Intelligent lithium ion modular energy storage cabinet combination optimization control method and system

Abstract

The invention relates to the technical field of lithium ion energy storage control, and discloses an intelligent lithium ion modular energy storage cabinet combination optimization control method and system. The method comprises the following steps: acquiring voltage, current and temperature data of each energy storage module in real time through distributed data acquisition nodes, and generating a dynamic operation parameter set; based on this, the charging and discharging state difference of each energy storage module is analyzed, the power balance demand between the modules is calculated, and a power distribution regulation and control instruction set is generated. And adjusting the charging and discharging power thresholds of the modules according to the instruction set, synchronously updating the energy flow path configuration between the modules, monitoring the temperature rise rate and the voltage fluctuation range after power adjustment, generating a thermal state correction parameter, and carrying out secondary calibration on the power thresholds according to the thermal state correction parameter. And performing cooperative charging and discharging operation based on the optimized instruction set, collecting real-time operation data, and generating a dynamic equilibrium compensation parameter. And reconstructing an energy flow path according to the compensation parameter, updating the power allocation weight of each module, and outputting a final weight to a control terminal.

Description

Technical Field

[0001] This invention relates to the field of lithium-ion energy storage control technology, specifically to a method and system for optimizing the combination of intelligent lithium-ion modular energy storage cabinets. Background Technology

[0002] With the rapid development of the new energy industry, lithium-ion energy storage cabinets, due to their high energy density and long cycle life, are widely used in power system peak shaving, new energy consumption, and user-side energy storage scenarios. In practical applications, to meet the power and capacity requirements of different scenarios, energy storage systems are usually constructed by combining multiple energy storage modules. However, due to differences in material properties and process precision during the manufacturing process of each energy storage module, and the influence of factors such as the number of charge-discharge cycles, ambient temperature fluctuations, and load changes during long-term operation, inconsistencies occur in the operating parameters such as voltage, current, and temperature of each module.

[0003] This inconsistency in parameters can lead to a series of problems: on the one hand, some energy storage modules may have shortened service life and increased safety hazards due to prolonged overcharging, over-discharging, or high-temperature conditions; on the other hand, differences in charge and discharge states between modules can cause a decrease in the overall energy utilization efficiency of the energy storage system, failing to fully utilize the performance of each module. Currently, control methods for modular energy storage cabinets mostly focus on the monitoring and regulation of single parameters, or adopt simple power sharing strategies, lacking comprehensive collection and analysis of the dynamic operating parameters of each module, making it difficult to formulate precise power allocation schemes based on the actual state differences between modules.

[0004] Existing control methods often neglect real-time monitoring of module temperature rise rate and voltage fluctuations after adjusting power distribution, failing to promptly correct thermal imbalances caused by power adjustments. Furthermore, during coordinated charging and discharging, they lack dynamic tracking of voltage consistency deviations and temperature gradient distributions, making it difficult to achieve dynamic balance between modules by reconstructing energy flow paths. These issues make it difficult for existing modular energy storage cabinets to balance system stability, safety, and energy utilization efficiency during long-term operation. They cannot meet the high-precision, high-reliability control requirements of energy storage systems in complex application scenarios, thus hindering the further promotion and application of lithium-ion modular energy storage cabinets in the new energy field. Summary of the Invention

[0005] The purpose of this invention is to provide a method and system for optimizing the combination of intelligent lithium-ion modular energy storage cabinets, so as to solve the problems mentioned in the background art.

[0006] To achieve the above objectives, the present invention provides a method for optimizing the combination control of intelligent lithium-ion modular energy storage cabinets, the method comprising:

[0007] The voltage, current and temperature data of each energy storage module are acquired in real time through distributed data acquisition nodes to generate a dynamic set of operating parameters.

[0008] Based on the dynamic operating parameter set analysis, the differences in the charging and discharging states of each energy storage module are analyzed, the power balance requirements between modules are calculated, and a power allocation control instruction set is generated.

[0009] Adjust the charging and discharging power thresholds of each energy storage module according to the power distribution control instruction set, and synchronously update the energy flow path configuration between modules;

[0010] Monitor the temperature rise rate and voltage fluctuation range of the energy storage module after power adjustment, and generate thermodynamic state correction parameters;

[0011] The charge and discharge power thresholds are calibrated a second time by combining the thermodynamic state correction parameters to generate an optimized power distribution control instruction set.

[0012] Based on the optimized power distribution control instruction set, the energy storage module performs coordinated charging and discharging operations and collects real-time coordinated operation data.

[0013] Analyze the voltage consistency deviation and temperature gradient distribution in real-time collaborative operation data to generate dynamic equalization compensation parameters;

[0014] The energy flow path configuration is reconstructed based on the dynamic equilibrium compensation parameters, and the power allocation weight of the energy storage module is updated.

[0015] The final power allocation weight is output to the energy storage module control terminal to complete the distributed collaborative control closed loop.

[0016] Preferably, the step of generating the dynamic operating parameter set specifically includes:

[0017] A fixed sampling period is set, and the instantaneous voltage, instantaneous current, and instantaneous temperature values ​​of each energy storage module are synchronously collected through distributed data acquisition nodes;

[0018] The instantaneous voltage, current, and temperature values ​​collected within the same sampling period are timestamped to align the data and remove abnormal jump data.

[0019] Calculate the rate of change of voltage, current and temperature for each energy storage module during a continuous sampling period, and generate a dynamic trend matrix.

[0020] Integrate the dynamic trend matrix of all energy storage modules to construct a dynamic operating parameter set.

[0021] Preferably, the step of generating the power allocation control instruction set specifically includes:

[0022] Extract the voltage and current change rates of each energy storage module from the dynamic operating parameter set, and calculate the voltage and current deviation coefficients between modules;

[0023] Power allocation priority is determined based on voltage deviation coefficient and current deviation coefficient, and energy storage modules with low voltage deviation coefficient and high current deviation coefficient are marked as priority control targets;

[0024] The power carrying capacity of the priority control object is evaluated by combining the instantaneous temperature value, and an initial power allocation ratio is generated;

[0025] A power allocation control instruction set is generated based on the initial power allocation ratio.

[0026] Preferably, the update step of the energy flow path configuration specifically includes:

[0027] Obtain the target charge and discharge power thresholds for each energy storage module in the power allocation and control instruction set;

[0028] Analyze the difference between the target charging / discharging power threshold and the current actual power to determine the energy flow direction and transmission magnitude;

[0029] Based on the energy flow direction and transmission magnitude, the physical connection paths between modules are replanned, and an energy flow path configuration table is generated.

[0030] The energy flow path configuration table is sent to the power routing executor.

[0031] Preferably, the step of generating the thermodynamic state correction parameters specifically includes:

[0032] Monitor the instantaneous temperature values ​​of each energy storage module after power adjustment, and calculate the absolute value of temperature rise per unit time;

[0033] By comparing the absolute value of the temperature rise with the preset safety threshold, energy storage modules that exceed the temperature rise limit can be identified.

[0034] Extract the voltage fluctuation range of the over-limit temperature rise energy storage module and calculate the thermo-coupling influence factor;

[0035] Thermodynamic state correction parameters are generated based on the thermodynamic coupling influence factor.

[0036] Preferably, the secondary calibration step specifically includes:

[0037] The thermal state correction parameters are weighted and fused with the initial power allocation ratio to generate the corrected power allocation ratio;

[0038] The charging and discharging power thresholds of each energy storage module are recalculated based on the revised power allocation ratio.

[0039] Update the power distribution control instruction set.

[0040] Preferably, the step of generating the dynamic equilibrium compensation parameters specifically includes:

[0041] Collect real-time voltage and temperature values ​​of each energy storage module during the coordinated charging and discharging operation;

[0042] Calculate the standard deviation of real-time voltage values ​​and the gradient difference between real-time temperature values ​​to generate a consistency evaluation index.

[0043] The compensation intensity and direction are determined based on the consistency assessment indicators, and dynamic equilibrium compensation parameters are generated.

[0044] Preferably, the update step of the power allocation weight specifically includes:

[0045] The dynamic balance compensation parameters are superimposed with the corrected power allocation ratio, and the power output ratio of each energy storage module is redistributed based on the superposition result to generate the final power allocation weight.

[0046] Preferably, the specific steps for completing the distributed collaborative control closed loop are as follows:

[0047] The final power allocation weights are encoded into control commands;

[0048] Control commands are broadcast to the control terminals of each energy storage module via a distributed communication network;

[0049] Verify the matching degree between the actual power output of each energy storage module and the command, and complete the closed-loop confirmation.

[0050] Preferably, the present invention also includes an intelligent lithium-ion modular energy storage cabinet combination optimization control system, including a memory, a processor, and a computer program stored in the memory and running on the processor. When the processor executes the computer program, it implements the steps of the above-described intelligent lithium-ion modular energy storage cabinet combination optimization control method.

[0051] Compared with the prior art, the beneficial effects of the present invention are:

[0052] The intelligent lithium-ion modular energy storage cabinet combination optimization control method acquires voltage, current, and temperature data of each energy storage module in real time through distributed data acquisition nodes, generating a dynamic operating parameter set. This allows for a comprehensive understanding of the real-time operating status of each module, avoiding control deviations caused by incomplete monitoring of a single parameter, and providing a more realistic basis for subsequent power allocation and path configuration. Based on the dynamic operating parameter set, the method analyzes the differences in charging and discharging states of each energy storage module, calculates the power balancing requirements between modules, and generates a power allocation control command set. Differentiated power allocation schemes can be formulated according to the actual state differences of each module, avoiding overloading or underutilization of some modules under a simple power equalization strategy, and ensuring that the performance of each module is rationally utilized.

[0053] By adjusting the charging and discharging power thresholds of each energy storage module according to the power distribution control command set and synchronously updating the energy flow path configuration between modules, coordinated adjustment of power distribution and energy transmission paths can be achieved. This ensures that each module operates according to the predetermined power target and reduces energy loss during transmission. Monitoring the temperature rise rate and voltage fluctuation range of the energy storage modules after power adjustment generates thermodynamic state correction parameters, which can promptly capture the impact of power adjustment on the module's thermal state, avoiding problems such as excessive temperature rise or voltage fluctuation caused by power adjustment, and ensuring the safety and stability of module operation.

[0054] By combining thermal state correction parameters to perform secondary calibration of the charge and discharge power thresholds, an optimized power allocation control command set is generated. This allows for dynamic correction of the initial power allocation scheme, making the power allocation more consistent with the real-time thermal state of the modules and further improving the accuracy of power allocation. Based on the optimized power allocation control command set, the energy storage modules are executed in a coordinated charge and discharge manner, and real-time coordinated operation data is collected. This allows for continuous tracking of the system status during coordinated operation, providing data support for subsequent dynamic balancing adjustments.

[0055] Analyzing voltage consistency deviations and temperature gradient distributions in real-time collaborative operation data generates dynamic equilibrium compensation parameters. This accurately identifies parameter inconsistencies between modules during collaborative operation, providing direction for subsequent path reconstruction and weight updates. Reconstructing energy flow path configurations based on these dynamic equilibrium compensation parameters and updating the power allocation weights of energy storage modules enables dynamic optimization of system operation, mitigating parameter inconsistencies between modules and extending the overall lifespan of the modules.

[0056] The final power allocation weight is output to the energy storage module control terminal to complete the distributed collaborative control closed loop. This enables the formation of a complete control process from data acquisition, analysis, regulation to feedback, ensuring that the regulation measures in each link can be effectively implemented, improving the stability and reliability of the entire energy storage system, and improving the energy utilization efficiency of the system to better adapt to the energy storage application needs in different scenarios. Attached Figure Description

[0057] Figure 1 This is a schematic diagram illustrating the working principle of the intelligent lithium-ion modular energy storage cabinet combination optimization control method described in this invention.

[0058] Figure 2 A flowchart for the steps of generating a dynamic runtime parameter set;

[0059] Figure 3 A flowchart of the steps for generating a power allocation and control instruction set. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0061] Please see Figure 1 This invention provides a method and system for optimized control of intelligent lithium-ion modular energy storage cabinets. The method includes: distributed data acquisition nodes acquiring real-time voltage, current, and temperature data of each energy storage module; generating a dynamic operating parameter set; analyzing the differences in charge / discharge states of each energy storage module based on this parameter set; calculating the power balance requirements between modules; generating a power allocation control instruction set; adjusting the charge / discharge power thresholds of each energy storage module according to the instruction set; synchronously updating the energy flow path configuration between modules; monitoring the temperature rise rate and voltage fluctuation range of the energy storage modules after power adjustment; generating thermal state correction parameters; performing a secondary calibration of the charge / discharge power thresholds based on the correction parameters; generating an optimized power allocation control instruction set; executing coordinated charge / discharge operations of the energy storage modules based on the optimized instruction set; collecting real-time coordinated operation data; analyzing voltage consistency deviations and temperature gradient distributions in the data; generating dynamic balance compensation parameters; reconstructing the energy flow path configuration according to the compensation parameters; updating the power allocation weights of the energy storage modules; and outputting the final power allocation weights to the energy storage module control terminal, completing the distributed coordinated control closed loop. This method achieves dynamic power balance and thermal management between energy storage modules, improving the overall system efficiency and safety.

[0062] Example 1: See Figure 2 The dynamic parameter set generation process is based on a preset fixed sampling period. Distributed data acquisition nodes are deployed near the electrical interfaces of each energy storage module. These nodes typically consist of high-precision voltage sensors, Hall current sensors, and surface-mount temperature sensors. Their sampling trigger signals are globally synchronized by a master clock source, ensuring that modules in different physical locations can capture instantaneous voltage, current, and temperature values ​​within the same microsecond-level time window. The acquired raw data stream first enters a preprocessing buffer. The data packet header embeds a timestamp accurate to the millisecond level. The core task of the preprocessing stage is to perform timestamp alignment on these multiple data streams. This is because although the sampling actions are triggered synchronously, there may be slight differences in the network latency of the data reaching the processor. The alignment algorithm uses a reference time axis as a benchmark to rearrange and combine the voltage, current, and temperature data of each module according to the timestamp, forming strictly synchronized data frames.

[0063] After completing timestamp alignment, the system initiates an anomaly data removal process. Identification of anomalous jumps relies on a dynamic threshold model based on historical statistics. This model continuously learns the data fluctuation range of each sensor under normal operating conditions. For example, the change in instantaneous voltage values ​​over multiple consecutive cycles typically exhibits a certain inertial characteristic. If the rate of change of the voltage value collected in a certain cycle relative to the previous cycle suddenly exceeds the threshold allowed by physical laws (such as the battery electrochemical response limit), then this data point is marked as an anomalous jump and removed. The removal operation is not simply discarding data; instead, it combines normal values ​​from previous and subsequent cycles and uses linear interpolation or spline interpolation algorithms for reasonable repair to ensure the continuity of the data sequence is not disrupted, providing a complete and reliable data foundation for subsequent trend analysis. Subsequently, the system calculates the dynamic rate of change for each energy storage module within continuous sampling cycles. The voltage rate of change is calculated by comparing the difference between the instantaneous voltage value of the current cycle and the previous cycle, and then dividing by the time interval of the sampling cycle. This difference method effectively captures the instantaneous trend of voltage change. The calculation of the current rate of change follows the same principle, but special attention must be paid to the sign of the current direction. The calculation of the temperature change rate is relatively complex, requiring consideration of the inertial characteristics of heat transfer. Therefore, the algorithm introduces a sliding time window to perform linear regression analysis on temperature values ​​over multiple consecutive periods, using the slope as an estimate of the temperature change rate to smooth out short-term fluctuations caused by measurement noise. All these change rate data are ultimately organized into a dynamic trend matrix. Each row of the matrix uniquely corresponds to an identifier for an energy storage module, while the column vectors store the voltage change rate sequence, current change rate sequence, and temperature change rate sequence of that module in chronological order.

[0064] Next, the dynamic trend matrices of all energy storage modules are integrated to construct a global dynamic operating parameter set. This integration process is not a simple matrix stacking but involves multi-dimensional data fusion. The system generates a global state snapshot at each sampling moment, containing the voltage, current, and temperature change rates of all modules at that instant, thus forming a three-dimensional data cube. This data cube is the final dynamic operating parameter set, comprehensively characterizing the dynamic operating characteristics of the entire energy storage cluster from both spatial (different modules) and temporal (continuous cycles) dimensions. The data cube is stored in a high-performance in-memory database with an efficient indexing mechanism, enabling subsequent analysis programs to quickly query and access the dynamic trend of any module within any time period.

[0065] The entire data acquisition and processing chain employs a highly reliable design. Distributed data acquisition nodes typically possess local caching capabilities, allowing them to continue recording data even during temporary network interruptions. Once communication is restored, data is uploaded in batches and re-synchronized for time synchronization. The length of the fixed sampling period can be configured according to the actual application scenario. For frequency modulation applications requiring rapid response, the period can be set to milliseconds; for steady-state applications such as energy transfer, the period can be set to seconds or even longer to balance system load and control accuracy. Timestamp alignment processing relies on high-precision clock synchronization protocols, such as the IEEE 1588 Precision Time Protocol (PTP), ensuring that clock deviations between distributed nodes are controlled within microseconds. The algorithm for eliminating anomalous jump data has self-learning capabilities, and its dynamic threshold can adaptively adjust according to the aging of the battery, avoiding misjudging normal aging voltage drops as anomalous data. The calculated dynamic trend matrix not only includes the instantaneous rate of change but also includes a confidence index for each rate of change. This index is calculated based on the recent data stability of the sensor, and data points with lower confidence are assigned smaller weights in subsequent analysis. The completed set of dynamic operating parameters is finally encapsulated in a standardized data format (such as JSON or Protocol Buffers) and broadcast to all subsequent processing units in the system that need the data, such as the power equalization analysis module and the thermal management module, through a high-speed internal network, thus initiating the first substantive step in the entire optimization control process.

[0066] Example 2: See Figure 3The process involves two closely linked stages: the generation of the power allocation and control command set and the updating of the energy flow path configuration. The generation of the power allocation and control command set begins with a deep analysis of the dynamic operating parameter set. The calculation module extracts the voltage and current rate of change data for each energy storage module from the parameter set. The voltage deviation coefficient is calculated by statistically analyzing the voltage rate of change of all modules at the same sampling moment and calculating the Euclidean distance between the rate of change of each module and the system average rate of change. This coefficient quantifies the degree of deviation of the module's voltage dynamic response from the group average level. The current deviation coefficient is calculated using a similar principle, but focuses on evaluating the dynamic consistency of the module when it receives or outputs current. If the current rate of change of a module is consistently higher or lower than the system average level, the absolute value of its current deviation coefficient will increase. Based on the calculated voltage and current deviation coefficients, the system executes a power allocation priority decision-making logic. The decision rule uses low voltage deviation coefficients and high current deviation coefficients as core judgment criteria. A low voltage deviation coefficient indicates that the module's terminal voltage is more stable during dynamic processes, suggesting a healthier electrochemical state or a superior state of charge. Conversely, a high current deviation coefficient suggests that the module's current is fluctuating drastically, possibly due to sudden load changes or internal impedance variations, requiring external intervention to smooth out the fluctuations. Therefore, energy storage modules that simultaneously meet the conditions of low voltage deviation coefficients and high current deviation coefficients are marked as priority control targets, and the system assigns these targets a higher control priority flag.

[0067] After identifying the priority control target, its instantaneous temperature value needs to be used to evaluate its real-time power carrying capacity. The instantaneous temperature value is input into a pre-established battery electrothermal coupling model, which outputs the maximum safe charge and discharge current threshold allowed for the cell at the current temperature. The evaluation process also considers the influence of temperature on internal resistance and aging rate. For a module marked as a priority control target, if its instantaneous temperature value is within the optimal operating range, it is considered to have a high power carrying capacity and can be assigned more power adjustment tasks. If its temperature is close to the safety limit, its power carrying capacity will be reduced even if its electrical signal meets the priority conditions. Based on the combined priority and power carrying capacity assessment results, the system generates an initial power allocation ratio. This ratio is a weight vector, where each element corresponds to the proportion of the total power that an energy storage module should share. The generation algorithm adopts a constrained optimization method, aiming to allow modules with strong power carrying capacity and stable voltage to share more of the load while meeting the total power demand, and at the same time to smooth out the fluctuations of modules with large current fluctuations. This initial power allocation ratio is converted into a specific and executable set of power setpoints, namely the power allocation control instruction set. The instruction set specifies the target charging and discharging power threshold for each energy storage module in the next control cycle.

[0068] After the power distribution control instruction set is generated, the energy flow path configuration update process is immediately triggered. The update process first needs to obtain the target charging and discharging power threshold set for each module in the instruction set, and compare these target values ​​with the current actual power values ​​fed back by the sensors of each module in real time, and calculate the difference between the two. This difference not only determines the amount of power to be transmitted, but its positive or negative sign also directly indicates the direction of energy flow. A positive difference indicates that the module needs to absorb more power (charging), and a negative difference indicates that power needs to be released (discharging). After determining the energy flow direction and transmission magnitude, the system needs to replan the physical connection paths between modules to achieve the specified energy allocation. Path planning relies on a programmable power routing matrix, which consists of a series of bidirectional DC-DC converters and solid-state switches. The planning algorithm abstracts the energy storage cabinet system as a network graph, where nodes are energy storage modules and edges are possible energy transmission paths. The algorithm is based on the shortest path or maximum flow principle in graph theory, but the optimization objective is to minimize transmission losses, balance switching device losses, and avoid circulating currents. For modules that need to receive power, the algorithm will find one or more most suitable power source modules (i.e., modules that need to release power) and calculate the current value that needs to flow on each path. The results of path planning are compiled into a clear energy flow path configuration table. This table records in detail, in list form, which modules need to establish energy connections in the next control cycle, from which module to which module the energy flows, and the planned power value. The configuration table also includes the instruction codes and execution timestamps for controlling the corresponding switching devices.

[0069] This generated energy flow path configuration table is sent to the power routing actuator via an internal communication bus. The power routing actuator is typically a distributed controller network, with each controller responsible for operating a set of solid-state switches. The controller parses the received configuration table and executes the closing and opening operations of the switches at precisely synchronized times, thereby physically reconstructing the electrical connections between energy storage modules and establishing new energy flow paths. This process realizes the final mapping of power distribution commands from the digital domain to the physical domain, providing the hardware foundation for on-demand, efficient, and balanced energy flow. The core of the closed-loop automatic control, which is then physically executed, lies in dynamically adjusting the system topology and power distribution based on the real-time status of the modules.

[0070] Example 3: Generation of thermal state correction parameters and secondary calibration based on these parameters. This process is initiated after the initial power allocation command set is issued and executed. Its aim is to further optimize the power allocation strategy based on the actual thermal response of the energy storage modules to improve system operational safety. After the power adjustment command is executed, the system's embedded thermal monitoring unit collects the instantaneous temperature values ​​of each energy storage module at a higher frequency. This collection frequency is typically higher than the main control cycle frequency to capture more subtle thermal dynamic changes. The collected temperature data stream is sent to a first-in-first-out data buffer. Calculating the absolute value of temperature rise per unit time is the first key operation. This calculation is performed independently for each module. The algorithm takes the current temperature value and subtracts a reference temperature value. This reference temperature is defined as the temperature reading recorded at the moment the power adjustment command takes effect. Then, the resulting temperature difference is divided by the time interval from the reference time to the current time, thus obtaining a scalar value reflecting the average temperature rise per unit time of the module after the power change, i.e., the absolute value of temperature rise. This value directly characterizes the heat generation rate of the module after bearing a new power load.

[0071] The system compares the calculated absolute temperature rise of each module with a preset safety threshold. This safety threshold is not fixed but dynamically adjusted based on the cell's chemical system, historical operating data, and the current ambient temperature. For example, for lithium-ion batteries, this threshold might be set within the range of 1 to 3 degrees Celsius per minute. The comparison is performed by a comparator circuit or software logic, which outputs a Boolean flag indicating which modules have exceeded the safety boundary in terms of temperature rise rate. These modules are classified as over-limit temperature rise energy storage modules. For each identified over-limit temperature rise module, the system further extracts its voltage fluctuation range within the same observation time window. The voltage fluctuation range is obtained by finding the maximum and minimum voltage sampling values ​​within that time period and calculating their difference. This range reflects the stability of the module's terminal voltage. Subsequently, the system calculates a parameter called the thermo-coupling influence factor, which is used to quantify the degree of influence of temperature rise on the electrical stability of the module. The calculation process is based on a pre-stored empirical mapping relationship, which describes the correlation strength between temperature rise change and voltage fluctuation at different initial temperatures and operating points. For a given module with excessive temperature rise, if its higher absolute temperature rise value is accompanied by a larger voltage fluctuation range, a higher thermo-coupling influence factor will be calculated, indicating that the thermal effect has a more significant impact on its electrical performance.

[0072] Based on the calculated thermo-coupling influence factor, the system generates the final thermo-state correction parameter. This parameter is usually a normalized coefficient, and its value is positively correlated with the thermo-coupling influence factor. The generation logic is that for modules with a high thermo-coupling influence factor, their thermo-state correction parameter will be adjusted in the direction of limiting power. The correction parameter is encapsulated into a vector data structure, and each element corresponds to a correction instruction for an energy storage module.

[0073] Once the thermal state correction parameters are generated, they are immediately used to perform a secondary calibration of the currently effective initial power allocation ratio. The calibration process employs a weighted fusion algorithm, which takes the thermal state correction parameters and the initial power allocation ratio as inputs. Its mathematical expression is as follows:

[0074] P final_ratio (i)=P initial_ratio (i)·[1-α·Δ thermal (i)]

[0075] Where: symbol P final_ratio (i) represents the power allocation ratio coefficient determined after secondary calibration for the i-th energy storage unit. This coefficient determines the proportion of power that unit should bear in the total system power. Symbol P initial_ratio (i) represents the proportional coefficient obtained in the initial power allocation calculation of the i-th energy storage unit. It is the baseline value allocated without considering the detailed thermal state effects. The Greek letter α is a system-level weighting coefficient that adjusts the degree of influence of the thermal state correction on the final power allocation result. Its value needs to be set according to the overall thermal management strategy and safety requirements of the energy storage system. Symbol Δ thermal (i) Specifically refers to the value of the thermodynamic state correction parameter corresponding to the i-th energy storage unit, which directly reflects the current thermal load state of the unit.

[0076] The weighted fusion process essentially incorporates thermal stress factors into the power allocation decision. For modules with large thermodynamic state correction parameters (i.e., modules with rapid temperature rise and significant thermo-electric coupling effects), the above formula will calculate a scaling factor less than 1, thus affecting the final power allocation ratio P. final_ratio (i) relative to the initial proportion P initial_ratio(i) Power is reduced, resulting in a decrease in power output. Conversely, for modules with good thermal conditions, the power ratio may be maintained or only slightly adjusted. After generating the corrected power allocation ratio, the system needs to recalculate the specific charge and discharge power thresholds for each energy storage module based on this new ratio. The recalculation must be constrained by the total power demand of the system to ensure that the sum of the power thresholds of all modules still meets the global charge and discharge commands. The calculation process involves normalization of the ratio and verification of the power limits of each module. Finally, the system uses the newly calculated power thresholds to comprehensively update the original power allocation control command set. This updated command set includes optimized information on thermal management and is then issued for execution, thus completing a complete closed-loop optimization of power allocation that considers thermal effects. This process continuously cycles, dynamically responding to the constantly changing thermal state inside the energy storage system. The entire implementation process reflects a deep focus on the electro-thermal coupling effect. By embedding a thermal feedback loop into the power allocation strategy, the system's ability to prevent local overheating and ensure long-term operational reliability is significantly enhanced.

[0077] Example 4: The generation process of dynamic equalization compensation parameters is activated during the system's coordinated charge-discharge operation based on the optimized power allocation control instruction set. Its core task is to evaluate the voltage and temperature consistency level of each energy storage module under group operation and generate precise compensation instructions accordingly. Consider a subsystem containing four energy storage modules (labeled as Module A, Module B, Module C, and Module D) performing a discharge operation. The system continuously collects the real-time voltage and temperature values ​​of each module at millisecond intervals. The real-time voltage value is acquired through a high-precision analog-to-digital converter. The voltage signal of each module is isolated and filtered before being synchronously sampled and recorded. The real-time temperature value is acquired using multiple temperature sensors embedded in key hot spots of the battery module. Their readings are fused using an algorithm to obtain a single temperature value representing the overall thermal state of the module. Within a complete evaluation cycle, the system continuously collects several sets of such voltage and temperature data pairs, forming a time series.

[0078] The system calculates the statistical standard deviation of the real-time voltage values ​​of each module within the time period. This calculation is performed on the voltage values ​​of all online modules at each sampling time. For example, at time t1, the voltage values ​​of modules A, B, C, and D are read, and the standard deviation of these four values ​​is immediately calculated. This process is repeated at each sampling time. Finally, the average of all standard deviations over the entire time period is taken as the core indicator for evaluating voltage consistency. A smaller average standard deviation indicates better voltage synchronization and less variation among modules. The system also calculates the gradient difference of real-time temperature values. The gradient difference calculation aims to characterize the uniformity of temperature distribution among modules. It is not simply the maximum or minimum temperature difference, but rather quantifies the thermal gradient by analyzing the spatial distribution of the temperature field. The algorithm first determines the physical coordinates of each energy storage module within the cabinet, then assigns the temperature value of each module to its coordinates. Next, a continuous virtual temperature field surface is generated using spatial interpolation. The gradient difference is calculated by finding the direction and rate of change of the most drastic temperature change on this surface. Finally, the maximum temperature gradient value within the entire area is extracted as the gradient difference indicator. A large gradient difference indicates the presence of significant hot spots and uneven temperature distribution within the cabinet.

[0079] The voltage standard deviation and temperature gradient difference are combined to generate a comprehensive consistency assessment index. This combination typically involves assigning weight coefficients to these two indices, which have different physical meanings and dimensions, and then performing a weighted sum. The weight coefficients reflect the system's emphasis on voltage consistency and temperature uniformity. For example, in applications where electrical balance is more important, the weight of the voltage standard deviation might be set higher. The calculation results of the consistency assessment index are directly used to determine the strength and direction of subsequent compensation. The compensation strength is positively correlated with the index value; that is, the worse the consistency (the larger the index value), the greater the required compensation intervention. The determination of the compensation direction is more detailed, requiring separate calculations for voltage and temperature. For voltage, the direction is determined by comparing the voltage value of each module with the average voltage. Modules below the average voltage require compensatory reduction of discharge power or increase of charging power (positive compensation), while modules above the average voltage require the opposite (negative compensation). For temperature, the direction is determined by the deviation of each module's temperature from the average temperature or ambient temperature. Modules with excessively high temperatures require compensatory load reduction (negative compensation) to achieve cooling.

[0080] Based on the determined compensation intensity and direction, the system generates the final dynamic equilibrium compensation parameters. These parameters are typically a multi-dimensional vector, where each element corresponds to an energy storage module and includes the voltage and temperature compensation amounts for that module. These compensation amounts are standardized scalars for direct use by the subsequent control unit. Table 1 shows the raw data and calculated intermediate indicators for the four energy storage modules at the end of a hypothetical evaluation period. This data is used to generate the dynamic equilibrium compensation parameters.

[0081] Table 1: Operating data and consistency indicators of energy storage modules during coordinated charging and discharging operations

[0082]

[0083] Based on the data in the table above, the voltage standard deviation is 0.25V. The temperature gradient difference is reflected by calculating the maximum temperature difference (32.5-29.5=3.0℃) and combining it with location information. Assuming that the consistency evaluation index value calculated after weight allocation is at a medium level, this indicates that a medium intensity of compensation is required. The compensation direction is determined as follows: For voltage, module A has the highest voltage and needs to be negatively compensated (it is recommended to moderately increase its discharge load or reduce its charging current). Module B has the lowest voltage and needs to be positively compensated (it is recommended to moderately reduce its discharge load or increase its charging current). The voltages of modules C and D are close to the average value, and the compensation amount is very small. For temperature, module A has the highest temperature and needs to be negatively compensated (it is strongly recommended to reduce its load to help cool it down). Module C has the lowest temperature and its load-bearing potential is relatively large.

[0084] The final generated dynamic equalization compensation parameter vector will contain a synthetic compensation command for each module, which integrates voltage and temperature compensation directions. For example, the command for module A may contain a large negative compensation amount (due to its high voltage and high temperature), while the command for module B may contain a medium positive voltage compensation amount (due to its low voltage) and a slight negative temperature compensation amount (due to its temperature being slightly above average). These parameters are encapsulated in a data packet and transmitted to the power allocation weight update module, providing a direct basis for the next stage of optimization control. The entire implementation process reflects the ability to finely perceive and close-loop adjust the consistency of the system's collective behavior.

[0085] Example 5: The final two steps are updating the power allocation weights and completing the distributed cooperative control closed loop. This process transforms the dynamic equilibrium compensation parameters calculated in the previous stage into final executable control commands and verifies their execution effect. Updating the power allocation weights begins by superimposing the dynamic equilibrium compensation parameters with the corrected power allocation ratio. The dynamic equilibrium compensation parameter is a vector, with each element corresponding to a compensation suggestion for an energy storage module, usually existing in the form of a relative change. For example, the compensation parameter for a module might be "increase the power output ratio by 5%" or "decrease the power absorption ratio by 3%". The corrected power allocation ratio is also a vector, representing the theoretical power sharing share of each module after considering the thermal state. The superposition operation is not a simple arithmetic addition, but a weighted superposition process with saturation constraints. The arithmetic unit reads the corrected power allocation ratio value of each module, and then adjusts the ratio value upward or downward according to the direction and intensity indicated by the compensation parameters. For example, for a module with a current allocation ratio of 25%, if its dynamic balance compensation parameters indicate that positive compensation (i.e., increased load) is needed, its ratio is increased by 2 percentage points to 27%. At the same time, another module with a ratio of 25% but requiring negative compensation will have its ratio reduced by 2 percentage points to 23%. This process needs to follow the constraint that the total ratio sum is 100%. Therefore, the arithmetic unit will perform a round of normalization to ensure that the sum of the power allocation weights of all modules is still exactly equal to 100% after the adjustment, thereby generating the final power allocation weight vector.

[0086] After the final power allocation weight vector is generated, the system immediately initiates the distributed collaborative control closed-loop completion process. The primary task of this process is to encode the final power allocation weights into control commands that the underlying control terminals can recognize and execute. The encoding process relies on a predefined communication protocol. The data structure of the control commands typically includes fields such as a unique identifier for the target module, a timestamp of the command's effective date, the expected absolute or relative change in charging and discharging power, and the effective duration of the command. The encoder combines the final power allocation weights with the system's current total power demand to calculate the specific target power value for each module and fills it into the corresponding fields of the command frame. The encoded control commands are broadcast to the control terminals of each energy storage module through a distributed communication network. The communication network typically uses a topology such as industrial Ethernet or CAN bus with deterministic latency to ensure that the commands arrive at all terminals almost simultaneously. The broadcast message contains a global sequence number. After receiving the command, all terminals need to return an acknowledgment message with the same sequence number. If the main controller does not receive acknowledgments from all terminals within a preset time, it will trigger a command retransmission or fault handling process. Upon receiving the command, the control terminal of each energy storage module immediately drives the local power converter (such as a DC-DC converter or inverter) to adjust its output or input power to approach the target value. The adjustment process is not instantaneous, but involves a dynamic response process. During this period, the system continuously verifies the matching degree between the actual power output of each energy storage module and the command.

[0087] The verification process involves high-frequency acquisition of real-time current and voltage values ​​at the output terminals of each module and calculation of their instantaneous power. The main controller continuously compares the acquired actual power values ​​with the target power values ​​required by the command, calculating their instantaneous deviation. The matching degree is usually not evaluated based on absolute equality at a certain moment, but rather on whether the actual power curve can quickly and smoothly track the target power curve within a short time window. The integral value of the deviation (such as the cumulative error over a period of time) and the maximum dynamic deviation are commonly used evaluation indicators. When the system confirms that the matching degree between the actual power output of all energy storage modules and the command exceeds a preset threshold (for example, the actual power is stable within ±2% of the target power), and the overall system operation is stable, it is determined that the control command has been correctly executed. The closed-loop confirmation signal is set, marking the successful completion of a complete distributed collaborative control closed loop from state perception, decision analysis to execution verification. The system then prepares to enter the next control cycle, starting a new round of data acquisition and optimization, thereby achieving continuous adaptive optimization operation. The entire implementation process embodies the closed-loop control concept of instruction generation, reliable transmission, precise execution, and effect verification, ensuring that the optimization strategy can be accurately implemented and produce the expected results.

[0088] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0089] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for optimized control of intelligent lithium-ion modular energy storage cabinets, characterized in that, The method includes: The voltage, current and temperature data of each energy storage module are acquired in real time through distributed data acquisition nodes to generate a dynamic set of operating parameters. Based on the dynamic operating parameter set analysis, the differences in the charging and discharging states of each energy storage module are analyzed, the power balance requirements between modules are calculated, and a power allocation control instruction set is generated. Adjust the charging and discharging power thresholds of each energy storage module according to the power distribution control instruction set, and synchronously update the energy flow path configuration between modules; Monitor the temperature rise rate and voltage fluctuation range of the energy storage module after power adjustment, and generate thermodynamic state correction parameters; The charge and discharge power thresholds are calibrated a second time by combining the thermodynamic state correction parameters to generate an optimized power distribution control instruction set. Based on the optimized power distribution control instruction set, the energy storage module performs coordinated charging and discharging operations and collects real-time coordinated operation data. Analyze the voltage consistency deviation and temperature gradient distribution in real-time collaborative operation data to generate dynamic equalization compensation parameters; The energy flow path configuration is reconstructed based on the dynamic equilibrium compensation parameters, and the power allocation weight of the energy storage module is updated. The final power allocation weight is output to the energy storage module control terminal to complete the distributed collaborative control closed loop.

2. The intelligent lithium-ion modular energy storage cabinet combination optimization control method according to claim 1, characterized in that, The specific steps for generating the dynamic operating parameter set are as follows: A fixed sampling period is set, and the instantaneous voltage, instantaneous current, and instantaneous temperature values ​​of each energy storage module are synchronously collected through distributed data acquisition nodes; The instantaneous voltage, current, and temperature values ​​collected within the same sampling period are timestamped to align the data and remove abnormal jump data. Calculate the rate of change of voltage, current and temperature for each energy storage module during a continuous sampling period, and generate a dynamic trend matrix. Integrate the dynamic trend matrix of all energy storage modules to construct a dynamic operating parameter set.

3. The intelligent lithium-ion modular energy storage cabinet combination optimization control method according to claim 2, characterized in that, The specific steps for generating the power allocation and control instruction set are as follows: Extract the voltage and current change rates of each energy storage module from the dynamic operating parameter set, and calculate the voltage and current deviation coefficients between modules; Power allocation priority is determined based on voltage deviation coefficient and current deviation coefficient, and energy storage modules with low voltage deviation coefficient and high current deviation coefficient are marked as priority control targets; The power carrying capacity of the priority control object is evaluated by combining the instantaneous temperature value, and an initial power allocation ratio is generated; A power allocation control instruction set is generated based on the initial power allocation ratio.

4. The intelligent lithium-ion modular energy storage cabinet combination optimization control method according to claim 3, characterized in that, The specific steps for updating the energy flow path configuration are as follows: Obtain the target charge and discharge power thresholds for each energy storage module in the power allocation and control instruction set; Analyze the difference between the target charging / discharging power threshold and the current actual power to determine the energy flow direction and transmission magnitude; Based on the energy flow direction and transmission magnitude, the physical connection paths between modules are replanned, and an energy flow path configuration table is generated. The energy flow path configuration table is sent to the power routing executor.

5. The intelligent lithium-ion modular energy storage cabinet combination optimization control method according to claim 4, characterized in that, The specific steps for generating the thermodynamic state correction parameters are as follows: Monitor the instantaneous temperature values ​​of each energy storage module after power adjustment, and calculate the absolute value of temperature rise per unit time; By comparing the absolute value of the temperature rise with the preset safety threshold, energy storage modules that exceed the temperature rise limit can be identified. Extract the voltage fluctuation range of the over-limit temperature rise energy storage module and calculate the thermo-coupling influence factor; Thermodynamic state correction parameters are generated based on the thermodynamic coupling influence factor.

6. The intelligent lithium-ion modular energy storage cabinet combination optimization control method according to claim 5, characterized in that, The specific steps of the secondary calibration are as follows: The thermal state correction parameters are weighted and fused with the initial power allocation ratio to generate the corrected power allocation ratio; The charging and discharging power thresholds of each energy storage module are recalculated based on the revised power allocation ratio. Update the power distribution control instruction set.

7. The intelligent lithium-ion modular energy storage cabinet combination optimization control method according to claim 6, characterized in that, The specific steps for generating the dynamic equilibrium compensation parameters are as follows: Collect real-time voltage and temperature values ​​of each energy storage module during the coordinated charging and discharging operation; Calculate the standard deviation of real-time voltage values ​​and the gradient difference between real-time temperature values ​​to generate a consistency evaluation index. The compensation intensity and direction are determined based on the consistency assessment indicators, and dynamic equilibrium compensation parameters are generated.

8. The intelligent lithium-ion modular energy storage cabinet combination optimization control method according to claim 7, characterized in that, The specific steps for updating the power allocation weights are as follows: The dynamic balance compensation parameters are superimposed with the corrected power allocation ratio, and the power output ratio of each energy storage module is redistributed based on the superposition result to generate the final power allocation weight.

9. The intelligent lithium-ion modular energy storage cabinet combination optimization control method according to claim 8, characterized in that, The specific steps for completing the distributed collaborative control closed loop are as follows: The final power allocation weights are encoded into control commands; Control commands are broadcast to the control terminals of each energy storage module via a distributed communication network; Verify the matching degree between the actual power output of each energy storage module and the command, and complete the closed-loop confirmation.

10. A smart lithium-ion modular energy storage cabinet combination optimization control system, comprising a memory, a processor, and a computer program stored in the memory and running on the processor, characterized in that, When the processor executes the computer program, it implements the steps of the intelligent lithium-ion modular energy storage cabinet combination optimization control method as described in any one of claims 1 to 9.

Images