Battery management system and method for new energy storage

Abstract

The invention discloses a battery management system and method, and particularly relates to a battery management system and method for new energy storage. A slave control monitoring and balancing module; a hybrid equalization circuit; and a distributed precise heat management execution unit. According to the battery management system and the control method thereof provided by the invention, through deep cooperation of an innovative hardware architecture and an intelligent algorithm, multilayer and systematic remarkable beneficial effects are brought. According to the system, the consistency performance of the battery pack can be greatly improved, the hybrid equalization topology is combined with the rapid response of inductor equalization in the modules and the wide-area scheduling capability of capacitor equalization between the modules, and accurate redistribution of energy from a battery cell level to a cluster level is realized, so that the voltage inconsistency is actively inhibited within an extremely low range, and the system is suitable for large-scale popularization and application. Local over-charging and over-discharging are effectively avoided, and the overall capacity attenuation of the battery pack is fundamentally delayed.

Description

Technical Field

[0001] This invention relates to the field of battery management equipment technology, and in particular to a battery management system and method for new energy storage. Background Technology

[0002] With the continuous increase in the proportion of new energy power generation, electrochemical energy storage systems have become a key link in ensuring grid stability and achieving efficient energy utilization. Large-scale battery energy storage systems, composed of thousands or even tens of thousands of individual battery cells connected in series and parallel, rely heavily on the performance of the battery management system for their safety, lifespan, and efficiency. Existing battery management systems face numerous challenges in practical applications. Regarding state estimation, the commonly used equivalent circuit model based on fixed parameters and the ampere-hour integral method struggles to accurately describe the dynamic electrochemical processes within the battery, leading to a significant decrease in the accuracy of state of charge and state of health estimation under complex operating conditions, across the entire temperature range, and after battery aging. In terms of equalization management, energy-consuming passive equalization schemes suffer from high energy loss and low efficiency, while single active equalization topologies suffer from limited energy transfer paths and high costs, making it difficult to achieve rapid and efficient consistency maintenance in large battery clusters. Furthermore, the system's thermal management, state estimation, and equalization control modules are typically independent and lack coordination, failing to perform linked optimization control based on the actual state of the batteries. For example, if the system fails to proactively adjust its charging and discharging strategy and initiate directional equalization and precise temperature control in the early stages of an increase in the internal resistance of a battery cell, it can lead to an accelerated expansion of battery inconsistencies, ultimately affecting the lifespan and safety of the entire system. Therefore, developing a battery management system capable of high-precision state estimation, high-efficiency equalization management, and multi-module collaborative control is crucial for promoting the development of new energy storage technologies. Summary of the Invention

[0003] In order to solve the problems existing in the prior art, the present invention provides a battery management system and method for new energy storage, so as to solve the current technical problems.

[0004] The technical solution adopted by this invention to solve its technical problem is: This invention provides a battery management system and method for new energy storage, comprising: The main control unit includes a core chip and peripheral circuits. The slave control monitoring and balancing module includes: a data acquisition module; A hybrid equalization circuit, comprising: an architecture of multiple switch matrices combined with a central power inductor; A distributed precision thermal management execution unit, comprising: an actuator and a drive circuit.

[0005] Preferred management methods include the following: Step 1: High-precision synchronous data acquisition and preprocessing; Step 2: Multi-state joint estimation based on dynamic parameter identification; Step 3: Construction and solution of the multi-objective optimization decision engine; Step 4: Distribution and execution of control commands.

[0006] Preferably, it also includes system security monitoring and fault tolerance mechanisms, which include: hardware redundancy, software lockstep, and fault prediction and health management.

[0007] The beneficial effects of this invention are: The battery management system and control method provided by this invention, through the deep synergy of its innovative hardware architecture and intelligent algorithms, bring significant multi-level and systematic benefits. This system can greatly improve the consistency performance of the battery pack. Its hybrid equalization topology combines the rapid response of inductor equalization within modules with the wide-area scheduling capability of capacitor equalization between modules, achieving precise energy redistribution from the cell level to the cluster level. This actively suppresses voltage inconsistencies within an extremely low range, effectively avoiding local overcharging and over-discharging, and fundamentally delaying the overall capacity decay of the battery pack.

[0008] In terms of energy management, this invention achieves a leap from extensive to precise methods. The linkage between distributed and bonded semiconductor cooling chips and state estimation enables the system to perform point-to-point precise temperature control on specific heat-generating modules, rather than indiscriminately cooling the entire system. This on-demand thermal management strategy can significantly reduce the energy consumption of the temperature control system itself.

[0009] This invention represents a leap from reactive, post-event protection to proactive early warning and coordinated suppression. Its high-precision synchronous data acquisition and dynamic parameter identification capabilities enable it to keenly detect key signals indicative of early battery aging, such as slight increases in internal resistance, and issue timely warnings. More importantly, when a potential risk is identified, the system can activate a combination of strategies, including current derating, directional equalization, and enhanced cooling, to suppress fault development through multiple measures. While ensuring a safety baseline, it also maintains continuous system operation as much as possible, significantly enhancing system reliability and availability.

[0010] This system integrates previously isolated subsystems such as state estimation, equilibrium management, and thermal control into a cohesive whole through a multi-objective optimization decision-maker. When formulating strategies, the decision-maker globally considers equilibrium benefits and circuit losses, cooling effects and energy costs, seeking the optimal solution for overall system operating efficiency, thus avoiding efficiency friction caused by conflicting objectives between subsystems. This integrated collaborative optimization model, coupled with extended lifespan and reduced operation and maintenance costs, significantly lowers the levelized cost of electricity (LCOE) of energy storage and improves the project's total lifecycle return on investment.

[0011] Furthermore, the joint state estimation algorithm based on dynamic parameter identification enables the estimation accuracy of SOC and SOH to adapt to battery aging and complex operating conditions, resulting in more reliable results. This not only provides a solid data foundation for coordinated control within the system, but its high-reliability battery state data also provides indispensable core data support for the optimized scheduling of upper-level energy management systems, as well as advanced applications such as future battery asset residual value assessment and cascade utilization. Attached Figure Description

[0012] The above-described aspects and advantages of the present invention will become apparent and readily understood from the following description of the embodiments in conjunction with the accompanying drawings, in which: Figure 1 This is a schematic diagram of a battery management system and method for new energy storage according to an embodiment of the present invention. Detailed Implementation

[0013] 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.

[0014] This invention provides a battery management system and method for new energy storage, including hardware architecture and management methods.

[0015] The hardware architecture of a battery management system for new energy storage includes: The hardware of this system adopts a three-layer distributed architecture of master control, slave control, and single-unit monitoring.

[0016] Main controller unit: Core Chip Selection and Division of Labor: The main control unit employs a heterogeneous multi-core processor, such as TI's AM64x (dual-core A53 + dual-core R5F) or NXP's i.MX 8XLite (Arm Cortex-A35 + Cortex-M7). The high-performance application core (A53 / A35) runs a Linux operating system and is responsible for complex state estimation algorithms, historical data storage, and gigabit Ethernet or 5G communication with the energy management system (EMS) and cloud platform. The real-time control core (R5F / M7) runs on bare metal or a real-time operating system (such as FreeRTOS) and is responsible for hard real-time tasks, including: periodic and event-triggered communication with all slave modules via a high-speed CAN FD bus (up to 5Mbps); receiving and verifying synchronization data packets uploaded by slave modules; executing system-level fault protection algorithms (such as insulation detection and short-circuit judgment based on total voltage and current) and triggering safety relay actions within microseconds.

[0017] Peripheral circuitry: Equipped with an isolated CAN FD transceiver, a high-precision total voltage sampling circuit (16-bit ADC, error <0.1%), a shunt-type total current sampling circuit (1000A / 100mV, 16-bit ADC), and an external watchdog circuit to ensure system robustness.

[0018] Slave control monitoring and balancing module: Each slave module manages a battery module consisting of 24 lithium-ion cells connected in series (nominal voltage approximately 76.8V). This design is the core of the hardware of this invention.

[0019] High-precision data acquisition: The core chip used is the ADI LTC6815-1, which supports voltage measurement of up to 18 battery cells (up to 24 cells can be achieved through cascading). However, we use two LTC6815-1 chips in a redundant manner to measure 24 battery cells, improving reliability. Its main performance parameters include: voltage measurement range of 0V to 5V, typical total measurement error of less than 1.8mV. It has a built-in 22-bit Σ-Δ ADC and an integrated multiplexer. It also has a built-in accurate voltage reference source and can perform self-calibration before measurement. It can synchronously acquire the voltage of all battery cells, eliminating errors caused by asynchronous acquisition time. It integrates 9 general-purpose ADCs for connecting to NTC thermistors to accurately measure the temperature of key points in each battery cell or module.

[0020] Hybrid equalization circuit: In-module inductive active balancing circuit: Topology: A non-isolated BUCK-BOOST architecture using a multi-switch matrix and a central power inductor. Specifically, a 24-to-2 switch matrix composed of high-current MOSFETs (such as the Infineon OptiMOS™ series, with on-resistance Rds(on) < 1mΩ) can flexibly connect the central inductor between any two cells to be balanced. Operating modes: Taking energy transfer from cell C10 to cell C5 as an example: Charging stage: The controller drives the switch matrix to connect one end of inductor L to the positive and negative terminals of cell C10 through MOSFETs Q10a and Q10b, forming a loop. At this time, C10 charges inductor L, and the inductor current rises linearly. By controlling the duration of this stage (e.g., 50μs), the transferred energy can be precisely controlled. Discharging stage: The controller turns off Q10a and Q10b, and then immediately (with a dead time set to prevent shoot-through) turns on MOSFETs Q5a and Q5b connected to cell C5. The energy stored in the inductor is released to C5 through Q5a and Q5b. Key components: The central power inductor is an iron-silicon-aluminum magnetic ring inductor with an inductance of approximately 22μH and a saturation current greater than 20A. The MOSFET driver chip needs to have a high peak current output capability (e.g., 2A) to ensure fast switching and reduce switching losses.

[0021] Inter-module capacitive active balancing circuit: Topology: Employs a flying capacitor architecture. Each slave module is equipped with a high-voltage film capacitor (e.g., 10μF / 100V, MKP type) and connected to the total positive and total negative terminals of the module it manages via an H-bridge circuit consisting of four MOSFETs. Operating modes: Taking energy transfer from module A to module B as an example: Charging phase: Controlling the upper transistor of the H-bridge in module A and the lower transistor of the H-bridge in module B (or using a dedicated charging switch) connects capacitor Cfly in parallel to module A, rapidly charging the capacitor voltage to the module A voltage. Switching and discharging phase: Disconnecting all switches connected to module A, then controlling the upper transistor of the H-bridge in module B and the lower transistor of the H-bridge in module A connects the charged Cfly in parallel to module B. Since the Cfly voltage is higher than the module B voltage at this time, the capacitor discharges to module B. Control strategy: The main controller periodically scans the average voltage of all modules and schedules the corresponding capacitors of the module pairs with the highest and lowest voltages for energy transfer. This balancing method is highly efficient and is especially suitable for scenarios where there is a large pressure difference between modules.

[0022] Distributed precision thermal management execution unit: Actuator: 2-4 TEC1-12706 type semiconductor cooling chips are attached to the aluminum heat dissipation substrate of each battery module. The hot end of the chip is in contact with the liquid cooling plate through thermal grease, and the cold end directly cools the module.

[0023] Drive circuit: Integrated on the slave control module, it adopts an H-bridge DC voltage regulation circuit. Through PWM control, the magnitude and direction of the current flowing through the cooling chip can be precisely adjusted, thereby achieving bidirectional temperature control for either cooling or heating. The main controller can independently set the target temperature value for each module, realizing "on-demand cooling" and greatly reducing the overall energy consumption of the thermal management system.

[0024] Battery management methods for new energy storage include: Step 1: High-precision synchronous data acquisition and preprocessing The main control R5F / M7 core broadcasts a "synchronous acquisition" command frame every 100ms via CAN FD.

[0025] Upon receiving a command, all slave modules' LTC6815 chips simultaneously initiate all cell voltage and temperature measurements on the next internal clock edge.

[0026] After the measurement is completed, the data is packaged and uploaded to the main controller via daisy chain or isolated CAN FD.

[0027] The main controller performs CRC checks and validity checks on the received data (e.g., whether the voltage is within the reasonable range of 0V-5V). Abnormal data is marked and may trigger a retest.

[0028] Step 2: Multi-state joint estimation based on dynamic parameter identification Dynamic model parameter identification: The algorithm starts when the battery is in the constant current charge / discharge phase. It extracts the voltage and current sequences over a past period (e.g., 60 seconds) and uses a recursive least squares method with a forgetting factor to identify the parameters of the first-order RC equivalent circuit model online: {open-circuit voltage OCV(SOC), ohmic internal resistance R0, polarization resistance R1, polarization capacitance C1}. The forgetting factor is typically set to 0.995 to allow the model to track slow parameter changes.

[0029] Co-estimation of SOC and SOH: The identified dynamic model parameters are injected into a dual extended Kalman filter. One filter is responsible for estimating the system state (SOC, surface concentration, etc.), and the other filter is responsible for estimating the model parameters (R0, R1, C1).

[0030] SOH estimation is divided into capacity SOH and internal resistance SOH. Capacity SOH is updated by comparing the capacity calculated using the ampere-hour integration method with the capacity calibrated based on the dynamic model OCV-SOC curve. Internal resistance SOH is obtained by comparing the currently identified R0 (normalized to 25°C) with the battery's initial internal resistance. These updates are typically performed at the end of a complete charge-discharge cycle to ensure accuracy.

[0031] Step 3: Construction and solution of the multi-objective optimization decision engine When the state estimation identifies inconsistencies (such as a cell's SOC deviating from the cluster average by more than 3%, or the temperature being more than 5°C higher), the decision-maker is activated.

[0032] Constructing the objective function: The objective function J aims to minimize system inconsistencies while taking into account both efficiency and security.

[0033] J = w1 * Σ(voltage deviation)² + w2 * Σ(temperature deviation)² + w3 * (equilibrium energy loss) + w4 * (thermal management energy consumption) Among them, w1~w4 are weighting coefficients that can be dynamically adjusted according to the system's operating mode. For example, w1 (voltage consistency weight) will increase to prevent overcharging at the end of the charging process; w2 (temperature consistency weight) will increase in high-temperature environments.

[0034] Constraints: The SOC, voltage, and temperature of each battery cell must be within safe upper and lower limits. Maximum current capacity of the balancing circuit (e.g., inductor balancing current limit 5A, capacitor balancing current limit 2A). Maximum cooling / heating power of the semiconductor.

[0035] Solution and Decision: This is a constrained optimization problem. Due to the high real-time requirements of the system, we use a combination of heuristic rules and simplified model predictive control for rapid solution. The solution result is a set of optimal control actions, such as: Action A: For slave module 3, initiate inductor equalization, transferring energy from C14 and C16 to C15, with a target equalization current of 3A. Action B: Increase the power of the semiconductor cooler in module 3 to 70%. Action C: Request EMS to reduce the charging current of the entire battery cluster by 10%.

[0036] Step 4: Distribution and execution of control commands The master controller will optimize and determine the action instructions and send them to the corresponding slave control modules through CAN FD message packets synchronized with timestamps.

[0037] After receiving the command, the slave module's local MCU (such as an ARM Cortex-M3) is responsible for execution: Based on the balanced current target of the command, it generates a corresponding PWM signal to precisely control the MOSFET switching matrix and the inductor charging and discharging timing. Based on the temperature control power of the command, it generates a corresponding PWM to drive the thermoelectric cooler. The slave module monitors local parameters such as the balanced current, inductor temperature, and MOSFET temperature in real time, performs local protection, and feeds back the execution status to the master controller.

[0038] System security monitoring and fault tolerance mechanisms include: To ensure the system's absolute reliability, multiple layers of security protection have been implemented: Hardware redundancy: Key parameters (such as total voltage) are sampled using dual channels. A heartbeat monitoring system exists between the master and slave controllers; if communication times out, the slave controller can enter a degraded safety mode based on its local policy.

[0039] Software lockstep: The main control unit runs key protection algorithms in real time, and its calculation results are cross-validated with the calculation results of the application core.

[0040] Fault prediction and health management: The system continuously records the cumulative operating time of the equalization circuit, the number of MOSFET switching times, the temperature rise history of the inductor, etc., and predicts potential faults through data mining to achieve predictive maintenance.

[0041] The system workflow includes: Power-on initialization: system self-test (relay activation test, communication loop test), load battery parameters, initialize filter and state estimator.

[0042] Steady-state operation cycle: a. Data acquisition cycle (100ms): Execute step one. b. Status assessment cycle (200ms): Execute step two, update the status of all cells. c. Decision cycle (500ms or event trigger): If inconsistency is found, execute step three. d. Control execution cycle (real-time): Execute step four. e. Data recording and uploading (1s): Package key data and upload it to the cloud platform via Ethernet.

[0043] Fault handling: If a fault is detected in any cycle, the loop will immediately exit and the corresponding fault handling procedure will be entered, ensuring safety with the highest priority.

[0044] Finally, it should be noted that the above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing embodiments or make equivalent substitutions for some of the technical features. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A battery management system and method for new energy storage, characterized in that: include: The main control controller unit includes a core chip and peripheral circuits. The slave control monitoring and balancing module includes: a data acquisition module; A hybrid equalization circuit, comprising: an architecture of multiple switch matrices combined with a central power inductor; A distributed precision thermal management execution unit, comprising: an actuator and a drive circuit.

2. The battery management system and method for new energy storage according to claim 1, characterized in that: This includes the following management methods: Step 1: High-precision synchronous data acquisition and preprocessing; Step 2: Multi-state joint estimation based on dynamic parameter identification; Step 3: Construction and solution of the multi-objective optimization decision engine; Step 4: Distribution and execution of control commands.

3. The battery management system and method for new energy storage according to claim 1, characterized in that: It also includes system security monitoring and fault tolerance mechanisms, which include: hardware redundancy, software lockstep, and fault prediction and health management.

Images