heat production rate changes gently, fundamentally eliminating the conditions for the electrode material to bear alternating thermal stress due to the sharp change of current. This can greatly inhibit the fatigue damage of the battery microstructure, reduce the risk of thermal runaway, and significantly prolong the overall service life of the system.

[0046] In the present embodiment, the zoned counterflow liquid cooling subsystem includes a liquid cooling zone independently set for each battery cluster, and the cooling circuit of each liquid cooling zone adopts a counterflow arrangement mode with the inlet located at the top of the battery cluster and the outlet located at the bottom of the battery cluster. Further, the cooling circuits of the liquid cooling zones can be connected in series by pipelines, so that the liquid cooling zone far from the high-pressure tank has its cooling liquid inlet connected to the cooling liquid outlet of the liquid cooling zone adjacent to the high-pressure tank, thereby forming a cooling liquid temperature gradient field in the arrangement direction of the battery cluster array.

[0047] wherein the liquid cooling zone refers to a cooling unit independently set for each independent battery cluster and relatively independent of the fluid and control system. Each zone has its own independent liquid inlet pipe, liquid outlet pipe and flow regulating valve, thereby realizing precise thermal management for a single battery cluster.

[0048] Since the high-pressure tank is centrally arranged at one end of the container, and the battery clusters are arranged longitudinally along the length direction of the container, the liquid cooling zone adjacent to the high-pressure tank specifically refers to the liquid cooling zone serving the first battery cluster closest to the high-pressure tank in physical space. Correspondingly, the liquid cooling zone far from the high-pressure tank refers to the liquid cooling zone serving the subsequent battery clusters further away from the high-pressure tank along the length direction of the container. By the cooling liquid inlet of the liquid cooling zone far from the high-pressure tank being connected to the cooling liquid outlet of the liquid cooling zone adjacent to the high-pressure tank, it is meant that the liquid outlet pipeline of the former battery cluster is used as the liquid inlet pipeline of the latter battery cluster in the physical installation order of the battery cluster from the high-pressure tank, thereby realizing the serial connection of the pipelines.​

[0049] The counter-flow arrangement refers to the flow direction of the cooling medium being opposite to the optimal direction of heat transfer. In the present system, the cooling liquid inlet is arranged at the top of the battery cluster, and the outlet is arranged at the bottom, which is essentially different from the common bottom-in top-out (co-current) mode. The cooling liquid temperature gradient field refers to the temperature of the cooling liquid inlet flowing through different battery clusters being regularly high and low by connecting the pipelines of each liquid cooling subzone, thereby forming a temperature background field related to the spatial position.

[0050] Specifically, based on the layout of 12 battery clusters arranged longitudinally in the aforementioned standard container, the cooling system is reconfigured as follows in the present embodiment:

[0051] Subzone independence: the single parallel or series large loop serving all battery clusters is cancelled, and an independent liquid cooling micro-subzone is configured for each battery cluster. Each subzone includes a branch pipe connected to the main distribution pipeline, an electrically controlled regulating valve, a pipeline connected to the internal liquid cooling plate of the battery cluster module, and a temperature sensor.

[0052] Topological counter-flow: the cooling liquid branch pipe of each subzone is connected to the top interface of the battery cluster module, and flows out from the bottom interface after flowing through the internal liquid cooling plate. This design utilizes the basic principle that hot air rises and cold air descends, so that the coolest cooling liquid first contacts the upper region of the battery cluster with the highest expected temperature, and the temperature rises after absorbing heat along the way, and then flows through the lower region with a lower temperature, thereby achieving more uniform temperature distribution and higher heat exchange efficiency throughout the height of the battery cluster.

[0053] Pipeline series connection and gradient construction: the inlet and outlet liquid pipelines of all 12 independent subzones are connected in series. Specifically, the lowest temperature cooling liquid output by the liquid cooling air conditioning unit is first introduced to the battery cluster subzone closest to the high-pressure tank and power converter (this region usually has higher additional heat due to electrical losses). The high-temperature outlet liquid of this subzone is introduced to the adjacent next battery cluster subzone as its inlet liquid, and so on. Finally, a cooling liquid temperature gradient field with the inlet liquid temperature gradually increasing from one end to the other end is naturally formed in the length direction of the container.

[0054] To achieve precise control, the following data needs to be collected in the present embodiment: first, the inlet liquid temperature of each subzone, which is the key basic data for calculating the heat exchange temperature difference and evaluating the cooling potential of the subzone. The data can be collected by installing a PT100 temperature sensor after the electrically controlled regulating valve of the inlet liquid pipe of each subzone.

[0055] Second, the real-time flow rate of each subzone, which is the direct execution target for controlling the cooling intensity. The data can be collected by integrating a turbine flowmeter or an ultrasonic flowmeter at the electrically controlled regulating valve of each subzone or after it.

[0056] Third is the key point battery temperature, mainly for calibration and verification of cooling effect, and as a safety redundancy monitoring, can be through the representative module in each group of battery cluster, arranged a small amount of thermocouple or digital temperature sensor to collect the data.

[0057] The embodiment realizes the spatial matching of cooling resources and heat sources by preferentially distributing the low-temperature cooling liquid to the area with the heaviest heat production burden (near the electrical side) through the temperature gradient field formed in series, and the counter-flow arrangement makes the cooling liquid temperature rise process consistent with the natural temperature gradient (high in the upper part and low in the lower part) inside the battery cluster, which avoids the problem that the low-temperature cooling liquid cools the lower part first in the parallel-flow arrangement, resulting in insufficient cooling of the upper part, significantly improves the temperature uniformity of the battery cluster in the vertical direction, and reduces local hot spots.

[0058] In the embodiment, the global energy router is communicatively connected to the distributed power buffer module array and the partitioned counter-flow liquid cooling subsystem, and is configured to select part of the battery clusters as service units to execute the power instruction and synchronously adjust the cooling intensity of the corresponding liquid cooling partition according to the power instruction and the state of each power buffer module.

[0059] Further, the battery cluster array is pre-divided into at least two logical groups; the global energy router is configured to control one logical group as a service unit and control the remaining logical groups in a thermal recovery state when responding to the power instruction, and periodically or according to a state trigger to rotate between the logical groups. The global energy router preferentially selects a logical group with high available capacity of the corresponding power buffer module and low inlet temperature of the corresponding liquid cooling partition as a service unit. In the embodiment, 12 physical battery clusters can be virtually grouped on the software level, for example, they can be fixedly divided into 3 groups, each group containing 4 battery clusters arranged in series. The purpose of grouping is to implement clustered scheduling and management.

[0060] The global energy router is also an independent hardware controller (such as a high-performance industrial PLC or a special industrial computer) as a core intelligent control unit, which runs a special scheduling algorithm. It serves as the hub connecting the grid scheduling instruction-power buffer module-liquid cooling subsystem, and is responsible for decoupling and converting the macro grid demand into optimal collaborative control instructions for the bottom layer physical units. The service unit / service cluster refers to the battery cluster or the group it belongs to that is selected by the global energy router to actually undertake the task of responding to the grid power instruction in a certain period of time, and the battery cluster that is not selected is in a thermal recovery state.

[0061] Specifically, the implementation of the global energy router includes three parts: hardware interface, state evaluation model, and dynamic scheduling strategy.

[0062] The hardware interface part can include an uplink communication interface, a downlink acquisition interface, and a downlink control interface. The uplink communication interface can receive real-time power instructions from a power grid dispatching end through a power special protocol (such as IEC-60870-5-104 or IEC-61850) The downlink acquisition interface can acquire the flywheel rotating speed of all power buffer modules in real time through a high-speed field bus (such as EtherCAT or CAN-FD) The downlink acquisition interface can acquire the flywheel rotating speed of all power buffer modules in real time through a high-speed field bus (such as EtherCAT or CAN-FD) The downlink acquisition interface can acquire the flywheel rotating speed of all power buffer modules in real time through a high-speed field bus (such as EtherCAT or CAN-FD) The downlink acquisition interface can acquire the flywheel rotating speed of all power buffer modules in real time through a high-speed field bus (such as EtherCAT or CAN-FD) The downlink acquisition interface can acquire the flywheel rotating speed of all power buffer modules in real time through a high-speed field bus (such as EtherCAT or CAN-FD) The downlink acquisition interface can acquire the flywheel rotating speed of all power buffer modules in real time through a high-speed field bus (such as EtherCAT or CAN-FD) .

[0063] The global energy router preferentially selects a logical group with high available capacity of the corresponding power buffer module and low inlet temperature of the corresponding liquid cooling partition as a service unit. The core of the determination logic is to execute a dynamic optimization program. The program fuses the real-time available capacity state and the cooling liquid inlet temperature state of each logical group into a comparable priority index through an internal calculation process. The router constantly selects the logical group with the optimal priority index as the service unit by calculating and comparing the priority index of all logical groups in real time. The higher the average available capacity of the battery cluster in the group and the lower the average inlet temperature, the better the priority index. The specific quantitative calculation method (corresponding to the comprehensive state score) of the index is described in detail in the subsequent embodiment related to the control method.

[0064] The state evaluation model part can continuously calculate the comprehensive state score of each battery cluster through the global energy router First, the available capacity of the single battery needs to be calculated That is, the i-th battery cluster can be calculated according to the buffer module data, and the calculation formula is:

[0065] ;

[0066] Wherein, refers to the design rated maximum safe operating angular velocity (unit: rad / s) of the flywheel energy storage unit; refers to the design rated maximum operating voltage (unit: V) of the super capacitor unit; is a weight coefficient representing the contribution degree of the flywheel and the super capacitor. The greater the value, the stronger the buffering capacity of the cluster to withstand power impact (the greater the electrical potential), which corresponds to the aforementioned global energy router that can evaluate the real-time available capacity of each power buffering module according to the speed state of the flywheel energy storage unit or the voltage state of the supercapacitor unit.

[0067] Then the single cooling potential is calculated , and the calculation formula is:

[0068] ;

[0069] Among them, is the upper limit temperature of the cooling system design, which can be set to 35°C, is the minimum output temperature of the air conditioner, which can be set to 20°C. The obtained is lower, the value is higher, indicating that the heat dissipation condition is better.

[0070] Finally, the grouping comprehensive score is calculated : For all battery clusters belonging to the same logical grouping g, the average value of (α +β ) is calculated as the score of the grouping. Among them, α and β are adjustable coordination coefficients, which can be initially set to 0.5 to balance the importance of electricity and heat.

[0071] The program compares the values of all logical groupings, and the selection rule is: always mark the grouping with the highest value as the “candidate service grouping”. The rotation logic is: there is a “current service grouping” variable and a “minimum continuous service time ” (such as 5 minutes) inside the program. If the “candidate service grouping” is different from the “current service grouping”, and the continuous service time of the “current service grouping” has exceeded , rotation is triggered. When rotating, the “current service grouping” state is set to “hot recovery”, and the “candidate service grouping” state is set to “service”. If the minimum time is not reached, the original grouping is maintained, even if its score is not the highest, to avoid too frequent switching.

[0072] In this embodiment, the global energy router is configured to: when receiving the power instruction issued by the power grid scheduling end, decompose the corresponding frequency segment (specific frequency modulation instruction segment) in the power instruction into high-frequency components and low-frequency components, and preferentially allocate the high-frequency components to the distributed power buffering module array for buffering. When the buffering module resolves the sharp change of current, the battery cluster array bears the smooth current, and the liquid cooling system can more calmly and efficiently cope with the battery heat generation.

[0073] The specific frequency modulation instruction segment refers to the power instruction interval with specific characteristics in the time domain or frequency domain issued by the grid dispatching end. Its identification and determination are completed by real-time analysis of the global energy router, mainly based on one or more of the following conditions:

[0074] Time domain change rate condition: when the change amplitude |ΔP| of the real-time power instruction P_cmd(t) in a very short time (e.g. 100 milliseconds) exceeds a preset proportion (e.g. 20%) of the system rated power, it is determined that the time period and the subsequent short duration are specific frequency modulation instruction segments.

[0075] Frequency domain component condition: the router performs real-time spectral analysis on P_cmd(t), and when it detects that the total energy proportion of the components higher than a certain characteristic frequency (e.g. 0.5 Hz) in the power spectrum exceeds a preset threshold, it determines that the current instruction segment is a specific frequency modulation instruction segment. This characteristic frequency usually corresponds to the critical frequency of the thermal response lag of the liquid cooling system.

[0076] Instruction source marking condition: when the grid dispatching end issues certain auxiliary service instructions (such as fast regulation signals for primary frequency modulation) that require extremely fast response, it may attach a specific service type identifier in the communication protocol. When the router identifies this identifier, it determines that the subsequent associated power instruction segment is a specific frequency modulation instruction segment.

[0077] When any of the above conditions is met, the router starts the decomposition and distribution process.

[0078] Specifically, for each control cycle received , the router performs dynamic allocation: first, instruction decomposition, using a first-order high-speed digital filter to decompose into a low-frequency basic component and a high-frequency fluctuation component , with a cutoff frequency usually set to 0.1-0.5 Hz. Then, power distribution, which allocates to the BMS of each battery cluster in the current "service group" in proportion as its steady-state power reference. At the same time, allocates mainly to the power buffer module corresponding to each battery cluster in the group (sending ), taking advantage of its fast response characteristics for throughput.

[0079] After determining the service group, immediately adjust the cooling: send =85% instruction to the liquid cooling sub-area electric control valve corresponding to all battery clusters in the service group, which corresponds to the "intensive cooling" mode; send =40% instruction to the group in "thermal recovery" state, which corresponds to the "maintain cooling" mode.

[0080] It is worth mentioning that the global energy router can also dynamically adjust the thermal recovery duration of the logical group to which the battery cluster belongs according to the historical accumulated discharge depth and the highest temperature experienced by each battery cluster when triggering the logical group rotation.

[0081] Specifically, in the program, the health data of each battery cluster in all logical groups are compared at the time, for each logical group g, the program calculates the dynamic thermal recovery duration of the battery cluster according to the health data of the battery cluster contained therein.

[0082] ;

[0083] wherein, represents the basic thermal recovery duration; and respectively represent the average historical discharge depth factor and the average historical over-temperature degree of the battery cluster in the group; γ and δ are influence coefficients for adjusting the influence weight of the historical factors on the recovery duration.

[0084] The formula makes the group with more serious historical fatigue (more cumulative discharge and higher temperature experienced) obtain a longer standing recovery time after exiting the service state, so as to fully release the internal accumulated stress.

[0085] For dynamically adjusting the thermal recovery duration of the logical group to which the battery cluster belongs, the total charge / discharge ampere-hour of the battery cluster recorded by the BMS can be used to calculate the cumulative charge amount and normalize to obtain the average historical discharge depth factor of the battery cluster in the group; by analyzing the historical temperature alarm and sampling logs of the BMS, the maximum temperature value reached by each cluster can be extracted. These data are pushed to the global energy router through the regular reporting function of the BMS, and the influence coefficients γ and δ in the formula are fine-tuned according to the latest battery cluster performance degradation data, so that the predictive management model is more in line with the actual aging state.

[0086] Referring to Figure 2 , the embodiment also provides a control method for fast frequency modulation, which is applied to the battery energy storage system for fast frequency modulation disclosed in the embodiment, and the method can be executed by the global energy router. The method can include the following steps:

[0087] Step 1: receiving real-time power instructions from the grid dispatching end.

[0088] In the embodiment, the global energy router can continuously monitor and receive real-time power instructions issued by the grid dispatching system through the communication interface on the router . The instructions are usually issued in the form of set points with time tags, and the energy storage is required to track the changes within milliseconds to seconds. After the router analyzes and verifies the effectiveness of the instructions, the instructions are used as input trigger signals for the entire method flow.

[0089] Step two, obtain the available capacity state of each distributed power buffer module and the cooling state of each liquid cooling partition.

[0090] Among them, the available capacity state and the cooling state are two core input dimensions for the global energy router to make decisions. The available capacity state quantifies the potential of the power buffer module to absorb power fluctuations, and the cooling state quantifies the real-time cooling condition of the corresponding battery cluster. The comprehensive evaluation of the two aims to achieve electric-thermal collaborative optimization.

[0091] In this embodiment, before responding to the instruction, the global energy router can synchronize the collection of two key real-time state data sets through its downlink data bus: electrical buffer state: read the current speed of the flywheel from the controller of each distributed power buffer module and the current voltage of the super capacitor These raw data directly reflect the instantaneous energy storage and release capacity of each module; thermal management state: read the cooling liquid inlet temperature of each independent liquid cooling partition from the sensor This temperature is a key parameter for evaluating the immediate cooling potential of the partition.

[0092] Step three, based on the available capacity state and the cooling state, select a subset of all battery clusters as the current service cluster.

[0093] Further, the step of selecting the current service cluster based on the available capacity state and the cooling state can include:

[0094] Divide the battery clusters into at least two logical groups; calculate the comprehensive state score of each logical group, the comprehensive state score is positively correlated with the available capacity of the power buffer module corresponding to each battery cluster in the group, and is negatively correlated with the liquid cooling partition inlet temperature corresponding to each battery cluster; select the logical group with the highest comprehensive state score as the current service cluster.

[0095] In this embodiment, the global energy router can select all battery clusters contained in the logical group with the highest comprehensive state score as the current service cluster by comparing the values of all logical groups The group will be out of the thermal recovery state and turned to undertake power tasks.

[0096] Step four, distribute the real-time power instruction to the distributed power buffer module corresponding to the current service cluster.

[0097] In this embodiment, the router distributes the received in step one to the power buffer module corresponding to the current service cluster. Specifically, the router can generate a power setting value proportional to And the local controller of each module is sent control commands through the control bus to drive the flywheel and super capacitor to quickly absorb or release power, so as to suppress the high-frequency fluctuation component in the command and protect the battery cluster at the back end.

[0098] Step five, synchronously adjust the cooling intensity of the liquid cooling subzone corresponding to the current service cluster.

[0099] In this embodiment, when the power distribution command is distributed, the global energy router sends control commands to the flow regulating valve of each liquid cooling subzone bound to the current service cluster, and sets the opening degree set value of the flow regulating valve to a higher "active cooling" gear (for example, 80% of the rated opening degree), so as to immediately enhance the heat dissipation capacity of the part of the battery cluster. And the opening degree set value of the flow regulating valve of the liquid cooling subzone where the non-service cluster is located is set to a lower "maintenance cooling" gear (for example, 30% of the rated opening degree).

[0100] In summary, the first active defense line is formed on the electrical path by constructing a distributed power buffer module array, the high-frequency component with large changes in the power command from the power grid dispatching end is decoupled from the direct current flowing to the battery cluster, the current change rate borne by the battery is made to tend to be gentle through physical buffering, so that the battery internal severe heat fluctuation directly caused by the millisecond-level dramatic change of the current is inhibited from the source; the partitioned counter-flow liquid cooling subsystem constitutes a second heat dissipation defense line that is precisely matched, the distribution logic of the cooling medium is spontaneously matched with the spatial distribution of the system heat load through topology reconstruction, the counter-flow design improves the single cluster heat dissipation uniformity, and the temperature gradient field formed in series realizes the on-demand allocation of cooling capacity, so as to ensure that the heat generated by the battery can be efficiently and uniformly dissipated, and heat accumulation and local overheating caused by heat dissipation delay or unevenness are avoided; the global energy router as the brain of the system forms a third intelligent scheduling defense line, which dynamically selects the optimal working unit and synchronously strengthens the cooling of the working unit by real-time fusion of the electrical buffer capacity and the heat dissipation state information, and arranges a controlled heat recovery period for other units, so as to break the vicious cycle of continuous electrical and thermal stress accumulation and fatigue damage from the system operation strategy level. Through the seamless connection and closed-loop operation of the three defense lines, the external millisecond-level power impact is converted into a gentle heat load and a periodic stress relaxation in the battery, so as to solve the problem of local thermal runaway of the battery energy storage system in the fast frequency modulation scene, and ensure the safety and durability of the battery energy storage system in long-term and high-intensity fast frequency modulation service.

[0101] By introducing a power buffer module composed of a flywheel and a super capacitor, the flywheel responds to the power shortage or excess lasting for several seconds with its huge moment of inertia, and the super capacitor responds to the transient power peak with millisecond-level response speed, both of which are physically coaxially coupled to form a hybrid power system with complementary response characteristics, so that the battery cluster is completely decoupled from the high-frequency power fluctuations of the grid side and always works in a slowly changing current condition, eliminating the physical conditions that cause the active material particles and current collectors to produce peeling forces due to the sharp change of current, and improving the efficiency of full-band decomposition and processing of the grid power command.

[0102] By executing a dynamic grouping rotation scheduling strategy based on real-time electro-thermal state fusion evaluation through a global energy router, the battery cluster array is divided into logical groups and is given an alternating state of service and thermal recovery, the comprehensive state score of each group is calculated in real time and the group with the highest score is always prioritized to undertake power tasks, while ensuring that the remaining groups obtain complete thermal recovery time, this mechanism actively creates a periodic thermal stress relaxation window at the system level, so that the battery can effectively recover the microstructure stability after experiencing electrochemical and thermal load, thereby breaking the chain of continuous accumulation and irreversible deterioration of fatigue damage in traditional operation.

[0103] The above describes only the specific embodiments of the present application, but the protection scope of the present application is not limited thereto, any person skilled in the art can easily think of changes or replacements within the technical range disclosed by the present application, which should be covered within the protection scope of the present application. Therefore, the protection scope of the present application should be subject to the protection scope of the claims.

[0104] Finally: the above described only the preferred embodiments of the present application, and not for limiting the present application, any modification, equivalent replacement, improvement, etc. made within the spirit and principles of the present application, should be included within the protection scope of the present application.

Claims

1. A battery energy storage system for fast frequency modulation, comprising an array of battery clusters, a high voltage tank, and a power converter, characterized in that, The system comprises: a distributed power buffer module array, which is arranged one-to-one corresponding to the battery clusters in the battery cluster array and is electrically connected between the corresponding battery cluster and the high-voltage box, for absorbing or releasing the high-frequency fluctuation component in the power instruction from the grid scheduling end; a partitioned counter-flow liquid cooling subsystem, which comprises a liquid cooling partition independently arranged for each battery cluster, and the cooling loop of each liquid cooling partition adopts a counter-flow arrangement mode with the inlet located at the top of the battery cluster and the outlet located at the bottom of the battery cluster; a global energy router, which is communicatively connected to the distributed power buffer module array and the partitioned counter-flow liquid cooling subsystem, for selecting part of the battery clusters as service units to execute the power instruction and synchronously adjusting the cooling intensity of the corresponding liquid cooling partition according to the power instruction and the state of each power buffer module.

2. A battery energy storage system for fast frequency modulation according to claim 1, characterized in that, The distributed power buffer module is a flywheel-super capacitor composite module, which comprises a flywheel energy storage unit and a super capacitor unit connected coaxially.

3. A battery energy storage system for fast frequency modulation according to claim 2, characterized in that, The global energy router evaluates the real-time available capacity of each power buffer module according to the speed state of the flywheel energy storage unit or the voltage state of the super capacitor unit.

4. A battery energy storage system for fast frequency modulation according to claim 1, characterized in that, The cooling loop of each liquid cooling partition is connected in series through pipelines, so that the liquid cooling partition far from the high-voltage box is connected to the cooling liquid outlet of the liquid cooling partition adjacent to the high-voltage box, thereby forming a cooling liquid temperature gradient field in the arrangement direction of the battery cluster array.

5. A battery energy storage system for fast frequency modulation according to claim 1, wherein, The battery cluster array is pre-divided into at least two logical groups; the global energy router is configured to: when responding to the power instruction, control one logical group as a service unit, control the remaining logical groups in a hot recovery state, and periodically or according to a state trigger the rotation between logical groups.

6. A battery energy storage system for fast frequency modulation according to claim 5, wherein, The global energy router preferentially selects a logical group with high available capacity of the corresponding power buffer module and low inlet temperature of the corresponding liquid cooling partition as a service unit.

7. A battery energy storage system for fast frequency modulation according to claim 5, wherein, When triggering the rotation of logical groups, the global energy router also dynamically adjusts the hot recovery duration of the logical group to which each battery cluster belongs according to the historical cumulative discharge depth or the highest temperature experienced by each battery cluster.

8. A battery energy storage system for fast frequency modulation according to claim 1, wherein, The global energy router is configured to: when receiving the power instruction issued by the grid scheduling end, decompose the corresponding frequency segment in the power instruction into a high-frequency component and a low-frequency component, and preferentially allocate the high-frequency component to the buffer by the distributed power buffer module array.

9. A control method for fast frequency modulation, applied to the battery energy storage system for fast frequency modulation according to any one of claims 1-8, characterized in that, The method is performed by the global energy router and comprises the following steps: receiving real-time power instructions from the grid scheduling end; obtaining the available capacity state of each distributed power buffer module and the cooling state of each liquid cooling partition; selecting a subset from all battery clusters as a current service cluster based on the available capacity state and the cooling state; allocating real-time power instructions to the distributed power buffer module corresponding to the current service cluster; synchronously adjusting the cooling intensity of the liquid cooling partition corresponding to the current service cluster.

10. The control method for fast frequency modulation according to claim 9, wherein The step of selecting the current service cluster based on the available capacity state and the cooling state comprises: dividing the battery clusters into at least two logical groups; calculating the comprehensive state score of each logical group, which is positively correlated with the available capacity of the power buffer module corresponding to each battery cluster in the group and is negatively correlated with the inlet temperature of the liquid cooling partition corresponding to each battery cluster; selecting the logical group with the highest comprehensive state score as the current service cluster.

Images