A heavy truck-mounted BMS system battery fault prediction and alarm method and device

Abstract

A method and device for battery fault prediction and alarm in a heavy-duty truck on-board BMS system, relating to the field of battery management system technology, is disclosed. The method includes data acquisition, heavy-duty truck-specific characteristic quantity calculation, latent fault prediction, graded alarm and linkage handling, fault tracing, and data updating. Data acquisition involves the heavy-duty truck on-board BMS main control unit collecting various raw data from the power battery pack through an acquisition module at an adaptive cycle, with the acquisition cycle adjusted according to operating conditions. Heavy-duty truck-specific characteristic quantity calculation involves the heavy-duty truck on-board BMS main control unit processing the collected raw data, calculating differential pressure characteristics, temperature difference characteristics, voltage change characteristics, internal resistance characteristics, and operating condition correlation characteristics in real time, and outputting the calculated data. Latent fault prediction involves the heavy-duty truck on-board BMS main control unit calling a heavy-duty truck fault prediction model and comparing the calculated multi-dimensional characteristic quantity data with preset thresholds and fault prediction rules.

Description

Technical Field

[0001] This application relates to the field of battery management system technology, and in particular to a method and device for predicting and alarming battery faults in a heavy-duty truck on-board BMS system. Background Technology

[0002] The BMS (Battery Management System) is a crucial link connecting the vehicle's power battery and the electric vehicle. Existing heavy-duty truck BMS systems have significant drawbacks in fault prediction and alarm functions, failing to adapt to the specific scenarios of heavy-duty trucks: First, they rely on post-event threshold protection, which cannot predict latent faults and can easily lead to battery damage or safety accidents; second, single-parameter judgment is susceptible to operating conditions and electromagnetic interference, resulting in high false alarm (≥5%) and missed alarm (≥3%) rates; third, they lack a tiered alarm mechanism and cannot be linked to the vehicle control system; fourth, fault information cannot be synchronized to the fleet management platform, leading to untimely handling; and fifth, they have high implementation costs, poor adaptability, and require additional hardware. Summary of the Invention

[0003] This application provides a method and device for predicting and alarming battery faults in a heavy-duty truck-mounted BMS system, which is adapted to heavy-duty truck usage scenarios and can predict hidden faults in advance.

[0004] This application provides a method for predicting and alarming battery faults in a heavy-duty truck-mounted BMS system, including steps S1 to S6.

[0005] S1: Data Acquisition. The heavy-duty truck's onboard BMS main control unit collects various raw data from the power battery pack through the acquisition module at an adaptive cycle. The acquisition cycle is adjusted according to the operating conditions. S2: Heavy-duty Truck-Specific Characteristic Calculation. The heavy-duty truck's onboard BMS main control unit processes the collected raw data, calculates pressure difference characteristics, temperature difference characteristics, voltage change characteristics, internal resistance characteristics, and operating condition correlation characteristics in real time, and outputs the calculated data.

[0006] S3: Latent Fault Prediction. The heavy-duty truck's onboard BMS main control unit calls the heavy-duty truck fault prediction model, compares the calculated multi-dimensional feature data with preset thresholds and fault prediction rules, and completes the prediction by combining a dual anti-interference and anti-shake mechanism, and outputs the fault prediction result. S4: Tiered Alarm and Linked Response. Based on the severity of the fault prediction result, a three-tiered fault alarm is triggered, linking the vehicle control system and fleet management platform to execute response strategies.

[0007] S5: Fault Origin Tracing. After a fault alarm is triggered, the heavy-duty truck's onboard BMS main control unit automatically records all fault-related data, generates an origination report, and uploads it to the fleet management platform to facilitate fault origination and maintenance optimization. S6: Data Update. After the fault is resolved, the heavy-duty truck's onboard BMS main control unit automatically clears alarm records, updates cell parameters and prediction thresholds, and restarts the fault prediction mechanism to adapt to the subsequent driving conditions of the heavy-duty truck.

[0008] The method provided in this application solves five major problems in the fault prediction and alarm of existing heavy-duty truck on-board BMS, and constructs a full-process fault management system adapted to the specific scenarios of heavy-duty trucks. This method, through S1 adaptive cycle data acquisition, breaks the limitations of fixed acquisition modes and can dynamically adjust the acquisition frequency according to the heavy-duty truck's driving conditions, ensuring both the timeliness and accuracy of data acquisition while avoiding resource waste caused by invalid acquisition.

[0009] The calculation of S2 heavy-duty truck-specific characteristic quantities focuses on the core battery status parameters and operating condition-related parameters during the operation of the heavy-duty truck. Compared with the existing single parameter judgment, it achieves full coverage of multi-dimensional data, providing comprehensive and accurate data support for subsequent fault prediction.

[0010] The S3 latent fault prediction process, combined with a heavy truck fault prediction model and a dual anti-interference and anti-shake mechanism, effectively solves the problems of existing technologies' inability to predict latent faults and susceptibility to interference due to post-event threshold protection. It can identify early signs of battery faults in advance, thus avoiding battery damage and safety accidents.

[0011] The S4 graded alarm and linkage mechanism establishes a three-level graded mechanism, which not only solves the pain points of no graded alarm and inability to link with the vehicle control system, but also realizes linkage with the fleet management platform to ensure the timeliness of fault handling.

[0012] The S5 fault tracing function automatically records all fault-related data and generates a tracing report, providing data support for fault repair and model optimization, and improving the maintainability of the heavy-duty truck BMS system.

[0013] In the S6 data update process, alarm records can be automatically cleared and parameters and thresholds updated to ensure that the system can continuously adapt to the subsequent driving conditions of heavy trucks, thereby improving the system's adaptability and long-term stability.

[0014] In summary, the method provided in this application realizes closed-loop management of fault prediction, alarm, handling, source tracing and optimization, which comprehensively improves the fault control capability of heavy truck on-board BMS system and adapts to the usage requirements of complex driving scenarios of heavy trucks.

[0015] In some embodiments of this application, the raw data collected includes cell data, operating condition data, environmental data, and auxiliary data. After preliminary filtering, the raw data is stored in a local cache and simultaneously uploaded to the vehicle CAN bus and fleet management platform.

[0016] The raw data encompasses cell data, operating condition data, environmental data, and auxiliary data, comprehensively covering the core influencing factors in the operation of heavy-duty truck batteries. This solves the problem of incomplete data collection leading to inaccurate predictions in existing technologies. Simultaneously, after preliminary filtering, the raw data is both stored in a local cache and simultaneously uploaded to the vehicle's CAN bus and fleet management platform. This achieves both local data retention and remote synchronization, ensuring data security and providing multi-channel data support for subsequent fault prediction, coordinated response, and source tracing. This further enhances the system's data management capabilities and collaboration, laying a solid data foundation for the smooth operation of the entire fault prediction and alarm process.

[0017] In some embodiments of this application, in the latent fault prediction S3, time window filtering removes instantaneous glitches, and multi-cycle confirmation anti-jitter is used to avoid interference and misjudgment. Only when the fault prediction rules are met for N consecutive cycles and interference is eliminated can it be determined as a fault precursor of the corresponding type.

[0018] A dual anti-interference strategy is adopted, which combines time window filtering to eliminate instantaneous glitches and multi-cycle confirmation anti-jitter. Time window filtering can filter out abnormal data caused by instantaneous interference, avoiding misjudgments caused by instantaneous fluctuations. Multi-cycle confirmation anti-jitter only determines a fault precursor when the fault prediction rules are met for N consecutive cycles and interference is eliminated, further improving the accuracy of fault prediction, significantly reducing the false alarm rate and missed alarm rate, ensuring the reliability of fault prediction results, avoiding unnecessary maintenance due to false alarms, and preventing safety hazards caused by missed alarms, thereby improving the safety and stability of heavy truck battery operation.

[0019] In some embodiments of this application, the three-level classification of fault alarms includes: Level 1 warning only records and uploads; Level 2 alarm triggers audible and visual prompts, limits power, prohibits heavy loads, and reminds for nearby maintenance; Level 3 emergency alarm triggers continuous alarms, slows down the vehicle, cuts off power, and locks fault data.

[0020] Level 1 warning is suitable for minor abnormalities, avoiding unnecessary audio and visual prompts and vehicle restrictions, and ensuring the normal operation of heavy trucks; Level 2 alarm can promptly remind staff to handle the situation before the fault develops further, preventing the fault from worsening, while also taking into account the safety and economy of heavy truck operation; Level 3 emergency alarm can quickly take emergency measures in the event of a serious fault, minimizing the occurrence of safety accidents and protecting the safety of the vehicle, battery and personnel.

[0021] In some embodiments of this application, during data acquisition S1, the acquisition module collects raw data in real time. Heavy-duty trucks operate under complex and variable conditions, and the battery status also changes in real time with these conditions. Real-time acquisition of raw data can promptly capture abnormal fluctuations in the battery, providing real-time data support for subsequent characteristic quantity calculations and latent fault prediction. This avoids missed fault detection or untimely prediction due to data acquisition delays, further improving the timeliness of fault prediction, ensuring that early signs of faults can be identified, allowing sufficient time for fault handling, and reducing the probability of battery damage and safety accidents.

[0022] In some embodiments of this application, cell data includes single cell voltage, internal resistance, and capacity decay rate; operating condition data includes discharge rate, current, vehicle speed, and load; environmental data includes battery pack and external temperature, and temperature change rate; and auxiliary data includes timestamp, BMS operating status, and line status.

[0023] By refining the specific composition of cell data, operating condition data, environmental data, and auxiliary data, the data collection process becomes more standardized and accurate, avoiding omissions or redundancies. At the same time, the refined data types provide more accurate input for subsequent calculations of heavy-duty truck-specific features, ensuring the accuracy of feature calculations and thus improving the reliability of fault prediction. This further adapts to the complex driving conditions and battery operation requirements of heavy-duty trucks, providing stronger data support for the entire fault prediction and alarm process.

[0024] A battery fault prediction and alarm device for a heavy-duty truck on-board BMS system is provided, which is applied to the above-mentioned battery fault prediction and alarm method for a heavy-duty truck on-board BMS system. The battery fault prediction and alarm device for a heavy-duty truck on-board BMS system includes a data acquisition module, a main control unit, a storage module, an alarm module, a vehicle linkage module, and a remote communication module; the main control unit, storage module, alarm module, vehicle linkage module, and remote communication module are sequentially connected by signals.

[0025] The acquisition module can collect various raw data from the power battery pack according to an adaptive cycle, and transmit the raw data to the main control unit set in the heavy truck's on-board BMS via electrical connection. The calculated data after being processed by the main control unit is stored in the storage module. After feature comparison in the storage module, it is determined whether to trigger the alarm module. The alarm module can issue corresponding alarm information according to the determination result, and generate corresponding actions through the vehicle linkage module. The alarm information and calculated data are transmitted to the fleet management platform through the remote communication module.

[0026] The heavy-duty truck-mounted BMS system battery fault prediction and alarm device provided in this application solves the problems of high implementation cost, poor adaptability, and the need for additional hardware in existing technologies. The device constructs a complete hardware architecture through signal connections of a data acquisition module, a main control unit, a storage module, an alarm module, a vehicle linkage module, and a remote communication module. Each module has a clear division of labor and works collaboratively, enabling precise implementation of the various processes of the above-mentioned methods.

[0027] Meanwhile, the device requires no additional hardware and can be adapted to the existing heavy-duty truck BMS hardware architecture, reducing implementation costs and adaptation difficulties, and improving the device's practicality and economy. The device achieves linkage with the fleet management platform through the remote communication module and with the vehicle control system through the vehicle linkage module, ensuring that fault alarm information can be synchronized in a timely manner and that handling measures can be executed quickly, further improving the efficiency and safety of heavy-duty truck battery fault management.

[0028] In some embodiments of this application, the acquisition module includes a cell parameter acquisition submodule group, an operating condition parameter acquisition submodule group, and an environmental parameter acquisition submodule group, to acquire cell parameters, operating condition parameters, and environmental parameters respectively.

[0029] The data acquisition module is divided into three sub-modules: cell parameter acquisition, operating condition parameter acquisition, and environmental parameter acquisition. Each sub-module is specifically responsible for acquiring its corresponding type of data, which avoids interference between different types of data acquisition and improves the accuracy of data acquisition. At the same time, this modular design makes the maintenance and upgrading of the acquisition module more convenient. The acquisition parameters of each sub-module can be adjusted according to different models of heavy trucks and different operating conditions, further improving the adaptability of the device and ensuring that it can accurately acquire various types of raw data, providing reliable data support for subsequent fault prediction and alarm.

[0030] In some embodiments of this application, the storage module is equipped with corresponding heavy-duty truck fault prediction models and preset thresholds for different heavy-duty trucks. Different heavy-duty trucks have different vehicle models, operating conditions, and battery configurations, resulting in different fault prediction requirements and parameter thresholds. This claim clarifies that the storage module can be configured with corresponding heavy-duty truck fault prediction models and preset thresholds for different heavy-duty trucks, enabling the device to flexibly adapt to different types of heavy-duty trucks without requiring large-scale modifications, thus reducing adaptation costs. Simultaneously, targeted model and threshold settings can improve the accuracy of fault prediction, ensuring the device can accurately adapt to the operational needs of different heavy-duty trucks, further expanding the device's applicability.

[0031] A method for predicting and alarming battery faults in a heavy-duty truck-mounted BMS system, comprising S1 to S6.

[0032] S1: Data Acquisition. The heavy-duty truck onboard BMS main control unit collects various raw data of the power battery pack through the acquisition module at an adaptive cycle. The acquisition cycle is adjusted according to the working conditions. S2: Calculation of Heavy-Duty Truck-Specific Characteristic Quantities. The heavy-duty truck onboard BMS main control unit processes the collected raw data and calculates the differential pressure characteristics, temperature difference characteristics, voltage change characteristics, internal resistance characteristics, and working condition correlation characteristics in real time, and outputs the calculated data.

[0033] S3: Latent fault prediction. The heavy truck onboard BMS main control unit calls the heavy truck fault prediction model, compares the calculated multi-dimensional feature data with the preset threshold and fault prediction rules, and removes instantaneous spikes through time window filtering and anti-shake confirmation through multiple cycles to avoid interference and misjudgment. Only when the fault prediction rules are met for N consecutive cycles and interference is eliminated can it be determined as a fault precursor of the corresponding type and the fault prediction result is output.

[0034] S4: Tiered Alarm and Linked Response: Based on the severity of the fault prediction results, three levels of fault alarms are triggered, linking the vehicle control system and fleet management platform to execute response strategies; S5: Fault Source Tracing: After a fault alarm is triggered, the heavy-duty truck's onboard BMS main control unit automatically records all fault-related data, generates a source tracing report, and uploads it to the fleet management platform for fault source tracing and maintenance optimization; S6: Data Update: After the fault is handled, the heavy-duty truck's onboard BMS main control unit automatically clears alarm records, updates cell parameters and prediction thresholds, and restarts the fault prediction mechanism to adapt to the subsequent driving conditions of the heavy-duty truck.

[0035] The dual anti-interference and anti-shake mechanism (time window filtering and multi-cycle confirmation anti-shake) is directly integrated into the S3 latent fault prediction stage, clarifying the specific execution process of the anti-interference mechanism and making the fault prediction process standardized and rigorous. By eliminating instantaneous glitches through time window filtering and avoiding interference-induced misjudgments through multi-cycle confirmation anti-shake, the false alarm rate and false negative rate can be effectively reduced, solving the pain points of existing technologies that are affected by operating conditions and electromagnetic interference. At the same time, this method retains the closed-loop design of the entire process, realizing complete control of fault prediction, alarm, handling, tracing and optimization. It not only ensures the reliability of fault prediction, but also ensures the timeliness and pertinence of fault handling, further improving the fault management capability of heavy truck on-board BMS system, adapting to the complex driving scenarios of heavy trucks, and avoiding battery damage and safety accidents. Attached Figure Description

[0036] The accompanying drawings are provided to further understand the technical solutions of the present invention and constitute a part of the specification. They are used together with the embodiments of this application to explain the technical solutions of the present invention and do not constitute a limitation on the technical solutions of the present invention.

[0037] Figure 1This is a schematic block diagram illustrating the battery fault prediction and alarm method for a heavy-duty truck-mounted BMS system provided in an embodiment of this application.

[0038] Figure 2 This is a schematic block diagram of the raw data in the battery fault prediction and alarm method for heavy-duty truck BMS system provided in the embodiments of this application.

[0039] Figure 3 A schematic block diagram illustrating the module connection relationship of the battery fault prediction and alarm device for a heavy-duty truck BMS system provided in this embodiment of the application. Detailed Implementation

[0040] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. Based on the embodiments of this application, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of this application.

[0041] It should be noted that all directional indications (such as up, down, left, right, front, back, etc.) in the embodiments of the present invention are only used to explain the relative positional relationship and movement of each component in a certain specific posture (as shown in the figure). If the specific posture changes, the directional indication will also change accordingly.

[0042] The terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Therefore, a feature defined as "first" or "second" may explicitly or implicitly include one or more of that feature. In the description of this application, unless otherwise stated, "a plurality of" means two or more.

[0043] In the description of this application, it should be noted that, unless otherwise expressly specified and limited, the terms "connected" and "linked" should be interpreted broadly, for example, as a fixed connection, a detachable connection, or an integral connection. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances. Furthermore, when describing pipelines, the terms "connected" and "linked" as used in this application have the meaning of establishing electrical connection. The specific meaning needs to be understood in conjunction with the context.

[0044] In the embodiments of this application, the terms "exemplary" or "for example" are used to indicate that something is an example, illustration, or description. Any embodiment or design that is described as "exemplary" or "for example" in the embodiments of this application should not be construed as being more preferred or advantageous than other embodiments or design. Specifically, the use of the terms "exemplary" or "for example" is intended to present the relevant concepts in a specific manner.

[0045] The BMS (Battery Management System) is a crucial link connecting the vehicle's power battery and the electric vehicle. Existing heavy-duty truck BMS systems have significant drawbacks in fault prediction and alarm functions, failing to adapt to the specific scenarios of heavy-duty trucks: First, they rely on post-event threshold protection, which cannot predict latent faults and can easily lead to battery damage or safety accidents; second, single-parameter judgment is susceptible to operating conditions and electromagnetic interference, resulting in high false alarm (≥5%) and missed alarm (≥3%) rates; third, they lack a tiered alarm mechanism and cannot be linked to the vehicle control system; fourth, fault information cannot be synchronized to the fleet management platform, leading to untimely handling; and fifth, they have high implementation costs, poor adaptability, and require additional hardware.

[0046] To address the aforementioned issues, this application provides a method for predicting and alarming battery faults in a heavy-duty truck's onboard BMS system, and a device for predicting and alarming battery faults in a heavy-duty truck's onboard BMS system.

[0047] A method for predicting and alarming battery faults in a heavy-duty truck-mounted BMS system, comprising steps S1 to S6.

[0048] S1: Data Acquisition. The heavy-duty truck's onboard BMS main control unit collects various raw data from the power battery pack through the acquisition module at an adaptive cycle. The acquisition cycle is adjusted according to the operating conditions.

[0049] In data acquisition, the acquisition module can be made of high-temperature resistant and electromagnetic interference-resistant materials, with an overall flat rectangular structure for easy fitting into the heavy-duty truck battery pack. Its dimensions are adapted to the pre-reserved installation space of the battery pack, allowing for stable mounting on the side or inside. The acquisition module can consist of multiple parts, installed in different areas inside and outside the vehicle.

[0050] The acquisition module and the main control unit are connected by a waterproof cable. The connection method can be a plug-in snap-fit ​​connection, which not only ensures the stability of signal transmission, but also facilitates later inspection and maintenance. Its position is close to the power battery pack, which can shorten the data transmission distance, reduce signal loss, and ensure the timeliness of raw data acquisition.

[0051] The adaptive adjustment of the data acquisition cycle is based on the real-time operating conditions of the heavy truck. When the heavy truck is under heavy load, climbing, or other high-intensity operating conditions, the data acquisition cycle is shortened to 1-2 seconds to ensure that changes in battery status can be captured quickly. When the heavy truck is under constant speed, idling, or other smooth operating conditions, the data acquisition cycle can be adjusted to 5-10 seconds to avoid wasting resources due to invalid data acquisition and to achieve a balance between data acquisition efficiency and resource consumption.

[0052] S2: Heavy-duty truck-specific characteristic calculation. The heavy-duty truck onboard BMS main control unit processes the collected raw data and calculates pressure difference characteristics, temperature difference characteristics, voltage change characteristics, internal resistance characteristics, and operating condition correlation characteristics in real time, and outputs the calculated data.

[0053] In the calculation of the special feature quantity of heavy trucks, the main control unit can use an industrial-grade chip as the core component. The shell can be made of aluminum alloy, which has good heat dissipation and impact resistance. It has a cube-shaped structure and multiple interfaces on the surface for connecting with components such as the acquisition module, storage module, and alarm module.

[0054] The main control unit is installed in the electrical control cabinet in the heavy-duty truck cab. It should be in a dry, well-ventilated environment to avoid the impact of high temperature and humidity on its operation. The main control unit and the acquisition module transmit data via a CAN bus, enabling rapid reception and processing of raw data. It is connected to the storage module via a high-speed data cable to ensure that the calculated data can be stored in real time.

[0055] The characteristic calculation process closely matches the operating conditions of heavy-duty trucks. Among them, the voltage difference characteristic calculates the difference between the voltage of a single cell and the average voltage of the entire battery pack, as well as the maximum voltage difference of the entire battery pack; the temperature difference characteristic calculates the difference between the temperature of a single cell and the average temperature of the battery pack, the maximum temperature difference of the battery pack, and the temperature change rate; the voltage change characteristic calculates the voltage change rate of a single cell and the voltage surge / drop rate; the internal resistance characteristic calculates the real-time value of the internal resistance of a single cell and the internal resistance fluctuation amplitude; and the operating condition correlation characteristic calculates the duration of discharge rate and the current fluctuation amplitude. By comprehensively capturing battery status and operating condition changes through multi-dimensional characteristic quantities, accurate data support is provided for fault prediction.

[0056] S3: Latent fault prediction. The heavy truck on-board BMS main control unit calls the heavy truck fault prediction model, compares the calculated multi-dimensional feature data with the preset threshold and fault prediction rules, and completes the prediction by combining the dual anti-interference and anti-shake mechanism and outputs the fault prediction result.

[0057] In the prediction of hidden faults, the heavy truck fault prediction model is stored in the storage module. The storage module can be a solid-state drive, with a cuboid structure and small size. It can be directly installed next to the main control unit and connected to the main control unit via a data cable. It can quickly respond to the call commands of the main control unit and read model data and preset thresholds.

[0058] The dual anti-interference and anti-shake mechanism includes time window filtering and multi-cycle confirmation anti-shake. Time window filtering can eliminate instantaneous spikes and filter out instantaneous abnormal data caused by electromagnetic interference and road condition fluctuations, avoiding misjudgments caused by instantaneous fluctuations. Multi-cycle confirmation anti-shake continuously monitors data for multiple cycles. Only when N consecutive cycles meet the fault prediction rules and interference is eliminated is it determined to be a fault precursor of the corresponding type. The value of N can be adjusted according to the heavy truck model and operating conditions. It is generally set to 3-5 cycles to ensure the accuracy of fault prediction.

[0059] During the prediction process, the main control unit receives the preset thresholds and fault prediction rules transmitted by the storage module in real time, compares them one by one with the calculated feature data, and completes the prediction by combining the anti-interference and anti-shaking mechanism. The prediction results are transmitted to the alarm module and the vehicle linkage module in real time.

[0060] S4: Tiered alarm and linkage response. Based on the severity of the fault prediction results, a three-tiered fault alarm is triggered, which is linked with the vehicle control system and fleet management platform to execute the response strategy.

[0061] In the graded alarm and linkage response, the alarm module can be made of high-temperature resistant and waterproof plastic material, with a circular shape and indicator lights and buzzers on the surface. It is installed on the dashboard of the heavy truck cab, so that the driver can easily see and hear the alarm prompts. It can be connected to the main control unit through wires, and can quickly receive alarm commands transmitted by the main control unit and issue corresponding alarm signals.

[0062] The specific settings of the three-level alarm system can be tailored to the operational needs of heavy-duty trucks. Level 1 warnings only record and upload information; the alarm module does not issue audible or visual alerts. The main control unit simply records the fault information and uploads it to the fleet management platform, without affecting the normal operation of the heavy-duty truck. Level 2 alarms trigger audible and visual alerts; the yellow indicator light on the alarm module flashes, and the buzzer emits an intermittent warning sound. Simultaneously, the main control unit coordinates with the vehicle control system to limit the charging and discharging power of the heavy-duty truck, prohibit heavy loading, and remind the driver to have the truck inspected nearby. Level 3 emergency alarms trigger continuous alarms; the red indicator light on the alarm module remains constantly lit, and the buzzer emits a continuous, sharp warning sound. Simultaneously, the main control unit coordinates with the vehicle control system to gradually reduce the vehicle speed and bring it to a safe stop, disconnecting the charging and discharging relays, shutting off the power battery output, and locking the fault data to prevent accidents.

[0063] The vehicle control system and the main control unit are linked via a CAN bus, enabling the system to quickly receive commands from the main control unit and execute corresponding vehicle control operations. The fleet management platform and the main control unit synchronize data via a remote communication module. This module can be a 4G / 5G module with a metal casing, designed for interference resistance and stable signal transmission. It is mounted on the top of the heavy truck and connected to the main control unit via a wire. It can upload fault alarm information and prediction results to the fleet management platform in real time, facilitating remote monitoring and maintenance by management personnel.

[0064] S5: Fault tracing. After a fault alarm is triggered, the heavy-duty truck's onboard BMS main control unit automatically records all fault-related data, generates a tracing report, and uploads it to the fleet management platform to facilitate fault tracing and maintenance optimization.

[0065] During fault tracing, the main control unit immediately starts the data recording function when the alarm is triggered. The recorded data includes the fault occurrence time, the collected raw data, the calculated characteristic data, the fault prediction results, the alarm level and the handling measures, etc. All data are stored in the storage module and simultaneously uploaded to the fleet management platform through the remote communication module.

[0066] The source tracing report is automatically generated by the main control unit. It contains all the above data, adopts a standardized format, which makes it easy for managers to view and analyze. It can quickly trace the cause of the fault, optimize the maintenance plan, and reduce the occurrence of similar faults.

[0067] S6: Data update. After the fault handling is completed, the heavy truck's on-board BMS main control unit automatically clears the alarm records, updates the battery cell parameters and prediction thresholds, and restarts the fault prediction mechanism to adapt to the subsequent driving conditions of the heavy truck.

[0068] During data updates, once maintenance personnel have completed fault handling, they issue a completion command through the operating terminal in the heavy truck cab. Upon receiving the command, the main control unit automatically clears the alarm records in the storage module. Simultaneously, based on the actual cell status data obtained during fault handling, it updates the cell parameters and preset prediction thresholds to ensure that the prediction thresholds match the current cell status. Afterward, it restarts the fault prediction mechanism to continue real-time monitoring of the power battery pack, achieving continuous safety protection.

[0069] The method provided in this application solves five major problems in the fault prediction and alarm of existing heavy-duty truck on-board BMS, and constructs a full-process fault management system adapted to the specific scenarios of heavy-duty trucks. This method, through S1 adaptive cycle data acquisition, breaks the limitations of fixed acquisition modes and can dynamically adjust the acquisition frequency according to the heavy-duty truck's driving conditions, ensuring both the timeliness and accuracy of data acquisition while avoiding resource waste caused by invalid acquisition.

[0070] The calculation of S2 heavy-duty truck-specific characteristic quantities focuses on the core battery status parameters and operating condition-related parameters during the operation of the heavy-duty truck. Compared with the existing single parameter judgment, it achieves full coverage of multi-dimensional data, providing comprehensive and accurate data support for subsequent fault prediction.

[0071] The S3 latent fault prediction process, combined with a heavy truck fault prediction model and a dual anti-interference and anti-shake mechanism, effectively solves the problems of existing technologies' inability to predict latent faults and susceptibility to interference due to post-event threshold protection. It can identify early signs of battery faults in advance, thus avoiding battery damage and safety accidents.

[0072] The S4 graded alarm and linkage mechanism establishes a three-level graded mechanism, which not only solves the pain points of no graded alarm and inability to link with the vehicle control system, but also realizes linkage with the fleet management platform to ensure the timeliness of fault handling.

[0073] The S5 fault tracing function automatically records all fault-related data and generates a tracing report, providing data support for fault repair and model optimization, and improving the maintainability of the heavy-duty truck BMS system.

[0074] In the S6 data update process, alarm records can be automatically cleared and parameters and thresholds updated to ensure that the system can continuously adapt to the subsequent driving conditions of heavy trucks, thereby improving the system's adaptability and long-term stability.

[0075] In summary, the method provided in this application realizes closed-loop management of fault prediction, alarm, handling, source tracing and optimization, which comprehensively improves the fault control capability of heavy truck on-board BMS system and adapts to the usage requirements of complex driving scenarios of heavy trucks.

[0076] In some examples, the raw data collected includes battery cell data, operating condition data, environmental data, and auxiliary data. After preliminary filtering, the raw data is stored in a local cache and simultaneously uploaded to the vehicle CAN bus and fleet management platform.

[0077] The raw data encompasses cell data, operating condition data, environmental data, and auxiliary data, comprehensively covering the core influencing factors in the operation of heavy-duty truck batteries. This solves the problem of incomplete data collection leading to inaccurate predictions in existing technologies. Simultaneously, after preliminary filtering, the raw data is both stored in a local cache and simultaneously uploaded to the vehicle's CAN bus and fleet management platform. This achieves both local data retention and remote synchronization, ensuring data security and providing multi-channel data support for subsequent fault prediction, coordinated response, and source tracing. This further enhances the system's data management capabilities and collaboration, laying a solid data foundation for the smooth operation of the entire fault prediction and alarm process.

[0078] For example, cell data includes single-cell voltage, internal resistance, and capacity decay rate, which are collected by the cell parameter acquisition submodule group in the acquisition module; operating condition data includes discharge rate, current, vehicle speed, and load, which are collected by the operating condition parameter acquisition submodule group in the acquisition module; environmental data includes battery pack and external temperature, and temperature change rate, which are collected by the environmental parameter acquisition submodule group in the acquisition module; auxiliary data includes timestamp, BMS operating status, and line status, which are collected synchronously by the acquisition module.

[0079] Preliminary filtering can be completed by the built-in filtering circuit of the acquisition module, which can filter out noise and interference signals in the raw data to ensure the accuracy of the data. The filtered raw data is stored in the local cache of the main control unit for quick retrieval, and simultaneously uploaded to the vehicle CAN bus and fleet management platform to achieve multi-terminal data synchronization, providing multi-channel data support for subsequent fault prediction, linkage handling and source tracing.

[0080] In some examples, in the latent fault prediction S3, time window filtering removes instantaneous glitches, and multi-cycle confirmation anti-jitter is used to avoid interference and misjudgment. Only when the fault prediction rules are met for N consecutive cycles and interference is eliminated can it be determined as a fault precursor of the corresponding type.

[0081] A dual anti-interference strategy is adopted, which combines time window filtering to eliminate instantaneous glitches and multi-cycle confirmation anti-jitter. Time window filtering can filter out abnormal data caused by instantaneous interference, avoiding misjudgments caused by instantaneous fluctuations. Multi-cycle confirmation anti-jitter only determines a fault precursor when the fault prediction rules are met for N consecutive cycles and interference is eliminated, further improving the accuracy of fault prediction, significantly reducing the false alarm rate and missed alarm rate, ensuring the reliability of fault prediction results, avoiding unnecessary maintenance due to false alarms, and preventing safety hazards caused by missed alarms, thereby improving the safety and stability of heavy truck battery operation.

[0082] For example, time window filtering can use a sliding window filtering algorithm. The window size can be adjusted according to the acquisition cycle, and can be set to 3-5 acquisition cycles, which can effectively filter out abnormal data caused by instantaneous interference. Multi-cycle confirmation anti-shake works by continuously monitoring the characteristic data of N consecutive cycles. If the data of all cycles meets the fault prediction rules and the influence of external factors such as electromagnetic interference and road condition fluctuations is excluded, it is determined to be a fault precursor. If only a few cycles meet the rules, it is determined to be an interference signal and fault prediction is not triggered. Through this dual anti-interference design, the false alarm rate and false negative rate are greatly reduced, and the reliability of fault prediction is improved.

[0083] In some examples, the three-level fault alarm classification includes: Level 1 warning only records and uploads; Level 2 alarm triggers audible and visual prompts, limits power, prohibits heavy loads, and reminds for nearby maintenance; Level 3 emergency alarm triggers continuous alarms, slows down the vehicle, cuts off power, and locks fault data.

[0084] Level 1 warning is suitable for minor abnormalities, avoiding unnecessary audio and visual prompts and vehicle restrictions, and ensuring the normal operation of heavy trucks; Level 2 alarm can promptly remind staff to handle the situation before the fault develops further, preventing the fault from worsening, while also taking into account the safety and economy of heavy truck operation; Level 3 emergency alarm can quickly take emergency measures in the event of a serious fault, minimizing the occurrence of safety accidents and protecting the safety of the vehicle, battery and personnel.

[0085] It can be explained that the Level 1 warning is applicable to minor, latent faults, which only record and upload data without affecting the normal operation of heavy trucks and avoid unnecessary maintenance costs; the Level 2 alarm is applicable to moderate fault precursors, which remind drivers to deal with the fault in time through audible and visual prompts and vehicle restrictions, and prevent the fault from worsening; the Level 3 emergency alarm is applicable to severe fault precursors, which maximizes the safety of personnel, vehicles and batteries through emergency response measures, and the locked fault data provides an important basis for subsequent maintenance.

[0086] In some examples, in data acquisition S1, the acquisition module collects raw data in real time. Heavy-duty trucks operate under complex and variable conditions, and the battery status also changes in real time. Real-time acquisition of raw data can promptly capture abnormal fluctuations in the battery, providing real-time data support for subsequent characteristic quantity calculations and latent fault prediction. This avoids missed fault detection or untimely prediction due to data acquisition delays, further improving the timeliness of fault prediction, ensuring early identification of fault precursors, allowing sufficient time for fault handling, and reducing the probability of battery damage and safety accidents.

[0087] For example, the acquisition module adopts a real-time acquisition mode, which can continuously capture the state changes of the power battery pack and the working conditions of the heavy truck. It is not affected by the driving state of the heavy truck. Whether it is under heavy load, climbing, driving at a constant speed or idling, it can realize the real-time acquisition of raw data, ensuring that abnormal fluctuations of the battery can be captured in time, providing real-time data support for subsequent characteristic quantity calculation and fault prediction, and avoiding fault omission or untimely prediction due to data acquisition delay.

[0088] In some examples, cell data includes single cell voltage, internal resistance, and capacity decay rate; operating condition data includes discharge rate, current, vehicle speed, and load; environmental data includes battery pack and external temperature, and temperature change rate; and auxiliary data includes timestamps, BMS operating status, and line status.

[0089] By refining the specific composition of cell data, operating condition data, environmental data, and auxiliary data, the data collection process becomes more standardized and accurate, avoiding omissions or redundancies. At the same time, the refined data types provide more accurate input for subsequent calculations of heavy-duty truck-specific features, ensuring the accuracy of feature calculations and thus improving the reliability of fault prediction. This further adapts to the complex driving conditions and battery operation requirements of heavy-duty trucks, providing stronger data support for the entire fault prediction and alarm process.

[0090] For example, the voltage of a single cell reflects the cell's charging state and health status, the internal resistance reflects the cell's aging degree, and the capacity decay rate reflects the cell's lifespan; the discharge rate and current reflect the battery's discharge intensity, while vehicle speed and load reflect the heavy truck's operating conditions, directly affecting the battery's working state; the battery pack and external temperature, and temperature change rate reflect the battery's working environment, significantly impacting battery performance and safety; the timestamp records the time of data acquisition, the BMS operating status reflects the system's own operating status, and the line status reflects the connection status of the data transmission lines. All data complement each other, comprehensively covering the core influencing factors of battery operation.

[0091] A battery fault prediction and alarm device for a heavy-duty truck on-board BMS system is provided, which is applied to the above-mentioned battery fault prediction and alarm method for a heavy-duty truck on-board BMS system. The battery fault prediction and alarm device for a heavy-duty truck on-board BMS system includes a data acquisition module, a main control unit, a storage module, an alarm module, a vehicle linkage module, and a remote communication module; the main control unit, storage module, alarm module, vehicle linkage module, and remote communication module are sequentially connected by signals.

[0092] The acquisition module can collect various raw data from the power battery pack according to an adaptive cycle, and transmit the raw data to the main control unit set in the heavy truck's on-board BMS via electrical connection. The calculated data after being processed by the main control unit is stored in the storage module. After feature comparison in the storage module, it is determined whether to trigger the alarm module. The alarm module can issue corresponding alarm information according to the determination result, and generate corresponding actions through the vehicle linkage module. The alarm information and calculated data are transmitted to the fleet management platform through the remote communication module.

[0093] The acquisition module can collect various raw data from the power battery pack according to an adaptive cycle, and transmit the raw data to the main control unit of the heavy truck's on-board BMS via electrical connection. The acquisition module is made of high-strength engineering plastic material, which has the characteristics of high temperature resistance, impact resistance, waterproofing and electromagnetic interference resistance. It can adapt to the complex driving environment and extreme high and low temperature environment of heavy trucks, and its shape can be flat and rectangular.

[0094] The data acquisition module has multiple data acquisition interfaces for connecting to the battery cells of the power battery pack, the operating condition sensors of the heavy truck, and the environmental sensors. The interfaces are waterproof to prevent rain and dust from entering, ensuring the stability of the data acquisition.

[0095] The acquisition module and the main control unit are connected by shielded wires and a plug-in snap-fit ​​connection, which facilitates installation and disassembly and reduces the impact of electromagnetic interference on data transmission. The acquisition module is located close to the power battery pack to shorten the data transmission distance and ensure that the raw data can be transmitted to the main control unit quickly and accurately.

[0096] As the core component of the device, the main control unit is responsible for receiving the raw data transmitted by the acquisition module, performing operations such as characteristic quantity calculation, fault prediction, and command issuance. The main control unit can use an industrial-grade microprocessor as the core chip. The chip model can be selected according to the heavy truck model and functional requirements. The outer shell is made of aluminum alloy, which has good heat dissipation and anti-electromagnetic interference capabilities, and can operate stably in the complex electromagnetic environment of heavy trucks.

[0097] The main control unit is cube-shaped and can have multiple interfaces on its surface, including a data acquisition interface connected to the data acquisition module, a storage interface connected to the storage module, an alarm interface connected to the alarm module, a linkage interface connected to the vehicle linkage module, and a communication interface connected to the remote communication module. Each interface can adopt a standardized design to facilitate connection with other modules.

[0098] The main control unit can be installed in the electrical control cabinet of the heavy truck cab to be in a dry and ventilated environment, so as to avoid the impact of high temperature and humidity on its operation. The signal connection between it and each module can be made by shielded wires or CAN bus to ensure the stability and timeliness of signal transmission.

[0099] The storage module is used to store the collected raw data, calculated feature data, heavy truck fault prediction model, preset thresholds, fault records and traceability reports, etc. The storage module can use solid-state drives, which can adapt to the bumpy driving environment of heavy trucks. Its shape is cuboid and can be directly installed next to the main control unit. It is connected to the main control unit through a high-speed data cable and can quickly respond to the read and write commands of the main control unit to ensure real-time storage and retrieval of data.

[0100] The storage capacity of the storage module can be selected according to actual needs, generally not less than 128GB, which can meet the needs of long-term data storage. It also supports automatic data overwrite function. When the storage capacity is insufficient, it will automatically overwrite the oldest historical data to ensure that the system can continue to operate stably.

[0101] The alarm module is used to issue corresponding audible and visual alarm signals according to the instructions of the main control unit to remind the driver of the fault. The alarm module is made of flame-retardant plastic material, which is waterproof and high temperature resistant. It can be round in shape and can be equipped with red and yellow indicator lights and a buzzer on the surface. The red indicator light is used for the third-level emergency alarm, the yellow indicator light is used for the second-level alarm, and the buzzer is used to emit an alarm prompt sound.

[0102] The alarm module can be installed on the dashboard of the heavy truck cab, close to the driver's line of sight, so that the driver can easily observe and hear the alarm prompts. It is connected to the main control unit through a wire and can quickly receive alarm commands transmitted by the main control unit. It issues corresponding audible and visual signals according to the alarm level. For example, no audible and visual prompts are issued when the alarm is level one; when the alarm is level two, the yellow indicator light flashes and the buzzer emits an intermittent alarm sound; when the alarm is level three, the red indicator light stays on and the buzzer emits a continuous sharp alarm sound.

[0103] The vehicle linkage module is used to receive instructions from the main control unit and link the heavy-duty truck vehicle control system to execute the corresponding handling strategy. The vehicle linkage module can adopt an industrial-grade interface module with a metal shell, which has good anti-electromagnetic interference capability. It can be flat and rectangular in shape and installed in the heavy-duty truck vehicle control box. It is connected to the main control unit and the vehicle control system via CAN bus.

[0104] The vehicle linkage module can convert the main control unit's processing commands into signals that the vehicle control system can recognize, controlling the heavy truck's power output, speed, power switch, etc., and realizing operations such as limiting power, prohibiting heavy loads, slowing down the vehicle, and cutting off power, ensuring that emergency measures can be taken quickly when a fault occurs, and preventing the fault from worsening and safety accidents from happening.

[0105] The remote communication module transmits fault alarm information, fault prediction results, and source tracing reports to the fleet management platform, while also receiving instructions from the platform. The module is a 4G / 5G dual-mode communication module with a metal casing, featuring anti-interference, stable signal transmission, waterproofing, and dustproofing. Its rectangular shape and installation on the top of the heavy truck facilitate signal reception. Connected to the main control unit via a wire, it enables real-time data transmission, ensuring fleet management personnel can remotely monitor the truck's battery operating status and fault conditions.

[0106] The remote communication module supports breakpoint resume function. When the communication signal is interrupted, it can cache data and continue uploading after the signal is restored, ensuring the integrity of data transmission.

[0107] The heavy-duty truck-mounted BMS system battery fault prediction and alarm device provided in this application solves the problems of high implementation cost, poor adaptability, and the need for additional hardware in existing technologies. The device constructs a complete hardware architecture through signal connections of a data acquisition module, a main control unit, a storage module, an alarm module, a vehicle linkage module, and a remote communication module. Each module has a clear division of labor and works collaboratively, enabling precise implementation of the various processes of the above-mentioned methods.

[0108] Meanwhile, the device requires no additional hardware and can be adapted to the existing heavy-duty truck BMS hardware architecture, reducing implementation costs and adaptation difficulties, and improving the device's practicality and economy. The device achieves linkage with the fleet management platform through the remote communication module and with the vehicle control system through the vehicle linkage module, ensuring that fault alarm information can be synchronized in a timely manner and that handling measures can be executed quickly, further improving the efficiency and safety of heavy-duty truck battery fault management.

[0109] In some examples, the acquisition module includes a cell parameter acquisition submodule group, an operating condition parameter acquisition submodule group, and an environmental parameter acquisition submodule group to acquire cell parameters, operating condition parameters, and environmental parameters, respectively.

[0110] The data acquisition module is divided into three sub-modules: cell parameter acquisition, operating condition parameter acquisition, and environmental parameter acquisition. Each sub-module is specifically responsible for acquiring its corresponding type of data, which avoids interference between different types of data acquisition and improves the accuracy of data acquisition. At the same time, this modular design makes the maintenance and upgrading of the acquisition module more convenient. The acquisition parameters of each sub-module can be adjusted according to different models of heavy trucks and different operating conditions, further improving the adaptability of the device and ensuring that it can accurately acquire various types of raw data, providing reliable data support for subsequent fault prediction and alarm.

[0111] For example, the acquisition module includes a cell parameter acquisition submodule group, an operating condition parameter acquisition submodule group, and an environmental parameter acquisition submodule group to acquire cell parameters, operating condition parameters, and environmental parameters respectively.

[0112] The cell parameter acquisition submodule can use high-precision voltage and internal resistance acquisition chips, which can accurately acquire the voltage, internal resistance and capacity decay rate of a single cell. Its acquisition interface is connected to each cell of the power battery pack.

[0113] The operating condition parameter acquisition submodule uses current, speed, and load sensors to acquire the discharge rate and current of the power battery pack, as well as the speed and load of the heavy truck. Its acquisition interface is connected to the operating condition sensors of the heavy truck.

[0114] The environmental parameter acquisition submodule uses a temperature sensor to collect the temperature and temperature change rate inside and outside the battery pack. Its acquisition interface is connected to the temperature sensor. The environmental parameter acquisition submodule can have multiple parts, and different numbers can be designed for different heavy truck models to ensure the accuracy of information collection.

[0115] In some examples, the storage module is equipped with corresponding heavy-duty truck fault prediction models and preset thresholds for different heavy-duty trucks. Different heavy-duty trucks have different vehicle models, operating conditions, and battery configurations, resulting in varying fault prediction requirements and parameter thresholds. This claim clarifies that the storage module can be configured with corresponding heavy-duty truck fault prediction models and preset thresholds for different heavy-duty trucks, enabling the device to flexibly adapt to different types of heavy-duty trucks without requiring large-scale modifications, thus reducing adaptation costs. Simultaneously, targeted model and threshold settings improve the accuracy of fault prediction, ensuring the device can accurately adapt to the operational needs of different heavy-duty trucks, further expanding the device's applicability.

[0116] For example, different heavy-duty trucks have different models, battery configurations, and driving conditions, resulting in different battery fault types and characteristics. Therefore, the storage module pre-stores multiple fault prediction models and preset thresholds adapted to different heavy-duty trucks. Each model and threshold is optimized for the specific characteristics of a particular heavy-duty truck, enabling accurate identification of early signs of battery faults in that type of heavy-duty truck. When the device is installed on different heavy-duty trucks, the main control unit can automatically identify the truck model and call the corresponding fault prediction model and preset threshold in the storage module. This eliminates the need for large-scale modifications to the device, reducing adaptation costs and expanding its applicability.

[0117] A method for predicting and alarming battery faults in a heavy-duty truck-mounted BMS system, comprising S1 to S6.

[0118] S1: Data Acquisition. The heavy-duty truck onboard BMS main control unit collects various raw data of the power battery pack through the acquisition module at an adaptive cycle. The acquisition cycle is adjusted according to the working conditions. S2: Calculation of Heavy-Duty Truck-Specific Characteristic Quantities. The heavy-duty truck onboard BMS main control unit processes the collected raw data and calculates the differential pressure characteristics, temperature difference characteristics, voltage change characteristics, internal resistance characteristics, and working condition correlation characteristics in real time, and outputs the calculated data.

[0119] S3: Latent fault prediction. The heavy truck onboard BMS main control unit calls the heavy truck fault prediction model, compares the calculated multi-dimensional feature data with the preset threshold and fault prediction rules, and removes instantaneous spikes through time window filtering and anti-shake confirmation through multiple cycles to avoid interference and misjudgment. Only when the fault prediction rules are met for N consecutive cycles and interference is eliminated can it be determined as a fault precursor of the corresponding type and the fault prediction result is output.

[0120] S4: Tiered Alarm and Linked Response: Based on the severity of the fault prediction results, three levels of fault alarms are triggered, linking the vehicle control system and fleet management platform to execute response strategies; S5: Fault Source Tracing: After a fault alarm is triggered, the heavy-duty truck's onboard BMS main control unit automatically records all fault-related data, generates a source tracing report, and uploads it to the fleet management platform for fault source tracing and maintenance optimization; S6: Data Update: After the fault is handled, the heavy-duty truck's onboard BMS main control unit automatically clears alarm records, updates cell parameters and prediction thresholds, and restarts the fault prediction mechanism to adapt to the subsequent driving conditions of the heavy-duty truck.

[0121] The dual anti-interference and anti-shake mechanism (time window filtering and multi-cycle confirmation anti-shake) is directly integrated into the S3 latent fault prediction stage, clarifying the specific execution process of the anti-interference mechanism and making the fault prediction process standardized and rigorous. By eliminating instantaneous glitches through time window filtering and avoiding interference-induced misjudgments through multi-cycle confirmation anti-shake, the false alarm rate and false negative rate can be effectively reduced, solving the pain points of existing technologies that are affected by operating conditions and electromagnetic interference. At the same time, this method retains the closed-loop design of the entire process, realizing complete control of fault prediction, alarm, handling, tracing and optimization. It not only ensures the reliability of fault prediction, but also ensures the timeliness and pertinence of fault handling, further improving the fault management capability of heavy truck on-board BMS system, adapting to the complex driving scenarios of heavy trucks, and avoiding battery damage and safety accidents.

[0122] In some embodiments, a battery fault prediction and alarm device for a heavy-duty truck-mounted BMS system includes a data acquisition module, a main control unit, a storage module, an alarm module, a vehicle linkage module, and a remote communication module. Each module is connected in sequence and uses materials and structures adapted to the complex environment of heavy-duty trucks.

[0123] The data acquisition module is made of high-temperature resistant and electromagnetic interference resistant engineering plastic material. It has a flat rectangular structure and is installed inside the side of the heavy truck battery pack. It integrates cell parameter acquisition sub-module group, operating condition parameter acquisition sub-module group and environmental parameter acquisition sub-module group. The acquisition interface adopts a waterproof design and is connected to the power battery pack, heavy truck operating condition sensor and temperature sensor through shielded wires. It is connected to the main control unit by plug-in buckle.

[0124] The main control unit uses an industrial-grade microprocessor as its core chip, has an aluminum alloy shell, and is cubic in structure. It is installed in the electrical control cabinet of the heavy truck cab and has multiple standardized interfaces on its surface, which are connected to other modules via CAN bus and shielded wires.

[0125] The storage module uses a 128GB solid-state drive, has a cuboid structure, is installed next to the main control unit, and is connected to the main control unit via a high-speed data cable.

[0126] The alarm module is made of flame-retardant plastic, is round, and is installed on the dashboard in the cab near the driver's line of sight. It has red and yellow indicator lights and a buzzer on its surface, and is connected to the main control unit through wires.

[0127] The vehicle linkage module adopts an industrial-grade interface module with a metal shell and a flat rectangular shape. It is installed in the heavy-duty truck's control box and is connected to the main control unit and the vehicle control system via a CAN bus.

[0128] The remote communication module is a 4G / 5G dual-mode module with a metal casing, a cuboid shape, and is installed on the top of the heavy truck. It is connected to the main control unit via wires.

[0129] The storage module pre-stores a fault prediction model and preset thresholds adapted to this pure electric heavy truck. The prediction thresholds are set according to the battery configuration and driving conditions of this heavy truck. The N value for N consecutive cycles is set to 4 cycles. The acquisition cycle is adaptively adjusted according to the working conditions: 1.5 seconds when under heavy load and climbing, 8 seconds when driving at a constant speed, and 10 seconds when idling.

[0130] In some embodiments, a method for predicting and alarming battery faults in a heavy-duty truck-mounted BMS system is provided.

[0131] S1: Data Acquisition. The heavy-duty truck's onboard BMS main control unit collects various raw data from the power battery pack through the acquisition module at an adaptive cycle. The cell parameter acquisition submodule of the acquisition module collects single cell voltage, internal resistance, and capacity decay rate; the operating condition parameter acquisition submodule collects discharge rate, current, vehicle speed, and load; the environmental parameter acquisition submodule collects battery pack and external temperature and temperature change rate; and auxiliary data acquisition includes timestamps, BMS operating status, and circuit status.

[0132] The collected data is initially filtered by the built-in filtering circuit of the acquisition module, then stored in the local cache of the main control unit, and simultaneously uploaded to the vehicle's CAN bus and the fleet management platform. When the heavy truck is in a climbing and heavy-load condition, the acquisition cycle is adjusted to 1.5 seconds to quickly capture changes in battery status; when the heavy truck is in a high-speed constant-speed driving condition, the acquisition cycle is adjusted to 8 seconds to avoid invalid acquisition.

[0133] S2: Heavy-duty truck-specific characteristic calculation. The heavy-duty truck's onboard BMS main control unit processes the collected raw data, calculates pressure difference characteristics, temperature difference characteristics, voltage change characteristics, internal resistance characteristics, and operating condition correlation characteristics in real time, and outputs the calculated data.

[0134] Among them, the voltage difference feature calculates the difference between the voltage of a single cell and the average voltage of the entire battery pack, as well as the maximum voltage difference of the entire battery pack; the temperature difference feature calculates the difference between the temperature of a single cell and the average temperature of the battery pack, the maximum temperature difference of the battery pack, and the temperature change rate; the voltage change feature calculates the voltage change rate of a single cell and the voltage surge / drop rate; the internal resistance feature calculates the real-time value of the internal resistance of a single cell and the internal resistance fluctuation amplitude; and the operating condition correlation feature calculates the duration of the discharge rate and the current fluctuation amplitude. All feature calculations are tailored to the driving conditions of this heavy truck to ensure the relevance and accuracy of the data.

[0135] S3: Latent Fault Prediction. The heavy-duty truck's onboard BMS main control unit calls the fault prediction model adapted to this heavy-duty truck in the storage module, compares the calculated multi-dimensional feature data with the preset threshold and fault prediction rules, and completes the prediction by combining the dual anti-interference and anti-shake mechanism and outputs the fault prediction result.

[0136] The time window filtering adopts a sliding window filtering algorithm with a window size of 4 acquisition cycles to eliminate instantaneous spikes; the multi-cycle confirmation anti-shake continuously monitors for 4 acquisition cycles, and only when 4 consecutive cycles meet the fault prediction rules and the influence of electromagnetic interference and road condition fluctuations is it determined to be a fault precursor of the corresponding type.

[0137] For example, when the fluctuation amplitude of the internal resistance of a single cell exceeds the preset threshold for four consecutive cycles, and electromagnetic interference is eliminated, it is determined to be a precursor to cell aging fault, and the corresponding fault prediction result is output.

[0138] S4: Tiered Alarm and Linked Response. Based on the severity of the fault prediction, a three-tiered fault alarm is triggered, linking the vehicle control system and fleet management platform to execute response strategies.

[0139] When a minor, latent fault is predicted (such as a slight fluctuation in the voltage of a single battery cell), a Level 1 warning is triggered. The main control unit only records the fault information and uploads it to the fleet management platform. The alarm module does not issue any audible or visual alerts, and the heavy truck continues to operate normally.

[0140] When a moderate fault precursor is detected (such as excessive fluctuation in the internal resistance of the battery cell), a level two alarm is triggered. The yellow indicator light on the alarm module flashes, the buzzer emits an intermittent beep, the main control unit links with the vehicle control system to limit the charging and discharging power to 70% of the rated power, prohibit heavy loads, and simultaneously display the location and type of the faulty battery cell on the instrument panel in the cab, reminding the driver to have it inspected at the nearest repair shop.

[0141] When a serious fault precursor is detected (such as a rapid increase in battery pack temperature exceeding a preset safety threshold), a level three emergency alarm is triggered. The red indicator light on the alarm module remains constantly lit, and the buzzer emits a continuous, sharp warning sound. The main control unit, in conjunction with the vehicle control system, gradually reduces the vehicle speed and brings it to a safe stop within 30 seconds. The charging and discharging relays are disconnected, the power battery output is shut off, the fault data is locked, and the fault information is simultaneously uploaded to the fleet management platform, triggering an emergency alert from the management personnel.

[0142] S5: Fault Origin Tracing. After a fault alarm is triggered, the heavy-duty truck's onboard BMS main control unit automatically records all fault-related data, including the fault occurrence time, collected raw data, calculated characteristic data, fault prediction results, alarm level, and handling measures, etc., forming a standardized origin tracing report, which is uploaded to the fleet management platform to facilitate managers in tracing the cause of the fault and optimizing maintenance plans.

[0143] For example, when a Level 3 emergency alarm is triggered, the source tracing report records in detail the battery pack temperature change curve, current and voltage data, and the heavy truck's operating conditions at the time of the fault, providing maintenance personnel with accurate fault location information.

[0144] S6: Data Update. After the fault handling is completed, the maintenance personnel issue a handling completion command through the cab operating terminal. The heavy-duty truck's on-board BMS main control unit automatically clears the alarm records in the storage module, updates the battery cell parameters and prediction thresholds based on the actual status data of the battery cells obtained during the fault handling process, and restarts the fault prediction mechanism to adapt to the subsequent driving conditions of the heavy-duty truck.

[0145] For example, after maintenance personnel replace an aging battery cell, the main control unit updates the voltage and internal resistance parameters of the cell and adjusts the corresponding prediction threshold to ensure the accuracy of subsequent fault prediction.

[0146] Specifically, in some embodiments, a method for predicting and alarming battery faults in a heavy-duty truck on-board BMS system includes data acquisition, calculation of heavy-duty truck-specific characteristic quantities, prediction of latent faults, graded alarms and linkage handling, fault tracing, and data updating.

[0147] Data acquisition: The heavy-duty truck's onboard BMS main control unit collects various raw data from the power battery pack via a data acquisition module at an adaptive cycle, which is adjusted according to operating conditions. The data acquisition module and the main control unit are connected via waterproof shielded cables using a plug-in snap-fit ​​connection. This ensures stable signal transmission, effectively resists interference from the strong electromagnetic environment of heavy-duty trucks, and facilitates future maintenance. Its installation location close to the power battery pack shortens data transmission distance, reduces signal loss, and further ensures the timeliness and accuracy of raw data acquisition.

[0148] For heavy-duty trucks, the onboard BMS main control unit processes the collected raw data, calculating pressure difference characteristics, temperature difference characteristics, voltage change characteristics, internal resistance characteristics, and operating condition correlation characteristics in real time, and outputs the calculated data. The main control unit is installed in the electrical control cabinet in the heavy-duty truck cab, in a dry and ventilated environment to avoid the impact of high temperature and humidity on its operation. It communicates with the acquisition module via a CAN bus, offering high transmission speed and strong anti-interference capabilities, enabling rapid reception and processing of raw data. A high-speed data cable connects it to the storage module, ensuring that the calculated multi-dimensional characteristic data can be stored in real time, providing complete data support for subsequent fault prediction.

[0149] For latent fault prediction, the heavy-duty truck's onboard BMS main control unit calls the heavy-duty truck fault prediction model, compares the calculated multi-dimensional feature data with preset thresholds and fault prediction rules, and completes the prediction by combining a dual anti-interference and anti-shake mechanism and outputs the fault prediction result.

[0150] In existing technologies, conventional BMS lacks a dedicated fault prediction model and mostly adopts the prediction logic of passenger car BMS, without considering the characteristics of heavy truck operating conditions and without an effective anti-interference mechanism, resulting in the inability to predict latent faults and a high rate of false alarms and missed alarms.

[0151] This method undergoes targeted optimization. The heavy-duty truck fault prediction model is based on the characteristics of common hidden faults in heavy-duty trucks (such as cell aging, precursors to thermal runaway, and voltage imbalance). It is generated through training on a large amount of actual heavy-duty truck operating data and can accurately identify fault characteristics specific to heavy-duty trucks, distinguishing it from existing general-purpose models. The model is stored in a storage module made of solid-state drive material. Compared with existing conventional mechanical hard drives, the storage module has the characteristics of large storage capacity, fast read and write speed, and strong shock resistance. It has a cuboid structure and small size, which can be directly installed next to the main control unit. It is connected to the main control unit through a data cable and can quickly respond to the main control unit's call commands, read model data and preset thresholds, and ensure prediction efficiency.

[0152] The dual anti-interference and anti-shake mechanism is the core solution to the high false alarm and false negative rates of existing technologies. It includes time window filtering and multi-cycle confirmation anti-shake: the time window filtering adopts a sliding window filtering algorithm, and the window size can be adjusted according to the acquisition cycle, generally set to 3-5 acquisition cycles. It can effectively eliminate instantaneous spike data caused by electromagnetic interference and road condition fluctuations, and avoid misjudgment caused by instantaneous fluctuations. The multi-cycle confirmation anti-shake continuously monitors data for multiple cycles. Only when the fault prediction rules are met for N consecutive cycles and the influence of external factors such as electromagnetic interference and road condition fluctuations is excluded, is it determined to be a fault precursor of the corresponding type. The value of N can be adjusted according to the heavy truck model and operating conditions. It is generally set to 3-5 cycles to ensure the accuracy of fault prediction.

[0153] During the prediction process, the main control unit receives the preset thresholds and fault prediction rules transmitted by the storage module in real time, compares them one by one with the calculated feature data, and completes the prediction by combining anti-interference and anti-shaking mechanisms. The prediction results are transmitted to the alarm module and the vehicle linkage module in real time, providing accurate basis for subsequent hierarchical alarm and linkage handling.

[0154] The system features tiered alarms and coordinated response. Based on the severity of the fault prediction, a three-tiered fault alarm is triggered, which in turn coordinates with the vehicle control system and fleet management platform to execute response strategies.

[0155] Level 1 warning only records and uploads information. It is applicable to minor, latent faults (such as slight fluctuations in the voltage of a single cell that do not affect the normal operation of the battery). The alarm module does not issue audible or visual prompts. The main control unit only records the fault information and uploads it to the fleet management platform. This does not affect the normal operation of heavy trucks, avoids unnecessary maintenance costs, and provides basic data for fleet managers to facilitate later traceability and optimization.

[0156] The level 2 alarm triggers audible and visual prompts, limits power, prohibits heavy loads, and reminds the driver to have the nearest repair shop. It is suitable for moderate early signs of faults (such as excessive fluctuations in cell internal resistance or abnormal local temperature differences in the battery pack). The yellow indicator light on the alarm module flashes, and the buzzer emits an intermittent alarm sound. At the same time, the main control unit links with the vehicle control system to limit the charging and discharging power of the heavy truck and prohibit heavy-load climbing to prevent the fault from worsening. The location of the faulty cell and the fault type are displayed on the instrument panel in the cab, reminding the driver to find the nearest repair shop and reduce operational interruptions caused by the fault.

[0157] The Level 3 emergency alarm triggers continuous warnings, slows the vehicle to a stop, cuts off power, and locks fault data. It is suitable for severe fault precursors (such as rapid rise in battery pack temperature, sudden drop in voltage of a single cell, posing a risk of thermal runaway). The red indicator light on the alarm module stays on, and the buzzer emits a continuous, sharp warning sound. At the same time, the main control unit links with the vehicle control system to gradually reduce the vehicle speed and bring it to a safe stop, disconnecting the charging and discharging relays and shutting off the power battery output to minimize the risk of accidents. The locked fault data provides important information for subsequent maintenance and fault tracing.

[0158] The alarm module is made of high-temperature resistant, waterproof, and flame-retardant plastic material. Compared with existing ordinary plastic alarm modules, it has a longer service life and stronger adaptability. It is round in shape and has indicator lights and a buzzer on its surface. It is installed on the dashboard of the heavy truck cab, making it easy for the driver to see and hear the alarm prompts. It is connected to the main control unit through a wire and can quickly receive alarm commands transmitted by the main control unit and issue corresponding alarm signals.

[0159] The vehicle control system and the main control unit are linked via a CAN bus, enabling the system to quickly receive commands from the main control unit and execute corresponding vehicle control operations. The fleet management platform and the main control unit synchronize data via a remote communication module. This module is a 4G / 5G dual-mode module with a metal casing, featuring anti-interference and stable signal transmission. Installed on the top of the heavy truck, it facilitates the reception of communication signals. Connected to the main control unit via a wire, it can upload fault alarm information and prediction results to the fleet management platform in real time, enabling remote monitoring and advance maintenance by management personnel. This solves the problems of unsynchronized fault information and untimely handling in existing technologies.

[0160] For fault tracing, after a fault alarm is triggered, the heavy-duty truck's onboard BMS main control unit automatically records all fault-related data, generates a tracing report, and uploads it to the fleet management platform to facilitate fault tracing and maintenance optimization.

[0161] Upon triggering the alarm, the main control unit immediately activates the data recording function, recording comprehensive and detailed data, including the time of the fault occurrence, the collected raw data (cell parameters, operating parameters, environmental parameters, etc.), the calculated characteristic data, the fault prediction results, the alarm level and handling measures, etc. All data is stored in the storage module and simultaneously uploaded to the fleet management platform through the remote communication module to ensure that the data is not lost.

[0162] The source tracing report is automatically generated by the main control unit, adopts a standardized format, and includes all the above data. It is convenient for managers and maintenance personnel to view and analyze, and can quickly trace the cause of the fault, formulate targeted maintenance plans, and optimize the fault prediction model and preset thresholds based on the source tracing data to reduce the occurrence of similar faults and improve the stability of heavy truck battery operation.

[0163] After the data is updated and the fault is handled, the heavy-duty truck's onboard BMS main control unit automatically clears the alarm records, updates the battery cell parameters and prediction thresholds, and restarts the fault prediction mechanism to adapt to the subsequent driving conditions of the heavy-duty truck.

[0164] After maintenance personnel complete the fault handling, they issue a completion command through the operating terminal in the heavy truck cab. Upon receiving the command, the main control unit automatically clears the alarm records in the storage module, eliminating the need for manual operation, thus improving efficiency and reducing errors. Simultaneously, based on the actual cell status data obtained during the fault handling process (such as voltage and internal resistance parameters after replacing aging cells), the unit automatically updates the cell parameters and preset prediction thresholds to ensure that the prediction thresholds match the current cell status. Afterward, the fault prediction mechanism is restarted to continue real-time monitoring of the power battery pack, achieving continuous safety protection and adapting to different driving conditions of the heavy truck, further improving the accuracy of fault prediction.

[0165] For example, the raw data collected includes battery cell data, operating condition data, environmental data, and auxiliary data. After preliminary filtering, the raw data is stored in a local cache and simultaneously uploaded to the vehicle CAN bus and fleet management platform.

[0166] The types of data collected are comprehensive, covering four major categories of core data, making up for the shortcomings of existing technologies.

[0167] The cell data includes the voltage, internal resistance, and capacity decay rate of a single cell. It is collected by the cell parameter acquisition submodule in the acquisition module and can reflect the health status, aging degree, and service life of the cell.

[0168] Operating condition data, including discharge rate, current, vehicle speed, and load, is collected by the operating condition parameter acquisition submodule in the acquisition module, which can reflect the impact of heavy truck driving conditions on battery status.

[0169] Environmental data, including battery pack and external temperature and temperature change rate, is collected by the environmental parameter acquisition submodule group in the acquisition module. It can reflect the battery's working environment, and high and low temperature environments are important factors affecting battery performance and safety.

[0170] Auxiliary data includes timestamps, BMS operating status, and line status, which are collected synchronously by the acquisition module. The timestamps are used to record the time of data acquisition, which is convenient for fault tracing. The BMS operating status reflects the working status of the system itself, and the line status reflects the connection status of the data transmission line, which avoids abnormal data acquisition due to line failure.

[0171] The initial filtering is performed by the built-in filtering circuit of the acquisition module, which can filter out noise and interference signals in the raw data to ensure the accuracy of the data. The filtered raw data is stored in the local cache of the main control unit for quick retrieval, and simultaneously uploaded to the vehicle CAN bus and fleet management platform to achieve multi-terminal data synchronization. This provides multi-channel data support for subsequent fault prediction, linkage handling and source tracing, and further improves the reliability of the system.

[0172] For example, in the prediction of latent faults, time window filtering removes instantaneous spikes, and multi-cycle confirmation anti-jitter is used to avoid interference and misjudgment. Only when the fault prediction rules are met for N consecutive cycles and interference is eliminated can it be determined as a fault precursor of the corresponding type.

[0173] Existing conventional BMS lacks effective anti-interference and anti-shake mechanisms, relying solely on data from a single cycle to determine faults. This makes them susceptible to external factors such as electromagnetic interference and road condition fluctuations, leading to high false alarm and missed alarm rates. The method provided in this application employs a sliding window filtering algorithm for time window filtering. The window size can be adjusted according to the acquisition cycle, typically set to 3-5 acquisition cycles. This effectively filters out abnormal data caused by instantaneous interference, avoiding misjudgments due to instantaneous fluctuations. Multi-cycle confirmation anti-shake involves continuously monitoring characteristic data over N consecutive cycles. Only if the data from all cycles meets the fault prediction rules and the influence of external factors such as electromagnetic interference and road condition fluctuations is the fault precursor determined. If only a few cycles meet the rules, it is considered an interference signal and fault prediction is not triggered. This dual anti-interference design significantly reduces false alarm and missed alarm rates, substantially improving the reliability of fault prediction and adapting to the complex electromagnetic environment and driving conditions of heavy trucks.

[0174] For example, the three-level classification of fault alarms includes: Level 1 warning only records and uploads; Level 2 alarm triggers audible and visual prompts, limits power, prohibits heavy loads, and reminds for nearby maintenance; Level 3 emergency alarm triggers continuous alarms, slows down the vehicle, cuts off power, and locks fault data.

[0175] The three-level alarm design is tailored to the actual needs of heavy-duty truck operation. Level 1 warning is suitable for minor, latent faults, only recording and uploading data without affecting the normal operation of the heavy-duty truck, avoiding unnecessary maintenance costs. Level 2 alarm is suitable for moderate fault precursors, reminding drivers to handle the situation in a timely manner through audible and visual prompts and vehicle restrictions, preventing the fault from worsening, and balancing operational efficiency and safety. Level 3 emergency alarm is suitable for severe fault precursors, maximizing the safety of personnel, vehicles, and batteries through emergency response measures. The locked fault data provides important information for subsequent maintenance, achieving a balance between safety and efficiency.

[0176] For example, in the data acquisition phase, the acquisition module collects raw data in real time. The acquisition module adopts a real-time acquisition mode, continuously capturing changes in the state of the power battery pack and the operating conditions of the heavy truck. It is unaffected by the truck's driving status; whether under heavy load, climbing, constant speed, or idling, it can achieve real-time acquisition of raw data, ensuring timely detection of abnormal battery fluctuations. This provides real-time data support for subsequent characteristic quantity calculations and fault prediction, avoiding missed fault detections or untimely predictions due to data acquisition delays, and further improving the system's safety and reliability.

[0177] For example, cell data includes single cell voltage, internal resistance, and capacity decay rate; operating condition data includes discharge rate, current, vehicle speed, and load; environmental data includes battery pack and external temperature, and temperature change rate; auxiliary data includes timestamp, BMS operating status, and line status.

[0178] Various data complement each other, comprehensively covering the core influencing factors of battery operation: single cell voltage reflects the cell's charging and health status, internal resistance reflects the cell's aging degree, and capacity decay rate reflects the cell's lifespan; discharge rate and current reflect the battery's discharge intensity, vehicle speed and load reflect the heavy truck's operating conditions, directly affecting the battery's working status; battery pack and external temperature and temperature change rate reflect the battery's working environment, significantly impacting battery performance and safety; timestamps record the time of data acquisition, BMS operating status reflects the system's own operating status, and line status reflects the connection status of data transmission lines. Through multi-dimensional data acquisition, comprehensive and accurate data support is provided for fault prediction, addressing the pain point of incomplete data acquisition in existing technologies.

[0179] In the description of this specification, specific features, structures, materials, or characteristics may be combined in any suitable manner in one or more embodiments or examples.

[0180] The above description is merely a specific embodiment of this application, but the scope of protection of this application is not limited thereto. Any changes or substitutions within the technical scope disclosed in this application should be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for predicting and alarming battery faults in a heavy-duty truck onboard BMS system, characterized in that, S1: Data acquisition. The heavy-duty truck's onboard BMS main control unit collects various raw data from the power battery pack through the acquisition module at an adaptive cycle. The acquisition cycle is adjusted according to the working conditions. S2: Calculation of heavy-duty truck-specific characteristic quantities. The heavy-duty truck on-board BMS main control unit processes the collected raw data, calculates pressure difference characteristics, temperature difference characteristics, voltage change characteristics, internal resistance characteristics and operating condition correlation characteristics in real time, and outputs the calculated data. S3: Latent fault prediction. The heavy truck on-board BMS main control unit calls the heavy truck fault prediction model, compares the calculated multi-dimensional feature data with the preset threshold and fault prediction rules, and completes the prediction by combining the dual anti-interference and anti-shake mechanism and outputs the fault prediction result. S4: Tiered alarm and linkage response. Based on the severity of the fault prediction results, a three-tiered fault alarm is triggered, which is linked with the vehicle control system and fleet management platform to execute the response strategy. S5: Fault tracing. After a fault alarm is triggered, the heavy truck onboard BMS main control unit automatically records all fault-related data, generates a tracing report, and uploads it to the fleet management platform to facilitate fault tracing and maintenance optimization. S6: Data update. After the fault handling is completed, the heavy truck on-board BMS main control unit automatically clears the alarm records, updates the battery cell parameters and prediction thresholds, and restarts the fault prediction mechanism to adapt to the subsequent driving conditions of the heavy truck.

2. The method for predicting and alarming battery faults in a heavy-duty truck-mounted BMS system according to claim 1, characterized in that, The raw data collected includes battery cell data, operating condition data, environmental data, and auxiliary data. After preliminary filtering, the raw data is stored in a local cache and simultaneously uploaded to the vehicle CAN bus and fleet management platform.

3. The method for predicting and alarming battery faults in a heavy-duty truck-mounted BMS system according to claim 1, characterized in that, In the latent fault prediction S3, time window filtering removes instantaneous spikes, and multi-cycle confirmation anti-jitter is used to avoid interference and misjudgment. Only when the fault prediction rules are met for N consecutive cycles and interference is eliminated can it be determined as a fault precursor of the corresponding type.

4. The method for predicting and alarming battery faults in a heavy-duty truck-mounted BMS system according to any one of claims 1 to 3, characterized in that, The three-level classification of the fault alarm includes: Level 1 alerts only record and upload data; The level 2 alarm triggers an audible and visual alert, limits power, prohibits overload, and reminds the nearest maintenance personnel to perform maintenance. A Level 3 emergency alarm triggers continuous alerts, slows the vehicle to a stop, cuts off power, and simultaneously locks fault data.

5. The method for predicting and alarming battery faults in a heavy-duty truck-mounted BMS system according to claim 1, characterized in that, In the data acquisition S1, the acquisition module acquires raw data in real time.

6. The method for predicting and alarming battery faults in a heavy-duty truck-mounted BMS system according to claim 1, characterized in that, The cell data includes single-cell voltage, internal resistance, and capacity decay rate. The operating data includes discharge rate, current, vehicle speed, and load; The environmental data includes battery pack and external temperature, and temperature change rate; The auxiliary data includes timestamps, BMS operating status, and line status.

7. A heavy truck-mounted BMS system battery fault prediction and alarm device, characterized in that, The method for predicting and alarming battery faults in a heavy-duty truck-mounted BMS system as described in claims 1-6 includes a data acquisition module, a main control unit, a storage module, an alarm module, a vehicle linkage module, and a remote communication module. The main control unit, the storage module, the alarm module, the vehicle linkage module, and the remote communication module are sequentially connected by signals; The acquisition module can collect various raw data of the power battery pack at an adaptive cycle, and transmit the raw data to the main control unit set in the heavy truck's on-board BMS via an electrical connection. The calculated data after being processed by the main control unit is stored in the storage module. After feature comparison in the storage module, it is determined whether to trigger the alarm module. The alarm module can issue corresponding alarm information according to the determination result, and generate corresponding actions through the vehicle linkage module. The alarm information and calculated data are transmitted to the fleet management platform through the remote communication module.

8. The heavy-duty truck on-board BMS system battery fault prediction and alarm device according to claim 7, characterized in that, The acquisition module includes a cell parameter acquisition submodule group, an operating condition parameter acquisition submodule group, and an environmental parameter acquisition submodule group, to acquire cell parameters, operating condition parameters, and environmental parameters respectively.

9. The heavy-duty truck on-board BMS system battery fault prediction and alarm device according to claim 7, characterized in that, Depending on the type of heavy-duty truck, the storage module is equipped with a corresponding heavy-duty truck fault prediction model and a corresponding preset threshold.

10. A method for predicting and alarming battery faults in a heavy-duty truck onboard BMS system, characterized in that, S1: Data acquisition. The heavy-duty truck's onboard BMS main control unit collects various raw data from the power battery pack through the acquisition module at an adaptive cycle. The acquisition cycle is adjusted according to the working conditions. S2: Calculation of heavy-duty truck-specific characteristic quantities. The heavy-duty truck on-board BMS main control unit processes the collected raw data, calculates pressure difference characteristics, temperature difference characteristics, voltage change characteristics, internal resistance characteristics and operating condition correlation characteristics in real time, and outputs the calculated data. S3: Latent Fault Prediction. The heavy-duty truck onboard BMS main control unit calls the heavy-duty truck fault prediction model, compares the calculated multi-dimensional feature data with preset thresholds and fault prediction rules, and removes instantaneous spikes through time window filtering and performs multi-cycle confirmation and anti-shake to avoid interference and misjudgment. Only when the fault prediction rules are met for N consecutive cycles and interference is eliminated can it be determined as a fault precursor of the corresponding type and the fault prediction result be output. S4: Tiered alarm and linkage response. Based on the severity of the fault prediction results, three levels of fault alarms are triggered, and the vehicle control system and fleet management platform are linked to execute the response strategy. S5: Fault tracing. After a fault alarm is triggered, the heavy truck onboard BMS main control unit automatically records all fault-related data, generates a tracing report, and uploads it to the fleet management platform to facilitate fault tracing and maintenance optimization. S6: Data update. After the fault handling is completed, the heavy truck on-board BMS main control unit automatically clears the alarm records, updates the battery cell parameters and prediction thresholds, and restarts the fault prediction mechanism to adapt to the subsequent driving conditions of the heavy truck.

Images