A lithium battery pack bluetooth smart BMS protection board intelligent management system

Abstract

The application belongs to the technical field of battery management, and particularly relates to a lithium battery pack Bluetooth intelligent BMS protection board intelligent management system. The system comprises a parameter self-calibration module, a Bluetooth Mesh networking module, a wake-up synchronization module and a protection timing optimization module. The parameter self-calibration module updates equivalent circuit model parameters in real time through an online parameter identification algorithm. The Bluetooth Mesh networking module realizes multi-node wireless ad hoc networking and collaborative management. The wake-up synchronization module executes a clock synchronization protocol. Through a neighbor node state query mechanism in the wake-up synchronization module, parameter collaborative sharing during wake-up initialization is realized in the multi-node Mesh network. The first wake-up node can acquire actual state information of neighbor nodes in time and correct local estimation results accordingly, so as to remove systematic estimation errors caused by parameter island phenomenon, solve the risk of SOC estimation result jumping and improper protection threshold setting caused by multi-node wake-up asynchronization.

Description

Technical Field

[0001] This invention belongs to the field of battery management technology, specifically a smart management system for a lithium battery pack Bluetooth smart BMS protection board. Background Technology

[0002] Lithium batteries, with their outstanding advantages of high energy density, long cycle life and low self-discharge rate, have been widely used in electric vehicles, energy storage power stations, power tools and electric forklifts. As a control unit that ensures the safe, stable and efficient operation of lithium battery packs, the battery management system has key functions such as battery status monitoring, state of charge estimation, cell balancing and overvoltage, overcurrent and short circuit safety protection. Its control logic and operating mechanism directly determine the safety, service life and overall energy efficiency of the lithium battery system.

[0003] Large-scale energy storage and multi-device parallel power supply scenarios generally adopt a working mode of multi-battery pack parallel networking. Each battery pack is configured with an independent BMS management node to form a multi-node distributed mesh networking architecture. Existing multi-node BMS systems generally adopt a single-node independent sleep-wake strategy. After detecting a load disconnection and no communication interaction in a static working condition, each BMS node autonomously shuts down the main control function module and only retains the basic wake-up circuit for standby, thereby reducing the system's static power consumption.

[0004] In related technologies, the independent sleep-wake mechanism has the drawback of being unsuitable for multi-node networking scenarios. The wake-up sequence of each node has inherent random differences. The BMS node that wakes up first can only complete the initialization based on the historical static parameters stored locally, and cannot synchronize the actual operating status of neighboring nodes in real time. During the battery's resting period, phenomena such as self-discharge, temperature drift, and polarization recovery will continuously change the battery's state of charge, internal resistance, and open-circuit voltage parameters, causing the initialization parameters to deviate significantly from the actual battery operating conditions. This creates a parameter islanding effect and causes a systematic deviation in the state of charge estimation. At the same time, there is a vacuum window period during the node wake-up initialization phase when the protection function is not ready. If a load is connected at this time and a large instantaneous current is generated, it is easy to exceed the rated tolerance threshold of the power device and cause hardware damage, thereby affecting the state estimation accuracy and operational safety of the multi-cluster parallel lithium battery system.

[0005] Therefore, the present invention provides an intelligent management system for a lithium battery pack Bluetooth smart BMS protection board. Summary of the Invention

[0006] In order to overcome the shortcomings of the prior art, at least one technical problem raised in the background art is solved.

[0007] The technical solution adopted by this invention to solve its technical problem is: a lithium battery pack Bluetooth smart BMS protection board intelligent management system, the system including a processing link, the processing link including: Acquire battery terminal voltage and charge / discharge current response data under current operating conditions, and trigger a wake-up signal when a load connection is detected; Based on the online parameter identification algorithm, the collected voltage and current response data are fitted in real time, and the ohmic internal resistance parameters and polarization parameters in the equivalent circuit model are dynamically updated to establish the mapping relationship between parameters and temperature, state of charge and real-time current value. Based on the mapping relationship, a collaborative correction process is applied to the local SOC estimation results during wake-up initialization, so that the local parameter estimation converges to the actual parameter state of the neighboring nodes. After completing parameter co-calibration, the protection timing optimization unit is activated. When the battery voltage is detected to be lower than the over-discharge protection threshold, the power MOSFET is directly turned off through the hardware watchdog protection unit independent of the main control MCU, and a qualitative safety status indicator is output.

[0008] Preferably, the processing link performs real-time fitting of the acquired voltage and current response data, including: The ohmic internal resistance parameter R0, the first-order polarization internal resistance parameter R1, and the polarization capacitance parameter C1 in the equivalent circuit model are iteratively updated using the recursive least squares method or the extended Kalman filter. The processing link also includes a parameter validity determination unit to check the rationality of the identification results. When the identified parameters exceed the preset physical acceptable range, an abnormal handling process is triggered and the valid parameter values ​​from the previous moment are maintained.

[0009] Preferably, the processing link includes the following when performing collaborative correction processing: Broadcast status query requests to neighboring nodes in the Mesh network and receive battery status parameters returned by neighboring nodes; The local estimation results are corrected based on the received neighbor parameters to eliminate systematic estimation bias caused by parameter silos. The processing link also supports a neighbor node priority sorting function, prioritizing the acquisition of status information of the main BMS node or the battery pack node with the largest capacity.

[0010] Preferably, the processing link further includes the following before acquiring battery response data: Before the BMS enters sleep mode, a snapshot of the updated equivalent circuit model parameters at the current moment is saved to a non-volatile storage area. Synchronously record the timestamp of the snapshot save time and the current battery temperature; After being woken up, the processing link evaluates the timeliness of the parameters based on the timestamp and temperature.

[0011] Preferably, the processing link further includes the following before activating the protection timing optimization unit: After the Mesh network is established, a network time-based clock synchronization protocol is executed to control the system clock deviation of each node within a preset threshold range. When a wake-up signal is detected, the entire network of nodes is triggered to wake up in a coordinated manner via Bluetooth Mesh broadcast, thereby reducing the difference in wake-up time among the nodes.

[0012] Preferably, the hardware watchdog protection unit has a built-in independent voltage monitoring channel and current sampling channel, which can detect the overvoltage, undervoltage, overcurrent and short circuit status of the battery in real time; The shutdown signal of the hardware watchdog protection unit directly acts on the gate control circuit of the power MOSFET.

[0013] Preferably, each BMS node further includes a sleep power management module, which gradually shuts down the clock of the main control MCU and reduces the power supply voltage after the processing link detects a static state with no load connection and no communication request.

[0014] Preferably, the sleep power management module is configured with a Bluetooth Mesh Low Power Broadcast unit; During sleep, the processing link periodically sends heartbeat signals through the low-power broadcast unit to inform neighboring nodes that the node is in sleep mode rather than offline or faulty, thus maintaining the integrity of the Mesh network topology.

[0015] Preferably, the system further includes a multi-node parameter self-calibration subsystem, a multi-node Bluetooth Mesh networking subsystem, and a wake-up synchronization and coordination subsystem; The multi-node parameter self-calibration subsystem is electrically connected to the battery pack, and collects the battery pack's terminal voltage, charging and discharging current and surface temperature data in real time. It also integrates an online parameter identification engine. The multi-node Bluetooth Mesh networking subsystem establishes a Bluetooth Mesh protocol stack on each BMS node to realize the wireless self-organizing network function between nodes. After detecting a wake-up signal, the wake-up synchronization and coordination subsystem sends a synchronization wake-up command to all nodes in the network via Mesh broadcast. Each node calculates its own wake-up execution time based on the received broadcast time and the preset wake-up delay parameters.

[0016] Preferably, the system further includes a protection timing optimization subsystem, which operates independently of other subsystems and includes a hardware-level protection execution path; The hardware watchdog module in the protection timing optimization subsystem starts monitoring the battery status before the main control MCU is powered on. When the battery voltage is lower than the over-discharge protection threshold and an external load connection is detected, the power MOSFET is immediately turned off. The protection timing optimization subsystem also includes a wake-up timing control unit, which arranges the timing from the wake-up signal trigger to the protection function readiness and then to the completion of the main control MCU initialization, ensuring that the hardware protection path is in a monitoring state before the main control MCU completes initialization.

[0017] Compared with the prior art, the present invention has the following beneficial effects: 1. This invention introduces a parameter self-calibration module and uses an online parameter identification algorithm to update the equivalent circuit model parameters in real time, so that the parameters are dynamically consistent with the current actual state of the battery. This can solve the contradiction between the static nature of parameter estimation methods in existing BMS and the time-varying nature of the actual battery state, eliminate the systematic deviation generated when estimating SOC based on fixed parameters, improve the accuracy of SOC estimation, and effectively avoid the safety hazards caused by time-varying parameter mismatch and parameter cliff problems during wake-up.

[0018] 2. This invention achieves parameter collaborative sharing during wake-up initialization in a multi-node Mesh network through a neighbor node status query mechanism in the wake-up synchronization module. The first node to be woken up can obtain the actual status information of its neighbor nodes in a timely manner and correct its local estimation results accordingly, thereby eliminating the systematic estimation error caused by parameter silos. This can solve the risk of SOC estimation result jumps and improper protection threshold settings caused by the asynchronous wake-up of multiple nodes.

[0019] 3. This invention realizes the function of multi-node wireless self-organizing network through Bluetooth Mesh networking module, which breaks through the limitations of traditional wired bus in terms of wiring complexity and cost. It supports flexible deployment in large-scale application scenarios such as parallel connection of multiple battery clusters in energy storage power stations and parallel power supply of multiple packages in electric forklifts. At the same time, it maintains the integrity of Mesh network topology during sleep through low power broadcast heartbeat mechanism, and realizes multi-node collaborative management while ensuring microampere-level sleep power consumption.

[0020] 4. This invention, through the hardware watchdog protection unit in the protection timing optimization module, moves the activation time of the protection function to the instant of the wake-up signal trigger, and operates independently of the main control MCU. This increases the protection vacuum period within the inherent wake-up initialization timing. When the battery is in a deep over-discharge state and a load attempts to connect, it can quickly respond to the abnormality and turn off the power MOSFET, effectively preventing hardware damage caused by instantaneous high current. This can solve the protection vacuum problem during wake-up. Attached Figure Description

[0021] The invention will now be further described with reference to the accompanying drawings.

[0022] Figure 1 This is a structural block diagram of an intelligent management system for a lithium battery pack Bluetooth smart BMS protection board according to the present invention; Figure 2This is a schematic diagram of the structural framework of the multi-node Bluetooth Mesh networking subsystem in this invention. Detailed Implementation

[0023] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0024] Example 1: This embodiment is specifically applied to a scenario where multiple lithium battery packs are connected in parallel for power supply to an electric forklift. In this scenario, multiple lithium battery packs need to work together to provide power support for the forklift drive system. During operation, the electric forklift may face various working conditions such as emergency stop, start, and fork lifting. There are collaborative management requirements such as power distribution, load balancing, and fault isolation among multiple lithium battery packs. The intelligent management system of the lithium battery pack Bluetooth smart BMS protection board described in this embodiment can achieve accurate battery status management, reliable wake-up protection, and flexible multi-node collaborative control in the above-mentioned large-scale multi-node application scenario.

[0025] like Figure 1 and Figure 2 As shown, this system includes a processing link, a parameter self-calibration module, a Bluetooth Mesh networking module, a wake-up synchronization module, a protection timing optimization module, a sleep power consumption management module, a hardware watchdog protection unit, a data acquisition unit, a storage unit, and a power control unit. The processing link, acting as the system's control unit, is configured to coordinate the timing of operations across modules and execute battery state estimation and protection decision-making logic. The processing link is integrated into the main control MCU, which establishes communication connections with each functional module via the system bus.

[0026] The data acquisition unit is located at the front end of the system and is used to collect the operating parameters of the battery pack in real time. The data acquisition unit includes a voltage divider sampling circuit and a current sampling resistor arranged in series. The input terminal of the voltage divider sampling circuit is electrically connected to the positive output terminal of the battery pack and is used to attenuate the high voltage signal of the battery pack to the safe voltage range that the analog-to-digital converter of the main control MCU can accept according to a preset ratio. The current sampling resistor is set in the charging and discharging circuit of the battery pack. The two ends of the current sampling resistor are electrically connected to the input terminal of the differential amplifier circuit, and the output terminal of the differential amplifier circuit is electrically connected to the input terminal of the analog-to-digital converter of the main control MCU. It is used to amplify the weak voltage signal across the sampling resistor to a suitable signal amplitude. The temperature sampling channel in the data acquisition unit acquires the temperature signal of the battery surface through a thermistor or a digital temperature sensor, and the output pin of the temperature sensor is electrically connected to the general-purpose input port of the main control MCU. The data acquisition unit collects the battery's terminal voltage, charging and discharging current, and surface temperature in one complete transaction. The three types of data obtained are then transmitted to the processing link via the internal bus for further processing.

[0027] The parameter self-calibration module is configured to dynamically update the internal parameters of the equivalent circuit model based on the online parameter identification algorithm during normal battery operation. The equivalent circuit model adopts a common second-order RC equivalent circuit structure, which includes the ohmic internal resistance parameter R0, the first-order polarization internal resistance parameter R1, the polarization capacitance parameter C1, the second-order polarization internal resistance parameter R2, and the second-order polarization capacitance parameter C2. The ohmic internal resistance R0 characterizes the sum of the electrolyte resistance, separator resistance, and current collector contact resistance inside the battery. This parameter mainly affects the transient voltage response characteristics of the battery. The first-order polarization internal resistance R1 and polarization capacitance C1 together constitute a first-order RC network, which is used to describe the electrochemical polarization effect inside the battery. The product of the two, R1×C1, defines the time constant of the first-order polarization process. The second-order polarization internal resistance R2 and polarization capacitance C2 together form a second-order RC network, which is used to describe the concentration polarization effect inside the battery. After acquiring real-time voltage, current, and temperature data, the parameter self-calibration module performs parameter identification calculations based on the preset equivalent circuit model equations, and updates the values ​​of model parameters such as R0, R1, C1, R2, and C2 in real time.

[0028] The online parameter identification algorithm in the parameter self-calibration module is implemented using recursive least squares or extended Kalman filter. Recursive least squares is based on the least squares principle and corrects the estimated values ​​of model parameters by continuously introducing new observation data. Its recursive formula includes parameter update term and covariance matrix update term. The specific calculation process includes: calculating the innovation vector based on the input and output data at the current time, performing matrix multiplication of the innovation vector and the covariance matrix to obtain the parameter correction amount, superimposing the parameter correction amount on the parameter estimate at the previous time to obtain the parameter estimate at the current time, and updating the covariance matrix to reflect the degree of uncertainty of parameter estimation. The extended Kalman filter introduces nonlinear processing capabilities on the basis of the standard Kalman filter. By performing a first-order Taylor expansion approximation on the nonlinear system equations, it linearizes the nonlinear dynamic characteristics of the battery at the current state estimate, thereby achieving a recursive optimal estimate of the battery state. In the state vector of the extended Kalman filter, in addition to the key state variable SOC, it also contains internal parameters R0, R1, C1, etc. of the equivalent circuit model. The state vector is predicted and updated over time through the state transition equation, and is compared and corrected with the measured voltage data through the observation equation.

[0029] The parameter self-calibration module also includes a parameter validity determination unit, which is used to check the rationality of the identification results. The parameter validity determination unit presets the upper and lower limits of the physical acceptable range of various parameters. When the identified parameter value exceeds the preset range, the determination unit outputs an abnormality flag and triggers the abnormality handling process. In the exception handling process, the module suspends the use of the current identification result and instead uses the valid parameter values ​​from the previous moment to participate in the subsequent SOC estimation calculation. This avoids parameter mutation problems caused by the divergence of the identification algorithm or noise interference. The judgment logic of the parameter validity determination unit also includes a limit on the parameter change rate. When the absolute value of the difference between the identification parameter at the current moment and the value at the previous moment exceeds the preset threshold, it is also determined to be an abnormal state and the parameter freezing operation is performed.

[0030] The parameter self-calibration module also includes a parameter snapshot saving unit before hibernation. This unit is activated after the BMS detects a static state with no load connection and no communication request, and before the main control MCU enters hibernation mode. The snapshot saving unit fully reads the updated values ​​of the equivalent circuit model parameters R0, R1, C1, R2, and C2 at the current moment and writes them to the non-volatile storage area of ​​the storage unit. The non-volatile storage area is implemented using ferroelectric memory or electrically erasable programmable read-only memory, which can retain data for a long time without loss after the system is powered off. The snapshot saving unit also synchronously records the timestamp information of the snapshot saving time. The timestamp information comes from the output of the real-time clock module, which is driven by an independent low-frequency crystal oscillator and can continue to run even when the main control MCU is in sleep mode to maintain the time reference. The snapshot saving unit also records the current surface temperature value of the battery. This temperature value comes from the temperature sampling data collected by the data acquisition unit at the snapshot saving time. The snapshot saving unit writes the timestamp and temperature data together with the model parameters into the non-volatile storage area, providing a data basis for parameter timeliness evaluation after wake-up.

[0031] The Bluetooth Mesh networking module is configured to support multiple BMS nodes to achieve self-organization and interconnection through the Bluetooth Low Energy wireless protocol. Based on the Bluetooth 5.0 protocol stack, the Bluetooth Mesh networking module implements the roles of proxy node, relay node, low-power node and friend node required by the Mesh network. The Bluetooth Mesh networking module on each BMS node is configured by default as a combination of relay node role and low-power node role. It can both forward data packets from other nodes as a relay to extend the network coverage and maintain extremely low current consumption during sleep by taking advantage of the low power characteristics. The RF front-end of the Bluetooth Mesh networking module adopts a single-chip solution including a power amplifier and a low-noise amplifier. The transmit power is set to 4 dBmW and the receive sensitivity is set to -90 dBmW. It can achieve a reliable communication distance of more than 50 meters in open spaces. In industrial warehousing environments, due to the multipath effect and the obstruction of metal shelves, the typical communication distance is 15 to 30 meters.

[0032] The Bluetooth Mesh networking module also includes a neighbor node status query unit. This unit immediately performs the operation of obtaining neighbor node status information after the node is woken up. The neighbor node status query unit first reads the locally stored Mesh network topology information, which records the Bluetooth device addresses and signal strength sorting list of the known neighbor nodes of this node. The query unit sends status query request messages to each neighboring node sequentially based on the topology information. The message content includes the current status identifier of the node and the request type field. After receiving the status query request, the neighboring node encapsulates its own status data such as battery voltage, current, SOC, temperature, and model parameters into a response message and sends it back to the node that initiated the query. The first-wake-up node, after receiving the response message from the neighboring node, parses the message content and updates its local neighbor parameter cache. At the same time, it triggers the parameter island elimination algorithm to perform collaborative correction processing on the local estimation results.

[0033] The Bluetooth Mesh networking module also supports neighbor node priority sorting, with priority sorting rules determined according to a preset strategy. Option 1 is to prioritize nodes based on their roles. When configuring the system network, a master BMS node is specified. The master node has a higher priority than other slave nodes. Prioritizing the acquisition of the master node's status information can provide a benchmark reference for the status assessment of the entire system. Option 2 is based on capacity priority ranking. The battery pack node with the largest capacity undertakes a larger proportion of the load distribution in the multi-pack parallel system, and its state changes have a more significant impact on the overall system. Therefore, prioritizing the acquisition of the state information of large-capacity nodes can more accurately assess the system state. Option 3 is to prioritize based on signal strength. In a mesh network, the neighboring node with the strongest signal strength usually has a more stable communication link, and the data obtained from that node has higher reliability.

[0034] The wake-up synchronization module is configured to execute a multi-node clock synchronization protocol in a multi-node Mesh network to ensure that the wake-up time of each node is consistent. The wake-up synchronization module includes a multi-node clock synchronization unit, which automatically executes a network time-based clock synchronization protocol after the Mesh network is established. The clock synchronization protocol adopts a master-slave synchronization architecture. The master BMS node in the system acts as the clock source node. The slave nodes periodically receive synchronization messages from the clock source node and correct their local clocks accordingly. The master node broadcasts a time synchronization message to the Mesh network every second. The message content includes the master node's local timestamp and sequence number. After receiving a synchronization message, the slave node records the message's reception time and calculates the deviation between its local clock and the master node's clock based on the message's sending and receiving times and a pre-determined communication delay estimate. The slave node then performs linear correction on its local clock based on the calculated deviation, keeping the system clock deviation of each node within a preset threshold range of less than 10 milliseconds.

[0035] The wake-up time scheduling function of the wake-up synchronization module is activated when a wake-up signal is detected. The wake-up signal may come from any one of the following: an external load connection event, an internal timed wake-up event, or a remote communication wake-up event. When the wake-up synchronization module detects a wake-up signal, it first determines the current node's role in the Mesh network. If the current node is designated as the wake-up coordinating node, the wake-up synchronization module sends a synchronization wake-up command message to all nodes in the network via Bluetooth Mesh broadcast. The message contains the wake-up coordinating node's local trigger time and a preset network-wide synchronization wake-up delay parameter. If the current node is a regular slave node, the wake-up synchronization module only receives the synchronization wake-up command message from the wake-up coordinating node and calculates its own wake-up execution time based on the trigger time and delay parameter in the message. In this way, all nodes in the network can coordinate and execute wake-up actions under a unified time reference, reducing the difference in wake-up times among nodes to an acceptable range.

[0036] During the node wake-up initialization period, the wake-up synchronization module continuously listens to the parameter broadcast messages of neighboring nodes. After completing their own parameter updates, neighboring nodes will encapsulate the updated model parameters into Mesh network broadcast messages for distribution across the entire network. After receiving the parameter broadcast messages from neighboring nodes, the first wake-up node parses the message content and updates its local neighbor parameter cache in real time. The local cache records the parameter values ​​and timestamps of the most recent broadcast by each neighboring node, providing data input for the subsequent parameter island elimination algorithm. The wake-up synchronization module also includes a parameter island elimination algorithm, which is executed immediately after obtaining the neighbor parameters. Its logic is as follows: calculate the cooperative correction amount based on the degree of deviation between the local parameters and the neighbor parameters, and add the cooperative correction amount to the local SOC estimation result, so that the local parameter estimation converges to the actual parameter state of the neighbor nodes. The calculation of the cooperative correction amount takes into account the confidence weight of the local parameters and the validity label of the neighbor parameters, ensuring that the correction process will not introduce additional estimation errors due to the abnormal parameters of the neighbor nodes.

[0037] The protection timing optimization module is configured to activate the hardware-level protection function at the initial stage of wake-up, and to move the activation time of the protection function forward to the moment of wake-up signal trigger. The protection timing optimization module includes two components: wake-up timing control unit and hardware watchdog protection unit. The wake-up timing control unit is configured to precisely arrange the timing sequence from wake-up signal trigger to protection function readiness and then to the completion of main control MCU initialization. The timing control unit contains a finite state machine, which includes five states: power-off state, wake-up detection state, hardware protection activation state, main control MCU initialization state, and normal operation state. When the system is in a power-off state, the power supply to all functional modules is cut off, and the hardware watchdog protection unit is in a low-power monitoring mode. Once the wake-up timing control unit detects that the wake-up signal is valid, the state machine immediately transitions to the wake-up detection state. At this time, the wake-up timing control unit activates the power supply to the hardware watchdog protection unit and sends a power-on request signal to the power management module of the main control MCU.

[0038] The hardware watchdog protection unit operates independently of the main control MCU and includes an independent voltage monitoring channel, current sampling channel, and protection logic execution circuit. The voltage monitoring channel includes a voltage divider resistor network and a comparator circuit. The voltage divider resistor network attenuates the battery voltage by a fixed ratio and then inputs it to the non-inverting input of the comparator. The inverting input of the comparator is connected to the preset overvoltage protection threshold voltage and undervoltage protection threshold voltage. When the battery voltage exceeds the overvoltage threshold or falls below the undervoltage threshold, the comparator output level flips to generate an alarm signal. The current sampling channel includes a current sampling resistor and a differential amplifier circuit. The output signal of the differential amplifier is converted from analog to digital and then input to the overcurrent detection module and short-circuit detection module in the protection logic execution circuit. The overcurrent detection module compares the measured current value with the preset overcurrent protection threshold in real time. When the measured current exceeds the threshold and the duration exceeds the preset debouncing delay, it is determined to be an overcurrent fault and triggers the protection action. The short-circuit detection module uses a high-speed comparator. When it detects that a sudden large current caused by a load connection exceeds the preset short-circuit current threshold and the duration exceeds the microsecond-level debouncing delay, it immediately determines that a short-circuit fault and triggers an emergency shutdown action.

[0039] The sampling frequency of the hardware watchdog protection unit is set to no less than 10 kHz, that is, the battery voltage and current are fully sampled and judged every 0.1 milliseconds. The high sampling frequency ensures that the system can detect abnormal states and execute protection actions in a very short time. When the battery voltage is detected to be lower than the over-discharge protection threshold and an external load connection is detected, the protection logic execution circuit of the hardware watchdog protection unit immediately outputs a shutdown signal to the power control unit. The shutdown signal acts directly on the gate control circuit of the power MOSFET through the hardware connection. After receiving the shutdown signal, the gate drive circuit quickly pulls the gate voltage of the MOSFET down to below its threshold voltage, so that the MOSFET enters the shutdown state and cuts off the connection between the battery and the external load. The shutdown response time of the hardware watchdog protection unit is no more than 1 millisecond. The entire process from detecting an abnormal state to the complete shutdown of the MOSFET is completed in the hardware circuit and does not depend on any instruction response from the main control MCU processor.

[0040] The power control unit includes two sets of power MOSFETs connected in series in the positive and negative output circuits of the battery, referred to as the high-side MOSFET and the low-side MOSFET, respectively. The drain of the high-side MOSFET is electrically connected to the positive output terminal of the battery pack, and the source is electrically connected to the positive connection terminal of the load. The source of the low-side MOSFET is electrically connected to the negative output terminal of the battery pack, and the drain is electrically connected to the negative connection terminal of the load. Under normal operating conditions, both the high-side MOSFET and the low-side MOSFET are in the on state, and the output current of the battery pack flows to the external load through these two sets of MOSFETs. After the protection action is triggered, the shutdown signal output by the hardware watchdog protection unit simultaneously acts on the gate control circuits of the high-side MOSFET and the low-side MOSFET, causing both MOSFETs to switch to the shutdown state within 1 millisecond, cutting off the electrical connection between the battery and the load.

[0041] The sleep power management module is activated after detecting a static state with no load connection and no communication request. It gradually shuts down unnecessary functions in the system to reduce static power consumption. The sleep power management module first sends a sleep request command to the processing link. After completing the current data acquisition task and parameter saving task, the processing link enters the sleep preparation state. The sleep preparation state includes writing the currently valid model parameters into the storage unit, disabling the data transmission and reception function of the Bluetooth Mesh networking module, and configuring the main control MCU to enter a low-power sleep mode. After the main control MCU enters sleep mode, its core clock and system bus clock are cut off by the gating logic, and all peripheral modules except the real-time clock and wake-up circuit are powered off. The sleep power management module also includes a Bluetooth Mesh low power broadcast unit, which periodically sends a low power broadcast heartbeat signal during the main control MCU sleep period. The heartbeat broadcast sending interval is set to 5 seconds. The broadcast message content includes the device identifier of this node, the current working status identifier is sleep state, and the estimated remaining power of this node. After receiving the heartbeat broadcast, the neighboring node can determine that the node is in a normal sleep state rather than an offline fault state, thereby maintaining the integrity of the Mesh network topology. The average current consumption of the Bluetooth Mesh Low Power Broadcast Unit during sleep does not exceed 50 microamps. Including the static current of the real-time clock and wake-up circuit, the total standby current of the entire BMS node in sleep state does not exceed 100 microamps.

[0042] The storage unit is configured to permanently store the system's key configuration parameters and operating data. The storage unit consists of two parts: a volatile storage area and a non-volatile storage area. The volatile storage area is implemented using static random access memory and is used to temporarily store real-time acquired voltage, current, and temperature data, as well as calculated SOC values, model parameter values, and neighbor node parameter buffer data during system operation. The non-volatile storage area is implemented using ferroelectric memory and is used to permanently store system configuration parameters, network configuration information, model parameter snapshots, and operation log data. System configuration parameters include the set values ​​of each protection threshold, sampling frequency configuration, communication parameter configuration, and master and slave node role configuration, etc. Network configuration information includes the device identifier of the Mesh network, application key, and known neighbor node list. Model parameter snapshots include the R0, R1, C1, R2, and C2 values ​​saved before the most recent hibernation, as well as the corresponding timestamps and temperature values.

[0043] The intelligent management system for the lithium battery pack Bluetooth smart BMS protection board described in this embodiment includes the following main stages: system startup and initialization stage, parameter self-calibration stage, pre-sleep preparation stage, sleep stage, wake-up detection and protection activation stage, multi-node clock synchronization stage, neighbor parameter acquisition and parameter island elimination stage, and normal operation and collaborative management stage. The specific workflow of each stage and the collaborative operation process of each module are described in detail below.

[0044] During the system startup and initialization phase, when the system is powered on for the first time or resumes from deep sleep, the main control MCU first completes the basic initialization configuration of the internal registers and peripheral modules. The clock configuration module of the main control MCU switches the system clock to the preset operating frequency and configures the input / output mode and multiplexing function of each GPIO port. The initialization program further configures the sampling parameters of the data acquisition unit, including setting the conversion clock division coefficient of the ADC, enabling the sampling channel and arranging the sampling sequence. The initialization program also includes the protocol stack parameters of the Bluetooth Mesh networking module, including setting the broadcast name of the device, the device category identifier and the application layer key. The final step of the initialization process is to read system configuration parameters and model parameter snapshots from the non-volatile storage area of ​​the storage unit. The timeliness of the snapshots is determined based on the timestamps recorded in the snapshots. If the timestamps show that the freshness of the snapshot data is within an acceptable range, the parameter values ​​in the snapshots are restored to the model parameter cache area of ​​the volatile storage area as the initial values ​​for SOC estimation and parameter updates.

[0045] During the parameter self-calibration stage, the processing link continuously executes the online parameter identification algorithm to keep the model parameters dynamically consistent with the actual state of the battery. The data acquisition unit collects the battery's terminal voltage, charging and discharging current and surface temperature in real time at a sampling frequency of 100 Hz. The three types of data collected are transferred to the memory buffer of the main control MCU via DMA. The processing link reads the latest acquired data points from the memory buffer and calculates the predicted model terminal voltage at the current moment based on the mathematical equations of the equivalent circuit model. The predicted voltage at the model terminals is calculated based on the previous state of charge (SOC) estimate, the current measured current, and the internal parameters R0, R1, C1, R2, and C2 of the equivalent circuit model. The predicted voltage at the model terminals is then compared with the measured voltage to calculate the voltage error between them. The processing link inputs the voltage error into the recursive least squares algorithm or the extended Kalman filter algorithm to calculate the corrected model parameter update. The correction is then added to the parameter estimate at the previous time step to obtain the parameter estimate at the current time step. The processing link writes the updated parameter values ​​to the model parameter cache in the volatile storage area, and refreshes the current parameter record in the storage unit. The parameter validity determination unit performs a reasonableness check on the updated parameter values. When it detects that the parameter value exceeds the preset physical acceptable range or the parameter change rate exceeds the threshold, it triggers the exception handling process and maintains the use of the valid parameter values ​​from the previous moment.

[0046] During the pre-sleep preparation phase, when the system detects that the idle state without load connection and without communication request has lasted for more than the preset judgment time, the sleep power management module is activated to start the sleep process. The sleep power management module first sends a sleep request instruction to the processing link. After receiving the sleep request, the processing link stops receiving new data acquisition samples and completes the processing of the current batch of data and parameter updates. The processing link sends the current model parameter cache contents to the snapshot saving unit, and the snapshot saving unit writes the values ​​of R0, R1, C1, R2, and C2 completely to the non-volatile storage area of ​​the storage unit. The snapshot saving unit synchronously reads the current timestamp of the real-time clock module and writes the timestamp value and the current battery temperature value together into the storage address unit adjacent to the model parameters in the non-volatile storage area; The processing link disables the data transmission and reception function of the Bluetooth Mesh networking module, but retains the configuration state of the low-power broadcast unit to support heartbeat broadcasting during sleep. The processing link sends a command to the power management module of the main control MCU to enter deep sleep mode. The core clock of the main control MCU is cut off by gating, and only the real-time clock module and the wake-up circuit are powered.

[0047] During the sleep phase, the system is in a very low power quiescent state. The sleep power management module controls the Bluetooth Mesh Low Power Broadcast Unit to periodically send heartbeat broadcast signals. The heartbeat broadcast sending interval is set to 5 seconds. Each broadcast includes the device identifier of this node, the current working status identifier, and the estimated remaining power. The heartbeat broadcast transmission power is set to 0 dB / mW to reduce power consumption, and the effective payload length of the broadcast message is controlled within 31 bytes to shorten the transmission time. After receiving the heartbeat broadcast, the neighboring node updates its local network topology table and marks the node as a sleep state rather than an offline fault state. The real-time clock module continues to run during the sleep period to maintain the time reference, and the wake-up circuit continuously monitors external wake-up signals, including load connection detection signals, timed wake-up signals, and Bluetooth remote wake-up signals.

[0048] During the wake-up detection and protection activation phase, when the wake-up circuit detects that any wake-up signal is valid, the system immediately switches from the sleep state to the wake-up process, and the state machine of the wake-up timing control unit switches from the power-off state to the wake-up detection state. At this time, the wake-up timing control unit immediately activates the power supply of the hardware watchdog protection unit, causing it to leave the low-power monitoring mode and enter the normal working state. After activation, the hardware watchdog protection unit immediately begins to perform high-frequency sampling and anomaly detection of the battery voltage and current. The voltage monitoring channel compares the measured voltage value with the over-discharge protection threshold. When the battery voltage is detected to be lower than the over-discharge protection threshold, the battery is determined to be in a deep over-discharge state. The current sampling channel monitors the connection status of external loads in real time. When current inflow caused by load connection is detected, a discharge risk is determined. When the protection logic execution circuit of the hardware watchdog protection unit simultaneously meets the two conditions of deep over-discharge determination and load connection determination, it immediately outputs a shutdown signal to the gate drive circuit of the power control unit. The shutdown signal is directly applied to the power MOSFET through hardware wiring, turning off both MOSFETs within 1 millisecond and cutting off the electrical connection between the battery and the external load. The above protection actions of the hardware watchdog protection unit are implemented entirely by hardware circuitry, without relying on any instruction response from the main control MCU. Even if the main control MCU has not yet completed initialization, it can still perform reliable abnormal protection.

[0049] During the multi-node clock synchronization phase, after the main control MCU completes initialization, it starts the network discovery and neighbor node scanning functions of the Bluetooth Mesh networking module. After scanning the broadcast signals of neighbor nodes, the Bluetooth Mesh networking module parses the broadcast messages and establishes a list of neighbor nodes. The master node, acting as the clock source node, periodically broadcasts time synchronization messages to the Mesh network. The time synchronization message contains the master node's local timestamp and message sequence number. After receiving the time synchronization message, the slave node records the reception time and calculates the deviation between its local clock and the master node's clock based on the sending time, receiving time, and estimated communication delay. The nodes perform linear correction to their local clocks based on the deviation value, controlling the system clock deviation of each node to within 10 milliseconds. After clock synchronization is complete, the wake-up time scheduling function of the wake-up synchronization module is activated; When a wake-up signal is detected, the wake-up coordinating node sends a synchronous wake-up command message to all nodes in the network via Bluetooth Mesh broadcast. The message contains the local trigger time of the wake-up coordinating node and the preset collaborative wake-up delay parameters. All nodes in the network calculate their own wake-up execution time based on the received broadcast messages, ensuring that the wake-up actions of each node are coordinated and executed under a unified time reference.

[0050] During the neighbor parameter acquisition and parameter island elimination phase, the first wake-up node broadcasts a status query request to the neighbor nodes in the Mesh network through the neighbor node status query unit of the Bluetooth Mesh networking module. The status query request message includes the device identifier of this node, the request type field and the query parameter list. After receiving a status query request, a neighboring node encapsulates its own battery voltage, charging and discharging current, estimated SOC, surface temperature, and equivalent circuit model parameters R0, R1, and C1 into a status response message and sends it back to the node that initiated the query. After receiving the response message from the neighboring node, the first wake-up node parses the message content and updates its local neighbor parameter cache. At the same time, it outputs the latest data in the neighbor parameter cache to the parameter island elimination algorithm unit of the wake-up synchronization module. The parameter island elimination algorithm unit calculates the collaborative correction amount based on the degree of deviation between local parameters and neighbor parameters. The calculation formula of the collaborative correction amount includes a deviation calculation term and a confidence weighting term. Specifically, it calculates the deviation between the local SOC estimate and the neighbor node's SOC estimate, then calculates the confidence weight of the neighbor node based on the signal strength and parameter freshness of the neighbor node, and finally multiplies the deviation amount by the confidence weight to obtain the collaborative correction amount. The collaborative correction amount is superimposed on the local SOC estimation result, so that the local parameter estimation converges to the actual parameter state of the neighboring nodes. The neighboring node priority sorting function sorts the neighboring nodes according to the preset role priority, capacity priority or signal strength priority rules, and the first node to be woken up obtains the status information of the neighboring node with the highest priority.

[0051] During the normal operation and collaborative management phase, the system enters a normal working state, continuously executing functions such as parameter self-calibration, SOC estimation, and protection monitoring. The data acquisition unit continuously collects battery terminal voltage, charging and discharging current, and surface temperature data, providing real-time data input to the parameter self-calibration module. The parameter self-calibration module continuously updates the internal parameters of the equivalent circuit model based on an online parameter identification algorithm, ensuring that the model parameters remain dynamically consistent with the current actual state of the battery. The processing link executes the SOC estimation algorithm based on the updated model parameters and real-time collected data to calculate the current state of charge estimate of the battery. The SOC estimate is combined with the battery's rated capacity and current current value to calculate the remaining usable time estimate and driving range estimate of the battery. Under normal operating conditions, the Bluetooth Mesh networking module continuously monitors the parameter broadcast messages of neighboring nodes and updates the parameter data of neighboring nodes to the local cache. The protection timing optimization module continuously monitors the battery's operating status. When abnormal states such as overvoltage, undervoltage, overcurrent, or short circuit are detected, the hardware watchdog protection unit immediately executes a shutdown protection action. The sleep power management module continuously monitors the load connection status and communication request status. When it detects that the idle state has been in place for a longer than the judgment time, it triggers a new round of sleep process.

[0052] Example 2: In terms of system architecture construction, each battery cluster in this embodiment is equipped with an independent BMS node. Each BMS node is wirelessly interconnected through a Bluetooth Mesh network. The hardware configuration of each BMS node has been upgraded based on the first embodiment to adapt to the needs of large-scale networking. The main control MCU adopts a 32-bit microcontroller with a main frequency of 120 MHz, which has a larger internal storage space and stronger transaction processing capabilities, and can support more complex parameter identification algorithms and multi-node data fusion calculations. The storage unit capacity is expanded to 8 megabytes, with the volatile storage area implemented using 2 megabytes of static random access memory to cache high-frequency sampling data and intermediate calculation results; the non-volatile storage area is implemented using 4 megabytes of serial peripheral interface flash memory to store system configuration, operation logs and historical data. The RF front-end of the Bluetooth Mesh networking module adopts an external power amplifier solution, increasing the transmit power to 20 dBmW, which can achieve a reliable communication distance of more than 100 meters in the container array environment of the energy storage power station.

[0053] Regarding the extended configuration of the parameter self-calibration module, this embodiment adds a cluster-level parameter coordination function for the characteristics of large-scale energy storage systems. After completing local parameter self-calibration, each BMS node encapsulates the updated model parameters into a Mesh network broadcast message for distribution across the entire network. After collecting parameter data from all cluster nodes, the central management node executes a cluster-level parameter coordination algorithm. Based on the weighted average principle, the cluster-level parameter coordination algorithm sums the model parameters of each cluster node according to the proportion of its current capacity to obtain the equivalent model parameters of the entire system. After the equivalent model parameters are sent back to each cluster node, each cluster node performs a secondary correction on its local SOC estimation results based on the equivalent model parameters to eliminate the system-level SOC estimation bias caused by the inconsistent aging degree of each cluster's batteries.

[0054] Regarding the extended configuration of the wake-up synchronization module, this embodiment adds a hierarchical synchronization mechanism for large-scale node scenarios. This mechanism divides the entire Mesh network into multiple synchronization domains, each containing several geographically adjacent BMS nodes. Within each domain, a node is selected as the domain clock source, responsible for providing time synchronization services to other nodes within the domain. Time information is transmitted between multiple synchronization domains through inter-domain relay nodes, ultimately achieving clock synchronization across the entire network. Compared to a single-layer synchronization mechanism, the hierarchical synchronization mechanism reduces the propagation delay and collision probability of time synchronization messages in large networks, improving the accuracy and reliability of network-wide clock synchronization.

[0055] Regarding the extended configuration of the protection timing optimization module, this embodiment adds multi-level protection linkage function to meet the safety level requirements of energy storage power stations; The first level of protection is executed independently by the hardware watchdog protection unit of each cluster, achieving millisecond-level single-cluster protection response; The second level of protection is uniformly coordinated and executed by the central management node. When conditions involving multi-cluster linkage protection are detected, the central management node sends linkage protection commands to all relevant nodes through the Mesh network, achieving a system-level protection response within seconds. The multi-level protection linkage function ensures that the protection system can make timely and reliable action responses in both single-unit failure and system-level failure scenarios.

[0056] In terms of multi-node collaborative management, this embodiment implements inter-cluster power balancing and fault isolation functions. The inter-cluster power balancing function calculates the target output power allocation ratio of each cluster based on the SOC, terminal voltage, and temperature status of each cluster's batteries, and adjusts the power converter settings of each cluster through communication commands. The fault isolation function immediately isolates the cluster from the parallel system when a serious fault is detected in a battery cluster, and broadcasts the fault information to other cluster nodes to trigger corresponding protection strategy adjustments.

[0057] Example 3: In terms of system architecture, each electric forklift battery pack in this embodiment is equipped with an independent BMS node. Each BMS node is wirelessly interconnected through a Bluetooth Mesh network. Considering the vibration and impact characteristics of the electric forklift's working environment, the hardware design of each BMS node adopts reinforcement measures, including using a reinforced connector to replace the standard pin header connection, increasing the number of module fixing screws, and adding thermally conductive silicone filling to the bottom of key chips. The power control unit in this embodiment adopts a dual MOSFET series scheme. Each MOSFET has a rated current of 200 amps. The two MOSFETs connected in parallel can withstand a continuous discharge current of 400 amps and an instantaneous peak current of 800 amps, which meets the high current requirements when the forklift starts.

[0058] Regarding the optimized configuration of the parameter self-calibration module, this embodiment adds a dynamic operating condition recognition function for the high-rate discharge characteristics of forklift batteries. The dynamic operating condition identification function determines the type of operating condition of the battery based on the changing trends of the measured current and voltage values ​​at the current moment. The operating conditions include four types: stationary condition, low-current condition, high-current condition, and pulse condition. In the stationary condition, the absolute current value is less than 1 ampere. At this time, accurate measurement of the open-circuit voltage is primarily performed to calibrate the SOC reference. In the low-current condition, the absolute current value is between 1 and 20 amperes. In this case, the standard sampling frequency and parameter update rate are used. In the high-current condition, the absolute current value is between 20 and 200 amperes. At this time, the sampling frequency is increased to 500 Hz and the parameter update rate is accelerated to track the battery's rapid dynamic response. In the pulse condition, the current suddenly exceeds 200 amperes. This triggers the instantaneous parameter identification mode, using a sliding window algorithm to extract features from the voltage response before and after the pulse to identify the battery's high-rate internal resistance characteristics.

[0059] Regarding the optimized configuration of the wake-up synchronization module, this embodiment adds a fast wake-up channel function to meet the real-time requirements of forklift applications. After detecting the key switch signal or the charging gun insertion signal, the fast wake-up channel immediately sends an emergency wake-up command to all relevant nodes through the Bluetooth Mesh network. The priority of the emergency wake-up command is higher than that of the normal wake-up command. After receiving the emergency wake-up command, each node immediately exits the sleep mode and enters the wake-up process. The fast wake-up channel controls the response time of the system from a completely sleep state to a protection function ready state to within 50 milliseconds, which meets the response speed requirements of forklift startup.

[0060] Regarding the optimized configuration of the protection timing optimization module, this embodiment adds a pre-charge protection function to meet the safety requirements of forklift applications. Before the main contactor closes, the pre-charge protection function closes the pre-charge relay and pre-charge resistor to form a pre-charge circuit. The current-limiting resistor of the pre-charge circuit limits the pre-charge current within a safe range, so that the large capacitor on the load side is gradually charged to close to the battery voltage under controlled conditions. When the difference between the load-side voltage and the battery voltage is less than a preset threshold, it indicates that pre-charging is complete. At this point, the main contactor closes and the pre-charging relay opens. The pre-charging protection function effectively prevents instantaneous high-current surges caused by the instantaneous charging of a large capacitor, extending the service life of the power MOSFET and the main contactor.

[0061] The foregoing has shown and described the basic principles, main features, and advantages of the present invention. Those skilled in the art should understand that the present invention is not limited to the above embodiments. The embodiments and descriptions in the specification are merely illustrative of the principles of the invention. Various changes and modifications can be made to the invention without departing from its spirit and scope, and all such changes and modifications fall within the scope of the present invention as claimed. The scope of protection of the present invention is defined by the appended claims and their equivalents.

Claims

1. A smart management system for a lithium battery pack Bluetooth smart BMS protection board, characterized in that, The system includes a processing link, which includes: Acquire battery terminal voltage and charge / discharge current response data under current operating conditions, and trigger a wake-up signal when a load connection is detected; Based on the online parameter identification algorithm, the collected voltage and current response data are fitted in real time, and the ohmic internal resistance parameters and polarization parameters in the equivalent circuit model are dynamically updated to establish the mapping relationship between parameters and temperature, state of charge and real-time current value. Based on the mapping relationship, a collaborative correction process is applied to the local SOC estimation results during wake-up initialization, so that the local parameter estimation converges to the actual parameter state of the neighboring nodes. After completing parameter co-calibration, the protection timing optimization unit is activated. When the battery voltage is detected to be lower than the over-discharge protection threshold, the power MOSFET is directly turned off through the hardware watchdog protection unit independent of the main control MCU, and a qualitative safety status indicator is output.

2. The intelligent management system for a lithium battery pack Bluetooth smart BMS protection board according to claim 1, characterized in that, The processing link performs real-time fitting of the acquired voltage and current response data, including: The ohmic internal resistance parameter R0, the first-order polarization internal resistance parameter R1, and the polarization capacitance parameter C1 in the equivalent circuit model are iteratively updated using the recursive least squares method or the extended Kalman filter. The processing link also includes a parameter validity determination unit to check the rationality of the identification results. When the identified parameters exceed the preset physical acceptable range, an abnormal handling process is triggered and the valid parameter values ​​from the previous moment are maintained.

3. The intelligent management system for a lithium battery pack Bluetooth smart BMS protection board according to claim 1, characterized in that, The processing link includes the following when performing collaborative correction processing: Broadcast status query requests to neighboring nodes in the Mesh network and receive battery status parameters returned by neighboring nodes; The local estimation results are corrected based on the received neighbor parameters to eliminate systematic estimation bias caused by parameter silos. The processing link also supports a neighbor node priority sorting function, prioritizing the acquisition of status information of the main BMS node or the battery pack node with the largest capacity.

4. The intelligent management system for a lithium battery pack Bluetooth smart BMS protection board according to claim 1, characterized in that, The processing link also includes the following before acquiring battery response data: Before the BMS enters sleep mode, a snapshot of the updated equivalent circuit model parameters at the current moment is saved to a non-volatile storage area. Synchronously record the timestamp of the snapshot save time and the current battery temperature; After being woken up, the processing link evaluates the timeliness of the parameters based on the timestamp and temperature.

5. The intelligent management system for a lithium battery pack Bluetooth smart BMS protection board according to claim 1, characterized in that, The processing link further includes the following before activating the protection timing optimization unit: After the Mesh network is established, a network time-based clock synchronization protocol is executed to control the system clock deviation of each node within a preset threshold range. When a wake-up signal is detected, the entire network of nodes is triggered to wake up in a coordinated manner via Bluetooth Mesh broadcast, thereby reducing the difference in wake-up time among the nodes.

6. The intelligent management system for a lithium battery pack Bluetooth smart BMS protection board according to claim 1, characterized in that, The hardware watchdog protection unit has built-in independent voltage monitoring channels and current sampling channels, which can detect the overvoltage, undervoltage, overcurrent and short circuit status of the battery in real time. The shutdown signal of the hardware watchdog protection unit directly acts on the gate control circuit of the power MOSFET.

7. The intelligent management system for a lithium battery pack Bluetooth smart BMS protection board according to claim 1, characterized in that, Each BMS node also includes a sleep power management module. After the processing link detects a static state with no load connection and no communication request, it gradually shuts down the clock of the main control MCU and reduces the power supply voltage.

8. The intelligent management system for a lithium battery pack Bluetooth smart BMS protection board according to claim 7, characterized in that, The sleep power management module is equipped with a Bluetooth Mesh low power broadcast unit; During sleep, the processing link periodically sends heartbeat signals through the low-power broadcast unit to inform neighboring nodes that the node is in sleep mode rather than offline or faulty, thus maintaining the integrity of the Mesh network topology.

9. The intelligent management system for a lithium battery pack Bluetooth smart BMS protection board according to claim 1, characterized in that, The system also includes a multi-node parameter self-calibration subsystem, a multi-node Bluetooth Mesh networking subsystem, and a wake-up synchronization and coordination subsystem; The multi-node parameter self-calibration subsystem is electrically connected to the battery pack, and collects the battery pack's terminal voltage, charging and discharging current and surface temperature data in real time. It also integrates an online parameter identification engine. The multi-node Bluetooth Mesh networking subsystem establishes a Bluetooth Mesh protocol stack on each BMS node to realize the wireless self-organizing network function between nodes. After detecting a wake-up signal, the wake-up synchronization and coordination subsystem sends a synchronization wake-up command to all nodes in the network via Mesh broadcast. Each node calculates its own wake-up execution time based on the received broadcast time and the preset wake-up delay parameters.

10. The intelligent management system for a lithium battery pack Bluetooth smart BMS protection board according to claim 1, characterized in that, The system also includes a protection timing optimization subsystem, which operates independently of other subsystems and includes a hardware-level protection execution path. The hardware watchdog module in the protection timing optimization subsystem starts monitoring the battery status before the main control MCU is powered on. When the battery voltage is lower than the over-discharge protection threshold and an external load connection is detected, the power MOSFET is immediately turned off. The protection timing optimization subsystem also includes a wake-up timing control unit, which arranges the timing from the wake-up signal trigger to the protection function readiness and then to the completion of the main control MCU initialization, ensuring that the hardware protection path is in a monitoring state before the main control MCU completes initialization.

Images