Battery pack system and battery safety intelligent monitoring and protecting device

Abstract

The invention discloses a battery pack system and a battery safety intelligent monitoring and protecting device. The battery pack system comprises a main control unit BCU, a shunt, a relay assembly and a battery pack, the main control unit BCU adopts a Loihi2 architecture SNN chip, distributes neuron groups in proportion, and respectively processes voltage, temperature, stress and electrochemical impedance physical field signals. According to the invention, the main control unit BCU integrates the functions of the event-driven brain-like computing architecture and the intelligent power management unit, so that the battery pack is monitored and managed in all directions, and the stability and reliability of the battery system are improved; the shunt integrates a current insurance and multi-dimensional detection integrated module, so that the protection function of the battery system is enhanced, and the probability of fault occurrence is reduced; through the layered pulse decision tree and the three-level response strategy, accurate diagnosis and timely processing of the battery pack fault are realized, the fault diagnosis efficiency and the processing speed of the system are improved, and the safety and the reliability of the battery system are remarkably improved.

Description

Technical Field

[0001] This invention relates to the field of electric vehicle battery management, specifically to a battery pack system and a battery safety intelligent monitoring and protection device. Background Technology

[0002] With the rapid development of new energy vehicle technology, the driving range and charging infrastructure of electric vehicles have not yet met consumers' demands for long-distance driving and fast charging. To address this issue, in-vehicle power banks and portable electric vehicle charging devices have emerged. However, these devices typically require the vehicle to be parked while charging the battery, making it impossible to charge while driving. Furthermore, electric vehicle batteries experience performance degradation and capacity reduction in low-temperature environments, severely impacting driving range. While methods have been developed to use charging stations to provide power for battery pack heating or to utilize the electric vehicle's own lithium battery pack for preheating, these solutions have numerous limitations.

[0003] In the power battery system of electric vehicles, the Battery Management System (BMS) is one of the core technologies. Existing electric vehicle BMS solutions typically use dedicated front-end AD acquisition chips, which communicate with the MCU via SPI through daisy-chain cascading. However, this approach is costly and lacks flexibility. Meanwhile, relays in electric vehicles are used frequently and are prone to failure, making fault diagnosis particularly important. However, sometimes the limited total voltage detection channels of the battery management system prevent fault diagnosis of all relays.

[0004] Furthermore, as the high-voltage platforms of electric vehicles continue to increase, traditional 400V low-voltage charging piles are difficult to match the charging needs of 800V high-voltage platform vehicles. Meanwhile, vehicles with high power demands typically require multiple high-voltage battery packs. The series and parallel connection of these battery packs not only increases material costs and installation difficulty but also limits the overall performance and safety of electric vehicles. Therefore, there is an urgent need for a battery pack system capable of real-time monitoring, fault prediction, thermal management, and providing safety protection to improve the overall performance and safety of electric vehicles.

[0005] Several invention patents have been granted to address issues related to real-time monitoring, fault prediction, thermal management, and safety protection of battery pack systems. For example: CN106740203A discloses a portable mobile battery pack for electric vehicles, including a battery pack, a main positive relay, a main negative relay, a Hall sensor, a battery management system, and a vehicle plug. This patent provides a universal interface and control method for electric vehicles, is small in size and lightweight, easy to carry, and can serve as a general emergency power source for electric vehicles. However, this patent still suffers from insufficient battery pack range and practicality.

[0006] CN115675183A discloses an electrical control structure and cooling system based on a power battery pack, including a battery module group, a current sensor, a main positive relay, a main negative relay, a pre-charge relay, and a pre-charge resistor. This patent achieves more uniform heat dissipation, which greatly helps to enhance the stability and extend the lifespan of the power battery system. However, this patent still suffers from the problem of lacking comprehensive monitoring of the battery pack's temperature and status in its electrical control structure.

[0007] The existing technology has the following drawbacks: 1. Existing electric vehicle battery management systems lack comprehensive monitoring capabilities for battery pack temperature and status, making it impossible to grasp the actual working condition of the battery pack in real time. This can easily lead to a decline in battery performance and affect driving range.

[0008] 2. Traditional battery pack systems have limitations in fault diagnosis, making it difficult to effectively monitor and diagnose faults in all relays, which affects the reliability and safety of the system.

[0009] 3. Existing battery pack systems lack intelligent charging and discharging strategies, and cannot optimize charging and discharging operations according to the actual power status and needs of the battery pack, thus reducing the efficiency and lifespan of the battery pack.

[0010] 4. Current battery pack systems still have shortcomings in thermal management, lacking precise temperature control and heat dissipation capabilities, and are unable to effectively prevent battery pack overheating-related malfunctions and safety hazards.

[0011] 5. Existing battery pack systems lack the ability to acquire and process key parameters such as battery pack voltage and current with high precision, making it impossible to achieve real-time monitoring and precise control of the battery pack status, which affects the stability and reliability of the battery system. Summary of the Invention

[0012] Existing technologies lack comprehensive monitoring capabilities for battery pack temperature and status, failing to provide real-time insights into the actual operating conditions of the battery pack. This leads to problems such as degraded battery performance, incomplete fault diagnosis, and poor thermal management. Therefore, to address these issues, this invention provides a battery pack system with multi-dimensional monitoring, fault prediction, and thermal management functions.

[0013] A battery pack system includes a main control unit (BCU), a shunt, a relay assembly, and a battery pack; The main control unit (BCU) is used to collect voltage, current and other battery data of each battery pack; the relay assembly is used to control the charging and power supply of the battery pack. The relay assembly includes a pre-charge relay, a main positive relay, a main negative relay, and a charging relay. The pre-charge relay controls the pre-charge state of the battery pack, enabling pre-charge operations and preventing power-on shock. The main positive relay controls the output of the battery pack, providing power to external devices or equipment. The main negative relay controls the connection between the main negative terminal of the battery pack and the external circuit, working in conjunction with the main positive relay to achieve the overall switching of the power supply / charging circuit. The charging relay controls the connection between the charger and the charging circuit of the battery pack. The pre-charge relay, main positive relay, main negative relay, and charging relay are all connected to the main control unit (BCU) via status feedback lines. The shunt is used for overcurrent protection, current acquisition, contact resistance monitoring and fault prediction, and is connected in series between the main negative terminal of the battery pack and the input terminal of the main negative relay. The battery pack serves as an energy storage power source to provide power or can be charged by an external charger.

[0014] Preferably, the main control unit (BCU) adopts the Loihi2 architecture SNN chip with a total of 130K neurons. The dedicated neuron groups are allocated in a ratio of 4:3:2:1 to process four-dimensional physical field signals of voltage (52K neurons), temperature (39K neurons), stress (26K neurons), and electrochemical impedance (13K neurons).

[0015] Preferably, the shunt achieves non-destructive current measurement through a Hall sensor array and utilizes magnetic field distribution characteristics for online monitoring of contact resistance and fault prediction; the shunt adopts an array design combining SiC MOSFET pins and solid-state power relays, and the solid-state relay array of the relay assembly is an independent functional component.

[0016] Preferably, the pre-charge relay, the main positive relay, the main negative relay and the charging relay are all designed as an array of solid-state power relays and SiC MOSFETs, with an on-resistance of <1mΩ, a rated on-off current of ≥500A and a response time of <5ms.

[0017] Preferably, the battery pack includes a slave control unit (BMU), multiple individual battery cells, and several acquisition lines; the battery pack is equipped with a 12-bit ADC array, which converts different physical field signals into dedicated pulse sequences through asynchronous pulse timing coding technology; the acquisition end of the slave control unit (BMU) is connected to each individual battery cell through acquisition lines, and the output end is connected to the master control unit (BCU) through a CAN bus to upload the encoded signal to the BCU; the master control unit (BCU) processes the signal using a dedicated group of neurons allocated in a 4:3:2:1 ratio.

[0018] Preferably, the battery pack system further includes a vehicle controller, a charger, and a diagnostic host computer; The vehicle controller is used to connect to the main control unit (BCU) via the vehicle CAN bus, receive battery system data, realize overall control of the electric vehicle, and provide driving force in conjunction with the battery pack system. The charger is used to charge the individual cells of the battery pack system; The diagnostic host computer is used for fault identification and performance monitoring of the battery pack system.

[0019] Preferably, the main control unit (BCU) is connected to the positive and negative terminals of the battery pack via a total voltage detection line; the input and output detection lines of the shunt are connected to the shunt, and the loop current is calculated by collecting the voltage difference between the two ends based on the fixed resistance value of the shunt, thereby collecting the input and output current. The main control unit (BCU) communicates with the battery pack, vehicle controller, charger, and diagnostic host computer via CAN bus, CAN-FD bus (5Mbps), and Time-Sensitive Network (TSN) Ethernet. The dual redundancy switching mechanism detects the communication status by sending a heartbeat packet every 100ms. When three consecutive heartbeat packets are lost in CAN / CAN-FD communication or the communication delay exceeds 2ms, the system automatically switches to TSN Ethernet communication. During the switching process, a ring buffer mechanism is used to temporarily store critical data (buffer capacity ≥10KB). After the switch is completed, the buffered data is transmitted first.

[0020] The main control unit (BCU) also includes a sampling module, a processing module, a power display module, a temperature detection module, a stress detection module, and an electrochemical impedance detection module. The sampling module is used to collect data from the battery pack. The sampling module has multiple sampling input terminals that are connected to the battery pack to collect the voltage of each individual battery cell and protect it. The power display module is connected to the processing module and is used to display the battery capacity; The temperature detection module is used to monitor the battery pack temperature in real time; The stress detection module is used by the stress sensor to collect stress data of individual battery cells; The electrochemical impedance spectroscopy module is used by the electrochemical impedance sensor to collect electrochemical impedance data of individual battery cells.

[0021] A battery safety intelligent monitoring and protection device, wherein the battery is located in the aforementioned battery pack system, and the intelligent monitoring and protection system includes a data acquisition and processing module and a decision and response module, wherein the data acquisition and processing module includes a multi-dimensional data acquisition unit and a signal encoding and fusion unit; The multi-dimensional data acquisition unit is used to acquire individual cell voltage, temperature, stress and electrochemical impedance data through the Hall array sensor integrated in the shunt and the detection lines at both ends, the acquisition lines of the battery pack, the stress sensor and the electrochemical impedance sensor, respectively. The signal encoding and fusion unit is used to input the dedicated pulse sequence of the four-dimensional physical field signal into the BCU for processing by the corresponding dedicated neuron group; through cross-scale STDP learning rules (time window [-20ms, +20ms], weight update step size 0.01-0.1), unsupervised adjustment of synaptic weights (weight range 0-255) is performed to automatically learn the spatiotemporal correlation pattern of the four-dimensional physical field and fuse voltage fluctuation gradient, abnormal temperature rise, internal resistance mutation, stress change threshold, and electrochemical impedance anomaly. The decision and response module includes a hierarchical impulse decision tree and a three-level response module; Hierarchical impulse decision tree: The first layer, consisting of 100 neurons, is used to process normal-abnormal binary classification and filter abnormal signals. The second layer, consisting of 500 neurons, is used to identify fault patterns; The third layer, consisting of 200 neurons, quantifies the severity of the fault and achieves interpretability through a pulse firing frequency-fault level mapping relationship. The three-level response module is used to perform corresponding protective measures based on the fault type and severity: The logic of the Level 3 response module includes: Level L1: Local equalization adjustment, automatically corrects minor parameter deviations without reporting; Level 2: Sends a CAN alarm frame to the BCU, along with the fault type and severity score, so that the whole vehicle system can handle it together. Level L3: Directly triggers relay circuit breaking; Preferably, it also includes a prediction and early warning module, which includes a fault prediction model and a thermal runaway early warning unit; Fault prediction model: This model employs a Liquid State Machine architecture with 800 neurons in the battery reservoir to capture the nonlinear evolution trajectory of battery aging. Experimental conditions include a charge / discharge rate of 1C, a temperature of 25±2℃, and an aging failure criterion of capacity decay to 80% of the initial capacity. The model achieves fault precursor detection 30-50 cycles in advance, with a prediction error of <2.5%. Thermal runaway early warning unit: It is used to collect impedance changes through electrochemical impedance sensors, combine voltage fluctuation data, and establish a mapping model of impedance-voltage-SEI film thickness (mathematical correlation formula: d=k×(Z / Z0)×(U0 / U), where d is SEI film thickness, k is calibration coefficient, Z is real-time impedance, Z0 is initial impedance, U is real-time voltage, and U0 is initial voltage), to capture the correlation between SEI film growth (thickness change > 5nm) and macroscopic voltage fluctuation (< 2mV). Beneficial effects

[0022] 1. This invention integrates an event-driven neuromorphic computing architecture and an intelligent power management unit (iPMU) into the main control unit (BCU), thereby achieving comprehensive monitoring and management of the battery pack. This overcomes the problem of insufficient battery pack monitoring capabilities in the prior art, enabling real-time monitoring of the actual working status of the battery pack and effectively improving the stability and reliability of the battery system. 2. This invention adopts a shunt integrated current fuse and multi-dimensional detection module, and realizes non-destructive current measurement and contact resistance monitoring through Hall array sensor, replacing the traditional current detection method, which greatly enhances the protection function of the battery system and effectively reduces the probability of failure. 3. This invention achieves precise control of battery pack charging and power supply through a solid-state relay array composed of four types of solid-state power relays (SSRs) and SiC MOSFETs, eliminating mechanical contact wear and improving system lifespan and reliability. 4. This invention uses a 12-bit ADC array and a high-precision sensor to achieve high-precision acquisition and monitoring of parameters such as battery pack voltage, current, and temperature. It can detect abnormal conditions in the battery pack in a timely manner and effectively prevent battery failures caused by parameter deviations. 5. This invention achieves accurate diagnosis and timely handling of battery pack faults through hierarchical pulse decision tree and three-level response strategy, improving the system's fault diagnosis efficiency and processing speed, and significantly enhancing the safety and reliability of the battery system. Attached Figure Description

[0023] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0024] Figure 1 This is a schematic diagram of the structure of a battery pack system according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a battery pack system according to another embodiment of the present invention; Figure 3 This is a schematic diagram of a battery pack system according to the present invention. Detailed Implementation

[0025] To facilitate understanding of the present invention, a more complete description will be given below with reference to the accompanying drawings. Preferred embodiments of the invention are shown in the drawings. However, the invention can be implemented in many different forms and is not limited to the embodiments described herein. Rather, these embodiments are provided to provide a thorough and complete understanding of the disclosure of the invention.

[0026] It should be noted that when an element is referred to as being "fixed to" another element, it can be directly attached to the other element or there may be an intervening element. When an element is referred to as being "connected to" another element, it can be directly connected to the other element or there may be an intervening element. The terms "vertical," "horizontal," "left," "right," and similar expressions used herein are for illustrative purposes only and do not represent the only possible implementation.

[0027] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains. The terminology used herein in the description of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. The term "and / or" as used herein includes any and all combinations of one or more of the associated listed items.

[0028] Please see Figure 1 A battery pack system includes: a main control unit (BCU) 100 (Battery Control Unit), a shunt 200 (current fuse and multi-dimensional detection integrated module), a relay assembly 300, and several battery packs 400; the main control unit (BCU) integrates an event-driven neuromorphic computing architecture and intelligent power management unit (iPMU) functions, each battery pack communicates efficiently with the BCU via a CAN bus, and the vehicle controller 500, charger 600, and diagnostic host computer 700 are connected to the BCU through corresponding interfaces to form a complete on-board energy storage and control ecosystem.

[0029] The main control unit BCU100 is used to collect voltage, current and other battery data of each battery pack. The shunt 200 is an integrated module of current protection and multi-dimensional detection, connected in series between the main negative terminal of the battery pack and the input terminal of the main negative relay. It has the functions of overcurrent protection, current acquisition, contact resistance monitoring and fault prediction (its overcurrent protection function realizes the core function of the traditional current protection module, without the need to set up an additional independent current protection module). The relay assembly 300 is used to control the charging and power supply of the battery pack. The battery pack 400 serves as an energy storage power source to provide power or is charged by an external charger.

[0030] Please see Figure 1 The main control unit (BCU) collects the voltage, current, and other battery data of each battery pack.

[0031] The relay assembly 300 includes a pre-charge relay 310, a main positive relay 320, a main negative relay 340, and a charging relay 330. The pre-charge relay 310 is used to pre-charge the charging / power supply circuit and limit the inrush current (connected to a pre-charge resistor with a resistance of 10-20Ω). The main positive relay 320 controls the forward switching of the power supply circuit, outputting current to the vehicle load. The main negative relay 340 acts as a switch connecting the battery pack to the external environment, controlling the connection between the battery pack's main negative terminal and the external circuit, and works with the main positive relay to achieve the overall switching of the power supply / charging circuit. The charging relay 330 controls the forward switching of the charging circuit and receives the input current from the charger. The four types of relays are connected to the main control unit (BCU) via a status feedback line to report their on / off status; at the same time, they receive control commands from the BCU via the CAN bus to ensure the coordination between status monitoring and on / off control.

[0032] The interlocking logic of the relay components is shown in Table 1: Table 1 All four types of relays use solid-state power relays (SSRs) and SiC MOSFETs to form a solid-state relay array. The on-resistance is <1mΩ, the rated on-off current is ≥500A, and the response time is <5ms. This eliminates mechanical contact wear and ensures uniform parameters and synchronized control.

[0033] The main control unit BCU100 is directly connected to the positive and negative terminals of the battery pack via a total voltage detection line to achieve real-time monitoring of the total battery pack voltage. It is connected to the shunt via a shunt input detection line (connected to the negative terminal of the battery pack) and a shunt output detection line (connected to the input terminal of the negative relay). Based on the fixed resistance value of the shunt, the loop current is calculated by collecting the voltage difference between the two ends, accurately collecting the input and output currents to prevent abnormal current.

[0034] The main control unit (BCU) uses a Loihi2 architecture SNN chip with a total of 130K neurons, distributed in a 4:3:2:1 ratio to process four-dimensional physical field signals: voltage (52K neurons), temperature (39K neurons), stress (26K neurons), and electrochemical impedance (13K neurons). The chip's computing power supports 100 neurons achieving a response latency of <3ms, meeting the requirements of hierarchical decision tree operations. The chip's power consumption is ≤120mW, its operating temperature range is -40℃ to 125℃, and it has passed AEC-Q100 automotive-grade certification.

[0035] The shunt performs the following functions: Overcurrent protection: When the current exceeds the rated current of the circuit board, the circuit is automatically cut off to prevent circuit damage; Current acquisition: The loop current is calculated by the BCU based on the voltage difference output from the two detection lines, achieving accurate current acquisition (accuracy ±0.3%, bandwidth >50kHz). Condition monitoring: Contact resistance and early signs of failure are monitored via an integrated Hall array sensor (contact resistance monitoring resolution 0.1mΩ).

[0036] The shunt integrates Hall array sensor technology, integrating an 8×8 Hall sensor array (3mm spacing) on ​​its own copper busbar. It achieves non-destructive current measurement through magnetic field reconstruction algorithm, and at the same time uses magnetic field distribution characteristics to perform online monitoring of contact resistance and fault prediction, replacing the traditional current detection method and enhancing protection functions.

[0037] The current flows in the following direction: Power supply circuit (battery pack supplies power to vehicle load): Battery pack main positive terminal → main positive relay 320 → external vehicle load (such as motor controller) → main negative relay 340 → shunt 200 → battery pack main negative terminal; Charging circuit (charger charging battery pack): Charger positive terminal → Charging relay 330 → Battery pack positive terminal → Battery pack internal cells → Battery pack negative terminal → Shunt 200 → Negative relay 340 → Charger negative terminal; Power supply pre-charge circuit: Battery pack main positive terminal → pre-charge relay 310 → pre-charge resistor → external vehicle load → main negative relay 340 → shunt 200 → battery pack main negative terminal; Pre-charge circuit: Charger positive terminal → Pre-charge relay 310 (parallel branch on the charging side) → Pre-charge resistor → Battery pack positive terminal → Battery pack → Shunt 200 → Total negative relay 340 → Charger negative terminal.

[0038] In one of the battery packs, the battery pack 400 includes a slave control unit BMU410, multiple individual battery cells 420 with an “n series m parallel” design (example: 103 series × 2 parallel, individual cell voltage 3.7V, individual cell capacity 200Ah, total pack voltage 381.1V, total capacity 400Ah) and several acquisition lines 430, with each acquisition line corresponding to an individual battery cell. The acquisition line is responsible for collecting information such as current, voltage, and temperature of individual battery cells. The battery pack adds stress sensors and electrochemical impedance sensors (the sensors are installed on the tabs and sides of the individual battery cells and connected to the BMU via the acquisition line) to improve the acquisition of four-dimensional physical field data. A 12-bit ADC array is configured (using a 5V reference voltage, voltage resolution of 1.22mV), sampling rate of 10kHz. Different physical field signals are converted into dedicated pulse sequences through asynchronous pulse timing coding technology (voltage signals use phase coding, phase difference accuracy ±2°; temperature uses frequency coding, 1-100Hz; stress uses pulse interval coding, 1-50ms; impedance uses pulse amplitude modulation). The acquisition end of the slave control unit (BMU) is connected to each individual battery cell via the acquisition line, and the output end is connected to the BCU via the CAN bus. The encoded signal is uploaded to the BCU, where it is processed by a dedicated group of neurons allocated in a 4:3:2:1 ratio to achieve precise matching between signal representation and neuron operation.

[0039] This embodiment employs a series design of k battery packs (example: 2 battery packs in series, total voltage 762.2V, total capacity 400Ah, meeting the 760V vehicle power supply requirements), and is equipped with a battery pack balancing module (using an active balancing chip) to ensure voltage consistency after multiple packs are connected in series; at the same time, a liquid cooling system is configured with a serpentine flow channel design, a flow channel spacing of 50mm, a coolant flow rate of 5-8L / min, and a temperature control accuracy of ±2℃, ensuring the thermal management requirements of the large-capacity high-voltage battery pack. Compared with the existing single battery pack, the power supply stability and thermal safety are stronger.

[0040] Working principle of the invention (I) Data Acquisition and Processing Multi-dimensional data acquisition: Circuit contact resistance and current data are acquired through Hall array sensors integrated into the shunt and detection lines at both ends. Individual cell voltage, temperature, stress and electrochemical impedance data are acquired through the battery pack acquisition lines, stress sensors and electrochemical impedance sensors, achieving comprehensive coverage of multi-dimensional data with a sampling rate of 10kHz and a time resolution of 0.1ms.

[0041] Signal Encoding and Fusion: For the dedicated pulse sequence of the four-dimensional physical field signal, the corresponding dedicated neuron group is input into the BCU for processing; through the cross-scale STDP learning rule (time window [-20ms, +20ms], weight update step size 0.01-0.1), the synaptic weights are adjusted unsupervised (weight range 0-255), and the spatiotemporal correlation pattern of the four-dimensional physical field is automatically learned, fusing features such as voltage fluctuation gradient (>5mV / s), abnormal temperature rise (>0.3℃ / min), internal resistance mutation (>8%), stress change threshold, and electrochemical impedance anomaly.

[0042] (II) Decision-making and response mechanism Hierarchical pulsatile decision tree First layer (100 neurons): Processes normal-abnormal binary classification, with a response latency of <3ms, and quickly filters abnormal signals (based on the computing power of the Loihi2 chip, the computational load of 100 neurons can meet this response speed). The second layer (500 neurons) identifies 10 types of fault modes (including thermal runaway, abnormal contact resistance, cell aging, relay adhesion, etc.), with an accuracy rate of >94%. The third layer (200 neurons): quantifies the severity of the fault (5-level rating), and achieves interpretability through the mapping relationship of "pulse firing frequency - fault level" (1-5Hz corresponds to level 1, 6-10Hz corresponds to level 2, 11-15Hz corresponds to level 3, 16-20Hz corresponds to level 4, and >20Hz corresponds to level 5).

[0043] Level 3 Response Strategy Level L1: Local equalization adjustment, automatically corrects minor parameter deviations (voltage deviation <5mV, temperature deviation <1℃) without reporting; Level 2: Sends a CAN alarm frame to the BCU, along with the fault type and severity score, to facilitate coordinated processing by the entire vehicle system; Level L3: Directly triggers the relay to disconnect the circuit, quickly blocking risks and ensuring safety; Data transmission optimization: Only L2 / L3 level reports data to BCU, reducing data traffic by 70% compared to traditional solutions.

[0044] (III) Forecasting and Early Warning Fault prediction model: A liquid state machine architecture is adopted, with 800 reservoir neurons to capture the nonlinear evolution trajectory of battery aging. "Cycle cycle" is defined as "one complete charge + one complete discharge". The experimental conditions are charge / discharge rate of 1C, temperature of 25±2℃, and aging failure standard of capacity decay to 80% of the initial capacity. It can detect fault precursors 30-50 cycles in advance with a prediction error of <2.5%.

[0045] Thermal runaway early warning: Impedance changes are collected using an electrochemical impedance spectroscopy sensor (SEI film growth leads to an increase in impedance), and combined with voltage fluctuation data, an impedance-voltage-SEI film thickness mapping model is established (mathematical correlation formula: Where d is the SEI film thickness; k is the calibration coefficient; and Z is the real-time impedance. Let U be the initial impedance and U be the real-time voltage. The initial voltage is used to accurately capture the correlation between SEI film growth (thickness change > 5 nm) and macroscopic voltage fluctuations (< 2 mV); the detection accuracy of the 12-bit ADC array can meet the requirements for voltage fluctuation capture, the sensitivity of thermal runaway feature recognition is 8-12 times that of traditional simple temperature detection methods, it retains the fast response logic of temperature gradient > 5℃ / min, the early warning window is ≥ 5 minutes, the false alarm rate is < 5%, and the false alarm rate is < 0.1%.

[0046] Enhanced safety protection: The temperature detection module monitors the battery pack temperature in real time and cuts off the main circuit when it is too high; the shunt automatically protects itself after the current exceeds the limit; when the contact resistance is abnormal, the BCU triggers an early warning or circuit break to avoid the risk of overheating; it also has overcharge and over-discharge protection functions to prevent damage to individual cells.

[0047] The main control unit (BCU) includes a sampling module, a processing module, a power display module, a temperature detection module, a stress detection module, and an electrochemical impedance spectroscopy (EIS) detection module. It is equipped with a 32-bit MCU for auxiliary computation, performing edge SOC estimation and contactor adhesion detection, achieving a self-diagnostic coverage of 97%. Among these features: The sampling module has multiple sampling input terminals, which are connected to the battery pack to collect the voltage of each individual battery cell and protect it. The power display module is connected to the processing module and is used to display the battery capacity; The temperature detection module monitors the battery pack temperature in real time and cuts off the main circuit when the temperature is too high to avoid affecting battery life.

[0048] It supports dual redundant communication interfaces of CAN, CAN-FD (5Mbps) and TSN Ethernet, reducing communication latency; interface function allocation: CAN bus is used for real-time control data transmission between slave control unit (BMU) and vehicle controller, CAN-FD is used for high-speed data interaction, and TSN Ethernet is used for large-volume detection and upgrade data transmission of the diagnostic host computer. Communication stability is ensured through multi-interface collaboration.

[0049] Please see Figure 2The battery pack system also includes a vehicle controller 500, which is connected to the main control unit (BCU). It should be noted that the vehicle controller controls the vehicle's infotainment system. When the battery pack system and the vehicle controller are connected, they form an on-board system to control the electric vehicle. The system also includes an instrument cluster, forming a complete on-board system capable of supplying power and providing driving force to the electric vehicle. The vehicle controller connects to the main control unit (BCU) via the vehicle CAN bus, receives battery system data, and achieves overall control of the electric vehicle, cooperating with the battery pack system to provide driving force. Simultaneously, it can provide battery status information through the instrument cluster, forming a complete on-board control system.

[0050] Please see Figure 2 The battery pack system also includes a charger 600, which is connected to the main control unit (BCU). The charger 600 is used to charge the individual cells of the battery pack system, and the charging logic is as follows: Step ①: Close the main negative relay, then close the pre-charge relay (charging side branch) to complete the pre-charge and avoid charging surge; Step 2: After pre-charging is completed (the difference between the battery pack voltage and the charger output voltage is detected to be <5%), disconnect the pre-charging relay and simultaneously close the main positive relay and the charging relay; Step 3: The charger charges the individual battery cells through the charging circuit; Step 4: After charging is complete (the voltage of a single cell reaches 4.2V or SOC≥98%), the BCU first disconnects the charging relay and the main positive relay, and then disconnects the main negative relay to ensure that the charging process is safe and controllable.

[0051] Please see Figure 2 The battery pack system also includes a diagnostic host computer 700, which is connected to the main control unit (BCU). The diagnostic host computer 700 is connected to the main control unit (BCU) via TSN Ethernet, enabling comprehensive diagnosis and testing of the battery pack system. It has functions such as fault identification and performance monitoring, and utilizes high-speed Ethernet to transmit large amounts of testing information.

[0052] The main control unit BCU100 also includes a sampling module for acquiring battery pack data. The sampling module, connected to the battery pack, has multiple sampling input terminals connected to the battery pack to acquire the voltage of each individual battery cell and protect it. The negative terminal of the battery pack is connected to the execution switch module (corresponding to a relay assembly) via a shunt (i.e., an integrated current fuse and multi-dimensional detection module), and then connected to the starter motor via the execution switch module. The other end of the starter motor is directly connected to the positive terminal of the battery pack.

[0053] The main control unit (BCU) also includes a processing module, on which a power display module 108 is connected to help display the battery capacity. This embodiment also includes a temperature detection module for real-time monitoring of the battery pack temperature. If the battery pack temperature is too high, it will affect battery life. When the battery pack temperature is too high, the invention will cut off the main circuit through the temperature detection module. Example 1

[0054] A battery pack system includes a main control unit (BCU100), a shunt converter (200), a relay assembly (300), and a battery pack (400). The main control unit (BCU100) uses a Loihi2 architecture SNN chip with a total of 130K neurons, distributed in a 4:3:2:1 ratio to process four-dimensional physical field signals: voltage (52K neurons), temperature (39K neurons), stress (26K neurons), and electrochemical impedance (13K neurons). The chip's computing power supports 100 neurons to achieve a 2ms response latency, meeting the requirements of hierarchical decision tree operations. The chip consumes 110mW, operates within a temperature range of -45℃ to 120℃, and has passed AEC-Q100 automotive-grade certification.

[0055] The shunt 200 features an integrated design, including a current fuse and a multi-dimensional detection module. It achieves non-destructive current measurement through an 8×8 Hall sensor array (3mm spacing) and utilizes magnetic field distribution characteristics for online monitoring of contact resistance and fault prediction. The shunt employs an array design combining SiC MOSFET pins and solid-state power relays (SSRs), with an on-resistance of 0.8mΩ, a rated on / off current of 600A, and a response time of 4ms. The SSR array is a separate functional component from the main relay assembly.

[0056] The relay assembly 300 includes a pre-charge relay 310, a main positive relay 320, a main negative relay 340, and a charging relay 330. All of these utilize an array design composed of solid-state power relays (SSRs) and SiC MOSFETs, with an on-resistance of <1mΩ, a rated on / off current of ≥500A, and a response time of <5ms, ensuring parameter uniformity and synchronized control. The relay assembly is connected to the main control unit (BCU) via a status feedback line, providing real-time feedback of its on / off status and receiving control commands from the BCU via the CAN bus.

[0057] The battery pack 400 includes a slave control unit (BMU) 410, multiple individual battery cells 420 using a "41 series, 3 parallel" design, and several acquisition lines 430. Each individual battery cell uses a quaternary lithium-ion battery with a voltage of 3.7V and a capacity of 250Ah. The total voltage of the battery pack is 140V (41 series × 3.7V), and the total capacity is 600Ah (3 parallel × 250Ah). When two battery packs are connected in series, the total voltage is 280V, and the total capacity is 600Ah, meeting the 280V power supply requirements of the vehicle. The battery pack is equipped with a 12-bit ADC array, using a 5V reference voltage, a voltage resolution of 1.22mV, and a sampling rate of 10kHz. It converts different physical field signals into dedicated pulse sequences using asynchronous pulse timing coding technology.

[0058] The BMU410 slave control unit connects to each individual battery cell via a data acquisition line, and its output connects to the BCU via a CAN bus, uploading the encoded signal to the BCU. The BCU processes the signal using a dedicated group of neurons allocated in a 4:3:2:1 ratio, achieving a precise match between signal representation and neuron computation.

[0059] The system also includes a vehicle controller 500 and a charger 600. The vehicle controller connects to the main control unit (BCU) via the vehicle CAN bus to receive battery system data and achieve overall control of the electric vehicle. The charger uses an 800V platform and charges the battery pack through a charging circuit.

[0060] The diagnostic host computer 700 connects to the main control unit (BCU) via TSN Ethernet to achieve comprehensive diagnosis and testing of the battery pack system, providing functions such as fault identification and performance monitoring. The BCU supports dual-redundant communication interfaces: CAN, CAN-FD (5Mbps), and TSN Ethernet, with a communication latency of 1.5ms. The dual-redundancy switching mechanism involves detecting the communication status using a 100ms heartbeat packet. If three consecutive heartbeat packets are lost in CAN / CAN-FD communication or the communication latency exceeds 2ms, it automatically switches to TSN Ethernet communication. During the switching process, a ring buffer mechanism is used to temporarily store critical data (buffer capacity ≥10KB). After the switch is complete, the buffered data is transmitted first. Interface function allocation: the CAN bus is used for real-time control data transmission between the slave control unit (BMU) and the vehicle controller; CAN-FD is used for high-speed data interaction; and TSN Ethernet is used for large-volume detection and upgrade data transmission for the diagnostic host computer.

[0061] The main control unit BCU100 also includes a sampling module, a processing module, a power display module, a temperature detection module, a stress detection module, and an electrochemical impedance spectroscopy (EIS) detection module. The sampling module has multiple sampling input terminals, connects to the battery pack, and collects the voltage of each individual cell for protection. The power display module connects to the processing module and displays the battery capacity. The temperature detection module monitors the battery pack temperature in real time and cuts off the main circuit when the temperature is too high to prevent impact on battery life.

[0062] The system is also equipped with a liquid cooling system, featuring a serpentine flow channel design with a channel spacing of 50mm, a coolant flow rate of 6-9L / min, and a temperature control accuracy of ±1.5℃, ensuring the thermal management requirements of large-capacity high-voltage battery packs. Through intelligent management by the main control unit (BCU), real-time monitoring, fault prediction, thermal management, and safety protection of the battery pack are achieved, improving the stability and reliability of the battery system.

[0063] 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 claimed invention.

[0064] 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 claimed invention.

Claims

1. A battery pack system, characterized in that, It includes a main control unit (BCU) (100), a shunt (200), a relay assembly (300), and a battery pack (400). The main control unit (BCU) (100) is used to collect the voltage, current and other battery data of each battery pack (400); the relay assembly (300) is used to control the charging and power supply of the battery pack (400). The relay assembly (300) includes a pre-charge relay (310), a main positive relay (320), a main negative relay (340), and a charging relay (330). The pre-charge relay (310) is used to control the pre-charge state of the battery pack (400) to realize the pre-charge operation and avoid power-on shock. The main positive relay (320) is used to control the output of the battery pack (400) to realize the power supply to external devices or equipment. The main negative relay (340) is used to control the connection and disconnection between the main negative terminal of the battery pack (400) and the external circuit, and works with the main positive relay (320) to realize the overall switching of the power supply / charging circuit. The charging relay (330) is used to control the connection and disconnection of the charging circuit between the charger and the battery pack (400). The pre-charge relay (310), the main positive relay (320), the main negative relay (340), and the charging relay (330) are all connected to the main control unit (BCU) (100) through a status feedback line. The shunt (200) is used for overcurrent protection, current acquisition, contact resistance monitoring and fault prediction, and is connected in series between the total negative terminal of the battery pack (400) and the input terminal of the total negative relay; The battery pack (400) serves as an energy storage power source to provide power.

2. The battery pack system according to claim 1, characterized in that, The main control unit BCU (100) adopts the Loihi2 architecture SNN chip with a total of 130K neurons. The dedicated neuron groups are allocated in a ratio of 4:3:2:1 to process four-dimensional physical field signals of voltage, temperature, stress and electrochemical impedance respectively.

3. The battery pack system according to claim 2, characterized in that, The shunt (200) achieves non-destructive current measurement through a Hall sensor array and uses magnetic field distribution characteristics for online monitoring of contact resistance and fault prediction. The shunt (200) adopts an array design combining SiC MOSFET pins and solid-state power relays.

4. The battery pack system according to claim 3, characterized in that, The precharge relay (310), the total positive relay (320), the total negative relay (340) and the charging relay (330) are all designed as an array of solid-state power relays and SiC MOSFETs.

5. The battery pack system according to claim 4, characterized in that, The battery pack (400) includes a slave control unit (BMU) (410), multiple individual battery cells (420), and several acquisition lines (430). The battery pack (400) is equipped with a 12-bit ADC array, which converts different physical field signals into dedicated pulse sequences through asynchronous pulse timing coding technology. The acquisition end of the slave control unit (BMU) (410) is connected to each individual battery cell (420) through the acquisition line (430), and the output end is connected to the master control unit (BCU) (100) through the CAN bus to upload the encoded signal to the master control unit (BCU) (100). The master control unit (BCU) (100) processes the signal using a dedicated group of neurons allocated in a 4:3:2:1 ratio.

6. The battery pack system according to claim 1, characterized in that, The battery pack system also includes a vehicle controller (500), a charger (600), and a diagnostic host computer (700). The vehicle controller (500) is used to connect to the main control unit BCU (100) via the vehicle CAN bus, receive battery system data, and realize overall control of the electric vehicle; The charger (600) is used to charge the individual cells (420) of the battery pack system; The diagnostic host computer (700) is used for fault identification and performance monitoring of the battery pack system.

7. The battery pack system according to claim 6, characterized in that, The main control unit BCU (100) is connected to the positive terminal and the negative terminal of the battery pack (400) respectively through the total voltage detection line; the input detection line and the output detection line of the shunt (200) are connected to the shunt. Based on the fixed resistance value of the shunt (200), the loop current is calculated by collecting the voltage difference between the two ends, thereby collecting the input and output current. The main control unit (BCU) (100) communicates with the battery pack (400), vehicle controller (500), charger (600) and diagnostic host computer (700) via CAN bus, CAN-FD bus and TSN Ethernet.

8. The battery pack system according to claim 1, characterized in that, The main control unit (BCU) also includes a sampling module, a processing module, a power display module, a temperature detection module, a stress detection module, and an electrochemical impedance detection module. The sampling module is used to collect data from the battery pack (400). The sampling module has multiple sampling input terminals connected to the battery pack, and is used to collect the voltage of each individual battery and protect it. The power display module is connected to the processing module and is used to display the battery capacity; The temperature detection module is used to monitor the battery pack temperature in real time; The stress detection module is used by the stress sensor to collect stress data of a single battery cell (420); An electrochemical impedance detection module is used to collect electrochemical impedance data of a single cell (420) by an electrochemical impedance sensor.

9. A battery safety intelligent monitoring and protection device, characterized in that, The battery is disposed in the battery pack system according to any one of claims 1-8, and the intelligent monitoring and protection system includes a data acquisition and processing module and a decision and response module. The data acquisition and processing module includes a multi-dimensional data acquisition unit and a signal encoding and fusion unit. The multi-dimensional data acquisition unit is used to acquire four-dimensional data of voltage, temperature, stress and electrochemical impedance of a single cell (420) through the Hall array sensor integrated in the shunt (200) and the detection lines at both ends, the acquisition line (430) of the battery pack (400), the stress sensor and the electrochemical impedance sensor. The signal encoding and fusion unit is used to input the dedicated pulse sequence of the four-dimensional physical field signal into the main control unit BCU (100) for processing of the corresponding dedicated neuron group; through the cross-scale STDP learning rule, the synaptic weight is adjusted unsupervised, and the spatiotemporal correlation pattern of the four-dimensional physical field is automatically learned, and voltage fluctuation gradient, abnormal temperature rise, internal resistance mutation, stress change threshold, and electrochemical impedance anomaly are fused. The decision and response module includes a hierarchical impulse decision tree and a three-level response module; Hierarchical impulse decision tree: The first layer, consisting of 100 neurons, is used to process normal-abnormal binary classification and filter abnormal signals. The second layer, consisting of 500 neurons, is used to identify fault patterns; The third layer, consisting of 200 neurons, quantifies the severity of the fault and achieves interpretability through a pulse firing frequency-fault level mapping relationship. The three-level response module is used to perform corresponding protective measures based on the fault type and severity: The logic of the Level 3 response module includes: Level L1: Local equalization adjustment, automatically corrects minor parameter deviations without reporting; Level 2: Sends a CAN alarm frame to the BCU, along with the fault type and severity score, so that the whole vehicle system can handle it together. Level L3: Directly triggers relay circuit breaking.

10. The battery safety intelligent monitoring and protection device according to claim 9, characterized in that, It also includes a prediction and early warning module, which includes a fault prediction model and a thermal runaway early warning unit; Fault prediction model: Used with a liquid state machine architecture, configured with 800 reservoir neurons, to capture the nonlinear evolution trajectory of battery aging; Thermal runaway early warning unit: It is used to collect impedance changes through electrochemical impedance sensors, combine them with voltage fluctuation data, establish an impedance-voltage-SEI film thickness mapping model, and capture the correlation between SEI film growth and macroscopic voltage fluctuations.

Images