Battery power shortage detection method and system of vehicle, vehicle and storage medium

Abstract

The embodiment of the invention provides a vehicle battery power shortage detection method and system, a vehicle and a storage medium. The method comprises the steps that battery state data of a battery in the vehicle, power consumption data of electrical equipment in the vehicle, driving state data of the vehicle and environment data of the environment where the vehicle is located are collected; performing state detection on the battery based on the power lack detection condition and the battery state data by using a first prediction model to obtain a state detection result; using a second prediction model to carry out power shortage detection on the battery based on the power consumption data, the driving state data and the environment data to obtain a power shortage detection result; and in response to the power shortage detection result that the battery has the power shortage risk, generating power shortage prompt information based on the battery state data, the power consumption data, the driving state data and the environment data. According to the invention, the technical problem of low prediction efficiency of the power shortage of the battery in the prior art is solved.

Description

Technical Field

[0001] This application relates to the fields of vehicles and batteries, and more specifically, to a method, system, vehicle, and storage medium for detecting battery depletion in a vehicle. Background Technology

[0002] In recent years, the rapid development of intelligent vehicle technology has driven innovation in the automotive industry, with new energy vehicles leading the way. New energy vehicles, especially pure electric vehicles and plug-in hybrid vehicles, not only rely on efficient power management systems but also require an efficient battery health monitoring and prediction system to ensure their operational stability and safety. However, in terms of battery depletion prediction, related technologies face numerous challenges, resulting in low prediction efficiency.

[0003] There is currently no good solution to the above problems. Summary of the Invention

[0004] This application provides a method for detecting battery depletion in a vehicle, thereby addressing at least the technical problem of low prediction efficiency for battery depletion in related technologies.

[0005] According to one aspect of the embodiments of this application, a method for detecting battery depletion in a vehicle is provided, comprising: in response to receiving a vehicle depletion detection command, collecting battery state data of the vehicle's battery, power consumption data of electrical devices in the vehicle, driving state data of the vehicle, and environmental data of the environment in which the vehicle is located; performing a state detection on the battery based on depletion detection conditions and battery state data using a first prediction model to obtain a state detection result, wherein the state detection result is used to indicate whether the battery meets the depletion detection conditions; in response to the state detection result indicating that the battery meets the depletion detection conditions, performing a depletion detection on the battery based on power consumption data, driving state data, and environmental data using a second prediction model to obtain a depletion detection result, wherein the depletion detection result is used to indicate whether the battery is at risk of depletion; and in response to the depletion detection result indicating that the battery is at risk of depletion, generating a depletion warning message based on the battery state data, power consumption data, driving state data, and environmental data.

[0006] Furthermore, based on battery status data, power consumption data, driving status data, and environmental data, a low battery warning message is generated, including: locating the low battery risk based on battery status data, power consumption data, driving status data, and environmental data, and determining the risk type corresponding to the low battery risk; quantifying the low battery risk based on the risk type to obtain the battery risk level; and generating a low battery warning message based on the risk level and risk type.

[0007] Furthermore, based on the risk level and risk type, a low battery warning message is generated, including: obtaining vehicle scenario data, wherein the vehicle scenario data is used to represent vehicle usage data under different scenarios; determining the low battery warning scenario based on the vehicle scenario data and risk type; and generating the low battery warning message based on the low battery warning scenario, risk type, and risk level.

[0008] Furthermore, the method also includes: sorting the power consumption of multiple devices corresponding to the vehicle to obtain a power consumption sorting result, wherein different devices are used to implement different functions of the vehicle; determining at least one target device based on the power consumption sorting result and generating control instructions for at least one target device; sending the control instructions to at least one target device and receiving control results from at least one target device, wherein the control results are used to indicate whether at least one target device has been successfully turned off.

[0009] Furthermore, in response to the battery depletion detection result indicating a risk of battery depletion, the method further includes: acquiring the vehicle's current location information and / or vehicle navigation information; determining the target service point's service point information based on the current location information and / or vehicle navigation information, wherein the target service point is used to provide battery repair services; and generating a navigation path to the target service point in response to receiving a confirmation instruction for the service point information.

[0010] Furthermore, the method also includes: acquiring sample voltage data, sample current data, and sample temperature data of the first sample battery; constructing first sample state data and sample state detection results of the first sample battery based on the sample voltage data, sample current data, and sample temperature data; inputting the first sample state data into a first initial prediction model, using the first initial prediction model to perform state detection on the first sample battery, and obtaining a predicted state detection result; and adjusting the model parameters of the first initial prediction model based on the sample state detection result and the predicted state detection result to obtain a first prediction model.

[0011] Furthermore, the method also includes: acquiring the sample battery internal resistance, sample electrolyte state, sample driving data, and sample environment data of the second sample battery, wherein the sample driving data is the driving data of the vehicle where the second sample battery is located, and the sample environment data is the environmental data of the vehicle where the second sample battery is located; constructing second sample state data and sample depletion detection results of the second sample battery based on the sample battery internal resistance, sample electrolyte state, sample driving data, and sample environment data; inputting the second sample state data into the second initial prediction model, using the second initial prediction model to perform depletion detection on the second sample battery, and obtaining the predicted depletion detection result; adjusting the model parameters of the second initial prediction model based on the sample depletion detection result and the predicted depletion detection result, and obtaining the second prediction model.

[0012] According to another aspect of the embodiments of this application, a vehicle battery depletion detection system is also provided, comprising: a data acquisition module, which, in response to receiving a vehicle depletion detection command, acquires battery state data of the vehicle's battery, power consumption data of electrical devices in the vehicle, vehicle driving status data, and environmental data of the vehicle's environment; a first detection module, deployed locally in the vehicle, which uses a first prediction model to perform state detection on the battery based on depletion detection conditions and battery state data, and obtains a state detection result, wherein the state detection result is used to indicate whether the battery meets the depletion detection conditions; in response to the state detection result indicating that the battery meets the depletion detection conditions, sending power consumption data, driving status data, and environmental data to a second detection module; a second detection module, deployed in the cloud corresponding to the vehicle, which uses a second prediction model to perform depletion detection on the battery based on power consumption data, driving status data, and environmental data, and obtains a depletion detection result, wherein the depletion detection result is used to indicate whether the battery is at risk of depletion; and a prompting module, which, in response to the depletion detection result indicating that the battery is at risk of depletion, generates a depletion prompt message based on battery state data, power consumption data, driving status data, and environmental data.

[0013] According to another aspect of the embodiments of this application, a vehicle is also provided, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods in various embodiments of this application when it runs.

[0014] According to another aspect of the embodiments of this application, a computer-readable storage medium is also provided, the computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of this application.

[0015] According to another aspect of the embodiments of this application, a computer program product is also provided, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0016] According to another aspect of the embodiments of this application, a computer program product is also provided, including a non-volatile computer-readable storage medium storing a computer program, which, when executed by a processor, implements the methods in various embodiments of this application.

[0017] According to another aspect of the embodiments of this application, a computer program is also provided, which, when executed by a processor, implements the methods of the various embodiments of this application.

[0018] In this embodiment, in response to receiving a low battery detection command from the vehicle, battery status data of the vehicle's battery, power consumption data of the vehicle's electrical devices, vehicle driving status data, and environmental data of the vehicle's environment are collected. A first prediction model is used to perform a battery status detection based on the low battery detection conditions and battery status data to obtain a status detection result, wherein the status detection result is used to indicate whether the battery meets the low battery detection conditions. In response to the status detection result indicating that the battery meets the low battery detection conditions, a second prediction model is used to perform a low battery detection based on power consumption data, driving status data, and environmental data to obtain a low battery detection result, wherein the low battery detection result is used to indicate whether the battery is at risk of low battery. In response to the low battery detection result indicating that the battery is at risk of low battery, a low battery warning message is generated based on the battery status data, power consumption data, driving status data, and environmental data. By comprehensively analyzing the state detection results of the first prediction model and the low-power detection results of the second prediction model, and further combining the changing trends of power consumption data, the energy consumption patterns under driving status data, and the degree of influence of environmental data on battery performance, the specific factors leading to low-power risk and the severity level of the risk can be efficiently and accurately derived. This achieves the technical effect of improving the prediction efficiency of battery low-power, and thus solves the technical problem of low prediction efficiency of battery low-power in related technologies. Attached Figure Description

[0019] The accompanying drawings, which are included to provide a further understanding of this application and form part of this application, illustrate exemplary embodiments and are used to explain this application, but do not constitute an undue limitation of this application. In the drawings:

[0020] Figure 1 This is a flowchart of a battery depletion detection method for a vehicle according to an embodiment of this application;

[0021] Figure 2 This is a schematic diagram of the structure of a vehicle battery depletion prediction system according to an embodiment of this application;

[0022] Figure 3 This is a schematic diagram of a vehicle battery depletion detection device according to an embodiment of this application;

[0023] Figure 4 This is a schematic diagram of a vehicle battery depletion detection system according to an embodiment of this application. Detailed Implementation

[0024] To enable those skilled in the art to better understand the present application, the technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present application, and not all embodiments. Based on the embodiments in the present application, all other embodiments obtained by those of ordinary skill in the art without creative effort should fall within the scope of protection of the present application.

[0025] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such data can be interchanged where appropriate so that the embodiments of this application described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0026] According to an embodiment of this application, a method for detecting battery depletion in a vehicle is provided. It should be noted that the steps shown in the flowchart in the accompanying drawings can be executed in a computer system such as a set of computer-executable instructions. Furthermore, although a logical order is shown in the flowchart, in some cases, the steps shown or described may be executed in a different order than that shown here.

[0027] This embodiment provides a method for detecting low battery charge in a vehicle. Figure 1 This is a flowchart of a vehicle battery depletion detection method according to an embodiment of this application, such as... Figure 1 As shown, the process includes the following steps:

[0028] Step S102: In response to receiving the vehicle's low battery detection command, collect battery status data of the vehicle's battery, power consumption data of the vehicle's electrical equipment, vehicle driving status data, and environmental data of the vehicle's environment.

[0029] The battery status data mentioned above refers to various parameters reflecting the current operating status of the battery, including but not limited to battery voltage, current, temperature, and internal resistance. This data forms the basis for assessing battery health and diagnosing the risk of battery depletion.

[0030] The power consumption data of the aforementioned electrical devices covers energy consumption information of different electrical devices in the vehicle during operation or standby. Power consumption data includes real-time power, cumulative power consumption, start / stop status, and operating mode of the electrical devices, and is a key indicator for judging abnormal power consumption of electrical devices and indirectly affecting battery life.

[0031] The aforementioned driving status data refers to the dynamic information of the vehicle during driving, including but not limited to vehicle speed, acceleration, mileage, and road condition type (such as urban congestion, highway driving, steep mountain slopes, etc.). Driving status data helps to understand the vehicle's real-time energy consumption and the dynamic impact state of the battery, providing an important basis for analyzing the risk of battery depletion under the influence of multiple factors.

[0032] The environmental data mentioned above refers to the vehicle's external environmental conditions and natural factors that may affect the vehicle and battery, including temperature, humidity, light intensity, and atmospheric pressure. Environmental data has a significant impact on battery performance, especially under extreme climatic conditions, and can help the system assess the potential impact of environmental factors on battery life and charge / discharge efficiency.

[0033] Upon receiving a low battery detection command from the vehicle, the onboard system activates and begins collecting data on the vehicle's battery status, power consumption of electrical devices, driving status, and environmental conditions. This step serves as the data input phase for the entire warning system, ensuring comprehensive information about the battery's status is obtained from multiple perspectives.

[0034] In response to a low battery detection command from the vehicle, the system collects battery status data, power consumption data of electrical devices, vehicle driving status data, and environmental data of the vehicle's surroundings. By immediately initiating the data collection process upon receiving the low battery detection command, the system can quickly acquire comprehensive vehicle information, including battery status, electrical power consumption, driving conditions, and environmental conditions, thereby enabling early warning of low battery risk. This is particularly important for new energy vehicles, as modern vehicles are equipped with numerous complex electronic devices, and the health of the battery directly affects the vehicle's availability and safety. Early detection and prevention can avoid sudden malfunctions, reduce unnecessary repair costs, and improve the user experience.

[0035] The steps described above encompass the collection of battery status data, electrical device power consumption data, vehicle driving status data, and environmental data. This significantly expands the data scope of the early warning system, enabling a comprehensive assessment of battery health from multiple perspectives. Compared to traditional methods that rely solely on battery voltage, this multi-dimensional data collection approach can more accurately identify various complex factors leading to battery depletion, such as abnormal power consumption of electrical devices, high energy consumption caused by frequent vehicle start-stop cycles, and the impact of ambient temperature changes on battery performance. This improves the accuracy and reliability of the early warning system.

[0036] The collected multi-dimensional data not only helps the early warning system to accurately assess the risk of battery depletion, but also provides a basis for further management decisions. For example, based on the power consumption data of electrical equipment, the system can identify which devices are consuming power abnormally, and thus propose targeted energy-saving suggestions; combined with vehicle driving status and environmental data, the system can assess the impact of different driving modes and environmental conditions on the battery, and provide personalized battery use and maintenance recommendations. This not only improves the practicality of the early warning system, but also promotes refined battery management and extends its lifespan.

[0037] The vehicle's intelligent management system utilizes pre-installed sensors and data acquisition modules, covering the battery management system, electrical control units, navigation system, and environmental monitoring sensors. The data acquisition process is typically automated, requiring no manual driver intervention, ensuring real-time and continuous data transmission. For example, when the vehicle is parked, the intelligent system periodically checks battery voltage, temperature, and internal resistance, while also monitoring the power consumption of in-vehicle lighting, audio systems, and other electrical equipment, as well as the external environment (such as temperature and humidity). If the vehicle is in motion, the system also collects information on speed, acceleration, and route, which is uploaded to a cloud processing center in real time.

[0038] Step S104: Use the first prediction model to perform state detection on the battery based on the low-power detection conditions and battery state data to obtain the state detection result.

[0039] Among them, the state detection result is used to indicate whether the battery meets the low-charge detection conditions.

[0040] The first prediction model described above is a lightweight prediction algorithm, typically running locally in the vehicle, used to quickly and initially screen whether the threshold conditions for low battery detection are met. Based on battery state data, this model can rapidly assess whether the battery's current state is close to the edge of low battery when the vehicle starts, ensuring an immediate response.

[0041] A first predictive model deployed locally in the vehicle can be used to perform rapid state detection based on pre-defined depletion detection conditions (such as voltage thresholds and current change rates) and collected battery state data. The purpose of this step is to achieve real-time response through a lightweight model, determine whether the battery is close to a depleted state, and provide preliminary judgment for subsequent in-depth analysis.

[0042] The aforementioned low-charge detection conditions refer to a set of predefined standards or thresholds used to determine whether the current state of a vehicle battery is close to or has reached a critical point that may lead to a low-charge risk. These conditions are typically formulated based on the battery's normal operating range, the vehicle's power consumption, and the expected battery performance under specific conditions, aiming to identify the degree to which the battery state deviates from the normal range, thereby predicting potential low-charge situations.

[0043] The design of the first predictive model requires a rich dataset of battery states, including data under various normal and abnormal operating conditions. During model training, the data is first preprocessed, such as removing outliers, filling missing values, and normalizing. Then, based on the depletion detection conditions (such as voltage thresholds, current change rates, and temperature ranges), the data is labeled as either normal or depletion-risk states. Next, algorithms such as logistic regression, support vector machines, random forests, or neural networks are used to train the model, enabling it to identify which battery parameter combinations are strongly correlated with depletion risk, thus allowing for rapid response in subsequent applications.

[0044] In one alternative embodiment, the model receives the latest battery state data from the vehicle's local data acquisition system. This data may include parameters such as instantaneous battery voltage, current, temperature, and state of charge. The model's front end performs feature extraction on the input raw data, transforming the battery state data into more meaningful feature vectors for easier model understanding and processing. For example, in addition to directly using voltage values, the model may also calculate advanced features such as voltage change rate and the cross-influence factor between temperature and voltage. Based on the extracted feature vectors, the model predicts whether the current battery state meets the depletion detection criteria using an internal decision function or weight matrix. This process is typically very rapid, ensuring timely warnings. For example, the model might assess whether the battery voltage is below a certain threshold, or whether the battery temperature is under extreme conditions accompanied by an abnormal voltage drop. After the model's prediction, it generates a binary result, the state detection result, indicating whether the battery meets the depletion detection criteria. If it does, it means the battery may be about to reach or is already in a depleted state, requiring further analysis; if it does not, it indicates the battery is in good condition and requires no immediate action.

[0045] Step S106: In response to the state detection result that the battery meets the low-power detection conditions, the second prediction model is used to perform low-power detection on the battery based on power consumption data, driving status data and environmental data to obtain the low-power detection result.

[0046] Among them, the low-charge test result is used to indicate whether the battery is at risk of low charge.

[0047] The second prediction model mentioned above is more complex and in-depth than the first prediction model. It usually runs in the cloud and can process multi-dimensional information such as power consumption data of electrical equipment, driving status data and environmental data. Through machine learning or deep learning algorithms, it can accurately predict whether the battery is at risk of running out of power in a specific environment.

[0048] When the first prediction model's detection results indicate that the battery state meets the low-charge detection conditions, the system uploads the collected power consumption data, driving status data, and environmental data to the cloud for in-depth analysis using the second prediction model. Based on multi-dimensional data and through complex algorithms, the second prediction model comprehensively evaluates and determines whether the battery faces a genuine risk of low charge by accurately identifying the power consumption trends of electrical devices, energy consumption characteristics under driving modes, and the impact of environmental factors on battery performance.

[0049] Before receiving data, the second prediction model undergoes feature engineering, including data cleaning, outlier detection, feature selection, and transformation, to ensure data quality and applicability. For example, the model smooths power consumption data, eliminating noise from short-term fluctuations and extracting the true power consumption patterns of electrical devices. The model employs complex algorithms (such as deep neural networks and ensemble learning) to integrate data from different sources, forming a unified analytical framework. This means it can simultaneously consider dynamic power consumption changes during vehicle operation, the impact of recent weather patterns on battery performance, and the battery's own aging characteristics. The second prediction model is trained on historical datasets, covering various vehicle operating scenarios, usage patterns of different electrical devices, and battery performance data under a wide range of environmental conditions. Through iterative improvements, the model can more accurately identify combinations of multiple factors leading to battery depletion risk, improving prediction accuracy.

[0050] Considering real-time performance and data volume, this step employs a cloud-edge collaborative computing strategy. Specifically, after the first prediction model on the vehicle completes initial screening, it uploads key data to the cloud for in-depth analysis by a second prediction model. This approach ensures both processing speed and increased analytical complexity. The second prediction model doesn't simply statically determine whether data exceeds thresholds; instead, it dynamically assesses the battery's potential for depletion under upcoming usage conditions and electrical loads. For example, if it predicts the driver will be traversing a dense urban area with high power consumption from the air conditioning and entertainment systems, the model will predict the likelihood of future depletion based on these factors, even if the current battery level appears sufficient.

[0051] The second predictive model can handle more complex data relationships, identifying the underlying causes of battery drain risk rather than just superficial phenomena. By integrating multi-source data and advanced machine learning algorithms, the model can predict battery drain risk more accurately, significantly reducing false positive and false negative rates. Cloud-edge collaboration ensures the speed of data processing and the efficiency of in-depth analysis, enabling rapid response and deep analysis even under poor network conditions. The model can provide personalized battery drain risk assessments based on each driver's driving habits, driving environment, and vehicle configuration, improving the relevance and usability of the warnings.

[0052] The aforementioned second prediction model, through its powerful data processing and deep learning capabilities, can comprehensively assess the power consumption, driving status, and environmental factors of vehicle electrical equipment, accurately predict the risk of battery depletion, provide a scientific basis for subsequent early warning and prevention measures, and effectively improve the accuracy of early warning and the intelligence level of the system.

[0053] Step S108: In response to the low battery detection result indicating that the battery is at risk of low battery, a low battery warning message is generated based on battery status data, power consumption data, driving status data, and environmental data.

[0054] If the analysis results of the second predictive model confirm that the battery is at risk of depletion, the system will generate detailed depletion warning information based on the comprehensive data collected. The warning information not only includes the risk level, but may also include specific risk factor analysis, such as "prolonged idling leads to excessive energy loss" or "sudden drop in temperature at night affects battery performance," providing car owners with clear warning information and potential cause analysis to help them take preventive measures.

[0055] Once the second predictive model confirms a risk of battery depletion, it can generate and push a low-battery warning message based on battery status data, electrical device power consumption data, driving status data, and environmental data. First, the system identifies the specific factors causing the current risk of battery depletion from the output of the second predictive model. This may include abnormal drops in battery voltage and current, abnormally high power consumption of specific electrical devices, frequent low-speed driving in urban areas, and environmental conditions such as extreme temperatures. Based on the identified risk factors, the system generates a customized low-battery warning message. This message not only clearly indicates the risk of battery depletion but also details the source of the risk, such as "It has been detected that your vehicle's air conditioning is operating at high power continuously in low-temperature environments, which may cause a rapid drop in battery power." The system will then generate corresponding preventative suggestions based on the specific cause of the risk, combined with the user's driving habits and vehicle configuration. These suggestions aim to help users reduce battery consumption, such as suggesting turning off or adjusting the use of certain electrical devices under specific conditions, providing improved charging strategies, or suggesting changes in driving behavior to reduce battery load.

[0056] Furthermore, the generated alerts will be formatted into an easy-to-understand format, including text descriptions, icons, color coding, or simplified data charts, ensuring users can quickly interpret the information. For example, red indicates high risk, yellow indicates medium risk, and green indicates low risk, making the risk level immediately clear. To ensure users receive information promptly, the system will push low battery alerts through multiple channels, including in-vehicle displays, mobile applications, SMS or email, and even broadcast via voice assistant. This way, users can easily understand the vehicle's battery status regardless of their location. The system also features a feedback loop, allowing users to confirm or provide feedback on received alerts to further improve the accuracy of warnings and the effectiveness of preventative measures. For example, users can confirm whether suggested preventative measures have been taken or report whether the battery condition has improved, helping the system continuously learn and adjust its warning strategies.

[0057] When the second predictive model confirms a risk of battery depletion, the system immediately activates the risk factor identification module to analyze which factors are causing the decrease in battery power. Based on the risk factors, the information customization module generates a prompt message containing detailed risk descriptions and preventative recommendations. The information formatting module transforms the prompt message into intuitive user interface elements for easy understanding. The push system delivers the information through the user's preferred channel, ensuring timeliness and accessibility. After receiving the information and taking appropriate action, the user can inform the system through a feedback mechanism to improve future warnings and recommendations.

[0058] For example, suppose the system detects that the vehicle is parked in a low-temperature environment at night, the battery voltage is near a critical value, and the in-vehicle entertainment system is continuously running, causing additional power consumption. In this case, the system will generate the following message: "Battery Depletion Risk Warning: Your vehicle has been detected as parked in a low-temperature environment, and the in-vehicle entertainment system is running at high power, which may lead to insufficient battery power. It is recommended that you immediately turn off the entertainment system. If possible, please charge during the day when the temperature is higher. The system has reserved the nearest charging station for you; please check the mobile application for details."

[0059] Through the above process, the system can not only provide timely warnings of the risk of low battery power, but also provide specific preventive measures to help users take action to avoid the situation of low battery power, thereby ensuring the normal operation of the vehicle and improving the user's driving experience.

[0060] In one optional embodiment, after the system identifies a risk of battery depletion, it generates targeted depletion warning messages based on previously collected battery status data, electrical device power consumption data, driving status data, and environmental data. The system first performs secondary analysis on all collected data to extract the main factors contributing to the risk of depletion. For example, if power consumption data shows that the car audio system has consumed a large amount of power in a short period, driving status data indicates that the vehicle is in a congested urban area, and environmental data reveals that the current temperature is low, these factors will be integrated to analyze how they collectively contribute to the risk of battery depletion. Based on the above data analysis results, the system uses natural language generation technology to automatically generate clear and specific depletion warning messages. This information not only informs the user that the battery is currently at risk of depletion but also lists in detail the possible reasons, such as "high power consumption operation of the car audio system," "low-speed driving in urban areas," and "low-temperature environment," providing the user with action guidelines.

[0061] To cater to the diverse needs and preferences of users, the system can provide personalized preventative recommendations based on user history, vehicle configuration, and individual settings. For example, for users who frequently drive at night, the system may suggest reducing the use of interior lighting; for users living in cold regions, the system may recommend using a battery warmer or changing parking locations. To ensure the timeliness and accessibility of information, the system employs a multi-channel push strategy, utilizing in-vehicle screen displays, mobile application notifications, and voice assistant announcements to ensure users receive low battery warnings in the most convenient way, even while driving, allowing them to understand the situation through voice announcements and reducing distraction risks. Both the in-vehicle display and the mobile application feature an intuitive and easy-to-understand user interface design, using color coding, icons, and concise text descriptions to highlight the low battery risk level and preventative measures, enabling users to quickly understand and take action.

[0062] By combining the above technical features, the abstract risk of battery depletion can be transformed into specific and practical alerts, helping users to understand the battery status in real time and providing effective solutions. This avoids the inconvenience and safety hazards caused by insufficient battery power, and significantly improves the intelligent management level and user satisfaction of new energy vehicles.

[0063] In the field of new energy vehicles, frequent vehicle start-stop cycles, abnormal power consumption of electrical equipment, and severe weather conditions can all accelerate battery consumption, leading to the risk of battery depletion. This invention, by running a first predictive model locally in the vehicle, can quickly detect key indicators such as battery voltage and current to determine whether the battery depletion detection conditions are met. For example, if the battery voltage drops above a set threshold within a short period, the system considers the detection conditions met. Then, a second predictive model is used to deeply analyze the impact of electrical equipment power consumption, driving modes (such as frequent short-distance driving and prolonged high-speed driving), and external environmental factors (such as low and high temperatures) on the battery, accurately predicting the risk of battery depletion. Once a risk is confirmed, the system generates a battery depletion warning message containing the risk level and specific cause analysis. This message is immediately communicated to the owner via the in-vehicle screen, mobile application, or voice broadcast, guiding them to take appropriate preventative measures, such as turning off unnecessary electrical equipment, adjusting driving plans, or seeking professional repair services. This significantly improves the efficiency and accuracy of battery depletion warnings, enhancing the safety and convenience of vehicle use.

[0064] This application achieves efficient early warning and accurate analysis of battery depletion risk in vehicles by combining a dual-model strategy with multi-dimensional data collection. Compared to traditional early warning methods that rely on single detection conditions or simple threshold comparisons, this invention can more comprehensively consider battery state changes under complex scenarios such as electrical equipment power consumption, driving status, and environmental factors. This improves the accuracy and timeliness of early warnings, effectively reduces the risk of unexpected battery depletion, and enhances the operational safety and user satisfaction of new energy vehicles.

[0065] For example, suppose a new energy vehicle is parked at night in a low-temperature environment, and the vehicle's onboard system (such as a smart key system) issues a low-battery detection command. The system first uses a locally deployed first predictive model to quickly detect the battery voltage and temperature, making a preliminary judgment on whether the battery condition is close to the potential low-temperature low-battery conditions. If the first predictive model confirms that the battery condition is close to a threshold, the system uploads collected power consumption data from electrical devices (air conditioning, lights, standby devices, etc.), driving data (mileage and speed changes from the most recent drive), and environmental data (temperature and humidity) to the cloud. A second predictive model in the cloud further analyzes this data, comprehensively assessing the battery performance degradation under low-temperature conditions, the heat consumption of electrical devices, and the vehicle's static power consumption. If the system confirms a low-battery risk, it sends a warning via the owner's mobile application, informing the owner that "your vehicle's battery performance has deteriorated due to the low-temperature environment, and the air conditioning heating function has been used frequently. It is recommended to reduce the use of unnecessary electrical appliances or start the vehicle in advance to preheat the battery." In this way, by comprehensively analyzing multi-source information, the system not only warns of the low-battery risk but also provides specific reasons and preventative suggestions, improving the practicality of the warning and the owner's response efficiency.

[0066] Through the above steps, in response to receiving a low battery detection command from the vehicle, the system collects battery status data, power consumption data of electrical devices in the vehicle, vehicle driving status data, and environmental data of the vehicle's environment. A first prediction model is used to perform a battery status detection based on the low battery detection conditions and battery status data to obtain a status detection result, whereby the status detection result indicates whether the battery meets the low battery detection conditions. In response to the status detection result indicating that the battery meets the low battery detection conditions, a second prediction model is used to perform a low battery detection based on power consumption data, driving status data, and environmental data to obtain a low battery detection result, whereby the low battery detection result indicates whether the battery meets the low battery detection conditions. Does the battery have a risk of being depleted? In response to the depletion detection result indicating that the battery has a risk of being depleted, a depletion warning message is generated based on battery status data, power consumption data, driving status data, and environmental data. By comprehensively analyzing the status detection results of the first prediction model and the depletion detection results of the second prediction model, and further combining the changing trend of power consumption data, the energy consumption pattern under driving status data, and the degree of influence of environmental data on battery performance, the specific factors leading to the risk of being depleted and the severity level of the risk can be efficiently and accurately derived. This achieves the technical effect of improving the prediction efficiency of battery depletion, and thus solves the technical problem of low prediction efficiency of battery depletion in related technologies.

[0067] Optionally, based on battery status data, power consumption data, driving status data, and environmental data, a low battery warning message is generated, including: based on battery status data, power consumption data, driving status data, and environmental data, fault location is performed on the low battery risk, and the risk type corresponding to the low battery risk is determined; based on the risk type, the low battery risk is quantified to obtain the battery risk level; and based on the risk level and risk type, a low battery warning message is generated.

[0068] The aforementioned fault location refers to the process of accurately determining the location or cause of a problem by analyzing the operating status data of a system or equipment. In this context, it means that the system specifically identifies the factors that lead to the risk of power depletion, such as identifying which electrical appliance is abnormally consuming power or whether the battery's efficiency has decreased due to aging.

[0069] The above risk types are classifications of the nature of risks, which can help us better understand the essence of risks and formulate corresponding countermeasures. Regarding the risk of power depletion, risk types may include, but are not limited to, "battery aging," "abnormal power consumption of electrical equipment," "frequent use of high-power devices," and "environmental impact."

[0070] The aforementioned quantification process refers to the conversion of qualitative descriptions into quantitative values. In power loss risk assessment, quantification can refer to converting identified risk types into risk levels, such as using numbers 1-5 to represent extremely low, low, medium, high, and extremely high risks, respectively, to facilitate subsequent processing and decision-making.

[0071] The risk levels described above are quantified indicators used to represent the severity of a risk. In this system, risk levels indicate the urgency and impact of power loss risks, helping users quickly understand and take appropriate measures.

[0072] In one optional embodiment, the system first uses data analysis and machine learning algorithms based on collected battery status data, electrical device power consumption data, driving status data, and environmental data to perform in-depth fault localization of the current risk of battery depletion. This step aims to identify the root cause of the risk of battery depletion and classify these causes into different risk types, such as "excessive power consumption" and "damaged battery health." After determining the risk type, the system quantifies the identified risk of battery depletion, converting it into a specific risk level. This process considers the severity of the risk type, its frequency of occurrence, and its impact on battery life and vehicle operation. The risk level classification helps users more intuitively understand the urgency of the risk and guides them to take appropriate preventive measures.

[0073] Based on the risk level and risk type, the system will generate specific low battery warning messages, clearly describing the current risk level to the user and providing preventative suggestions based on the risk type. For example, if the risk level is increased due to abnormal power consumption by the vehicle's air conditioning in low temperatures, the warning message will suggest reducing air conditioning use or, under safe conditions, trying to restart the system to see if it restores normal power consumption.

[0074] The system begins with fault location, analyzing data such as battery status, appliance power consumption, and driving environment to accurately identify the specific factors leading to the risk of battery depletion, and then classifying these factors into risk types based on their different characteristics. Subsequently, the system converts risk types into risk levels, quantitatively expressing the severity of the risk. Finally, based on the risk level and type, the system generates user-friendly prompts, not only alerting users to the existence of the risk but also providing actionable suggestions to mitigate it.

[0075] By meticulously analyzing the various factors contributing to the risk of power loss, fault location ensures accurate diagnosis, avoiding false warnings and ineffective suggestions. Quantifying risk levels helps users understand the urgency of the current risk, guiding them to prioritize and address high-risk situations and improving the effectiveness of preventative measures. The alerts generated based on risk type and level provide users with specific and feasible prevention strategies, not only solving the current problem but also providing guidance for taking preventative measures in similar situations.

[0076] When faced with the risk of battery depletion in new energy vehicles, the system performs in-depth analysis of battery status data, power consumption data, driving status data, and environmental data to perform fault location, risk type determination, quantitative processing, and information generation, providing accurate and user-friendly early warnings.

[0077] The system employs pattern recognition techniques from artificial intelligence, such as convolutional neural networks or long short-term memory networks in deep learning, combined with a rule engine and expert system, to perform multi-dimensional analysis of data and identify the specific causes of power loss risks. For example, by analyzing the trend of power consumption data over time, the system identifies the phenomenon of abnormally high power consumption of in-vehicle entertainment systems at night, classifying it as a risk type of "abnormal operation of high-power devices at night." This process relies on a pattern library built on big data, automatically identifying the source of risk through pattern matching and anomaly detection algorithms.

[0078] The system quantifies the identified risk types and uses machine learning algorithms such as decision tree models based on risk factor weights or Support Vector Machines (SVM) to assess the urgency and severity of each risk type, thereby determining the battery's risk level. For example, for the "damaged battery health" risk type, the system assesses factors such as the battery's remaining capacity, temperature fluctuation range, and charging cycles, quantifying their contribution to the risk of battery depletion, and ultimately classifying it as a "high" risk level. The purpose of this classification is to better guide users in taking preventative measures and prioritizing high-risk situations.

[0079] Based on the determined risk level and type, the system uses Natural Language Processing (NLP) technology, combined with user behavior analysis and preference settings, to automatically generate easy-to-understand low battery warning messages. The information generation module considers the channel through which the user receives the information (such as in-vehicle display, mobile application, or voice broadcast) and adjusts the urgency and detail of the information according to the risk level. For example, for a high-risk low battery warning, the message will emphasize the necessity of immediate action and provide direct and specific preventative advice, such as "Immediately turn off non-essential in-vehicle devices and go to the nearest charging station to charge."

[0080] The entire early warning system integrates a dynamic learning mechanism, continuously collecting user feedback and real-time data to improve fault location algorithms, risk quantification models, and information generation strategies. The system can identify which preventative measures are most effective in specific situations, thus providing more personalized recommendations in future warnings and enhancing the system's practicality and user satisfaction.

[0081] Optionally, based on the risk level and risk type, a low battery warning message is generated, including: obtaining vehicle scenario data, wherein the vehicle scenario data is used to represent vehicle usage data under different scenarios; determining the low battery warning scenario based on the vehicle scenario data and risk type; and generating the low battery warning message based on the low battery warning scenario, risk type, and risk level.

[0082] The vehicle scenario data mentioned above is a data set describing the vehicle's operating status under different application scenarios, including but not limited to mileage, road conditions, weather conditions, driver habits, etc., which is crucial for understanding the vehicle's usage and efficiency under specific conditions.

[0083] The aforementioned low battery warning scenarios are determined by the system based on vehicle scenario data and risk types, representing the most suitable scenarios to issue low battery warnings to users. For example, when a vehicle is parked for an extended period and the ambient temperature is low, the system may classify this as a high-risk low battery scenario and require an immediate warning message to the user.

[0084] In one optional embodiment, vehicle scenario data is first collected from sensors, user input, and other data sources. This data helps construct a usage profile of the vehicle in different scenarios, including but not limited to user driving habits, vehicle usage frequency, and changes in environmental conditions. Combining the collected risk types and vehicle scenario data, the system uses scenario analysis algorithms to determine in which specific scenarios a low battery warning should be issued. This step considers the potential impact of the risk and the triggering conditions of the scenario to ensure the timeliness and relevance of the warning. After determining the warning scenario, the system generates a user-friendly prompt message based on the risk type, risk level, and current scenario. This message not only includes a warning but also provides a specific risk description and prevention suggestions, aiming to help users take effective measures in specific scenarios to avoid low battery situations.

[0085] The system not only assesses the risk of low battery power based on battery status, power consumption, driving conditions, and environmental data, but also takes into account the vehicle's usage scenarios to ensure customized warning messages and scenario applicability. By collecting and analyzing vehicle scenario data, the system can more accurately determine when, where, and under what circumstances a low battery warning is most effective, thereby generating scenario-appropriate and targeted prompts to help users take quick action when faced with insufficient battery power, reduce unnecessary power consumption, and protect battery health.

[0086] By combining vehicle scenario data with the type of battery depletion risk, the system can provide more accurate and practical early warning information, enhancing user responsiveness and vehicle adaptive management capabilities. This intelligent early warning strategy not only reduces false alarms and avoids unnecessary anxiety, but also enables users to take more effective preventative measures based on their vehicle usage and environmental conditions, extending battery life and improving vehicle performance.

[0087] For example, consider a car owner who frequently parks their car outdoors on cold winter nights and uses the vehicle's intelligent driver assistance systems (such as adaptive cruise control) frequently during the day. In this scenario, vehicle scenario data reflects the usage habits of low-temperature environments and high-power devices. The system analysis reveals that these factors combined cause the vehicle's battery level to drop rapidly in a short period, indicating a high risk level. Therefore, the system automatically identifies "cold nighttime outdoor parking and high power consumption of the intelligent driver assistance system" as a low battery warning scenario and sends a alert to the owner: "We have detected that you frequently park outdoors on cold nights and use the intelligent driver assistance system frequently, which will significantly increase the risk of battery drain. The current risk level is high. Please consider parking the vehicle in a warmer garage and minimize the use of the intelligent driver assistance system unless absolutely necessary to protect the battery. The system has planned the nearest and most affordable charging station for you, and we recommend that you go there to charge immediately after parking."

[0088] By integrating detailed vehicle scenario data, the system can provide highly applicable and comprehensive early warning information, helping car owners maintain a relatively good battery condition even when facing complex usage conditions, thereby significantly improving the user experience and safety of new energy vehicles.

[0089] In one optional embodiment, time series analysis and clustering algorithms can be used to extract scene features from historical driving data, such as driving time periods (morning and evening rush hours, late at night, etc.), regular driving routes, average speed ranges, and ambient temperature ranges. Simultaneously, by combining real-time GPS data, meteorological information, and user behavior patterns, scene data is dynamically updated to reflect the latest vehicle usage. Employing context-aware technology, combining risk types with vehicle scene data, the system intelligently determines which scenarios pose a greater risk of battery depletion. Through scene matching algorithms, such as rule-based scene classifiers or deep learning scene recognition models, the system can accurately locate typical battery depletion risk scenarios such as "frequent short-distance travel at night in low temperatures" or "long-term high-speed driving," providing crucial context for generating warning information.

[0090] Utilizing Natural Language Generation (NLG) technology, customized warning messages are generated based on the low battery warning scenario, risk type, and risk level. The NLG module combines user preferences (such as voice, text, and chart formats), vehicle condition, and expected driving behavior to formulate the most appropriate warning strategy. For example, for the "high-risk level" "battery aging" risk type, the warning message will include an urgent suggestion to have the battery checked as soon as possible; while for the "medium-risk level" "parking in low temperatures overnight," it may suggest activating the vehicle's low-temperature battery protection mode to reduce power loss.

[0091] Through the in-depth implementation of the aforementioned technical features, the system can more precisely depict the vehicle's usage environment and accurately match risk types and scenarios, thereby generating more targeted and actionable warning information. This effectively encourages users to take preventative measures and avoid the inconvenience caused by battery depletion. For example, in the scenario of "frequent short-distance travel at night in low temperatures," the system identifies the dual risks of decreased battery charging efficiency and increased power consumption of electrical devices in low-temperature environments. Combined with a "medium" risk level, it generates the following suggestion: "Given the recent frequent short-distance travel at night in low temperatures, your battery power consumption has increased. It is recommended to check the remaining power before traveling, consider charging in advance, and minimize the use of air conditioning to reduce power consumption. The current risk level is medium. Please plan your driving route and parking location reasonably to protect battery health." This type of information not only warns of risks but also provides practical coping strategies, greatly improving the practicality of the warning system and the user experience.

[0092] Optionally, the method further includes: sorting the power consumption of multiple devices corresponding to the vehicle to obtain a power consumption sorting result, wherein different devices are used to implement different functions of the vehicle; determining at least one target device based on the power consumption sorting result and generating control instructions for at least one target device; sending the control instructions to at least one target device and receiving control results from at least one target device, wherein the control results are used to indicate whether at least one target device has been successfully turned off.

[0093] The aforementioned power consumption refers to the electrical energy consumed by various devices on the vehicle during operation, and is an important indicator for measuring equipment efficiency and battery load.

[0094] The power consumption ranking results mentioned above are a list of all devices on the vehicle arranged according to their power consumption through algorithm analysis, used to identify which devices are high power consumers.

[0095] The target devices mentioned above refer to those that the system determines need to control power consumption to reduce the burden on the battery, based on the power consumption ranking results. These devices are usually selected from high-power devices.

[0096] The aforementioned control commands are signals sent from the system to the target device, instructing the device to change its operating state, such as shutting down, entering a low-power mode, or adjusting operating parameters.

[0097] The control result mentioned above refers to the feedback received by the system from the target device, confirming whether the device has successfully adjusted its power consumption state according to the control command.

[0098] The system collects and analyzes the real-time power consumption of all electrical devices in the vehicle, and uses sorting algorithms (such as quicksort and heapsort) to sort the devices by power consumption, generating a power consumption ranking result. This process helps identify the devices with the highest power consumption in the vehicle, providing a basis for subsequent power consumption control. Based on the power consumption ranking result, the system can determine which devices are high-power devices in the current battery state, and then identify them as target devices. Subsequently, the system automatically generates corresponding control commands based on the characteristics of the target devices, such as requiring the in-vehicle entertainment system to enter standby mode, adjusting the temperature setting of the in-vehicle air conditioning, or turning off non-critical equipment.

[0099] Control commands are sent to the target device via the vehicle's communication network. The device adjusts its power consumption state according to the received commands and feeds back the control results (i.e., changes in power consumption state) to the system, forming a closed-loop control mechanism. The system monitors whether the device successfully follows the control commands and whether the power consumption changes are as expected, ensuring that battery-saving measures are effectively implemented.

[0100] Through dynamic power consumption management and control, intelligent energy saving is achieved for vehicles in different usage scenarios. By prioritizing power consumption and identifying target devices, the system can accurately identify and control high-power-consuming equipment, reducing unnecessary power loss and extending battery range. Simultaneously, a closed-loop feedback mechanism between control commands and results ensures the effectiveness of command execution, improving system reliability and energy efficiency management.

[0101] Considering the trend towards intelligentization in new energy vehicles and the increasing number of in-vehicle devices, the system intelligently prioritizes the real-time power consumption of different devices in the vehicle, identifying high-power devices such as large touchscreen in-vehicle information systems, advanced audio systems, and heated seats. Based on the vehicle's current battery status and driving environment (e.g., low temperature, low battery charge), the system designates these high-power devices as target devices, automatically sending control commands to require them to enter low-power mode or temporarily shut down. Subsequently, the system receives and verifies the control results of the target devices to ensure that energy-saving measures are effectively implemented.

[0102] The methods described above significantly improve the precision and proactivity of battery power management, helping to extend vehicle availability during critical moments. Especially in situations of low battery power, by shutting down unnecessary high-power devices, sufficient power can be reserved for more critical functions (such as the powertrain and safety systems), ensuring driving safety and the normal operation of essential functions. Furthermore, in the long term, intelligent power management also helps to improve overall battery lifespan, reduce charging frequency, and save energy costs.

[0103] For example, suppose it's a cold morning and the vehicle's battery status indicates low power, while devices such as the in-vehicle entertainment system, heated seats, and rear window defroster are all running. The system first collects the current power consumption of these devices, prioritizes them, and identifies the heated seats and rear window defroster as target devices due to their higher power consumption caused by operating in low temperatures. The system then automatically generates control commands, instructing the heated seats and rear window defroster to reduce power consumption or temporarily shut down to minimize energy consumption. Upon receiving the commands, the devices automatically adjust their status and feed the control results back to the system. Through this series of operations, the vehicle can prioritize the normal operation of the powertrain and safety systems when battery power is limited, while reminding the user to reduce the use of unnecessary devices to avoid the vehicle failing to start due to low battery, thereby improving vehicle usability and user safety.

[0104] By utilizing high-precision current sensors and power monitoring modules built into the vehicle, the current requirements and power consumption of various electrical devices are captured in real time and transmitted to the vehicle system or cloud-based big data platform. Data analytics techniques, such as real-time stream processing frameworks (e.g., Apache Kafka or Apache Flink), are used to continuously monitor and dynamically prioritize device power consumption, ensuring data real-time performance and accuracy. The power consumption ranking algorithm considers the instantaneous and average power consumption of devices, as well as their usage frequency and importance, constructing a multi-dimensional power consumption assessment model to generate a ranking result reflecting the power consumption priority of the devices.

[0105] In one alternative embodiment, the system can combine the vehicle's current battery status, driving mode, and predicted trip requirements. It can then use machine learning algorithms such as decision trees and support vector machines, or rule-based logical judgments, to select high-power, non-immediately-needed devices from the power consumption ranking results as target devices. For example, when battery power is low and the expected trip is short, the system might identify entertainment system devices such as in-car audio systems and ambient lighting as targets to reduce unnecessary power consumption.

[0106] The system generates customized control commands based on the characteristics of the target device, such as requiring the in-vehicle entertainment system to enter low-power mode, adjust the in-vehicle air conditioning temperature setting, or shut down non-critical equipment. These commands are sent to the target device via in-vehicle network communication technology, triggering its energy management mechanism. Simultaneously, a device response mechanism is introduced; upon receiving the command, the device must return an execution confirmation signal to ensure correct execution. The system also continuously monitors the control results, assessing whether the power consumption reduction has achieved the expected target and its impact on the overall vehicle operation, achieving closed-loop management from control command generation to execution result feedback.

[0107] After the device executes control commands, the system continuously collects data on the adjusted power consumption through a sensor network, compares and analyzes this data with the power consumption data before control, and evaluates the energy-saving effect. Furthermore, the system integrates user feedback, device health status, and real-time environmental parameters to dynamically adjust power consumption improvement strategies. For example, if the user confirms that the impact on comfort is minimal, the power consumption of non-critical equipment is further reduced; conversely, if the user reports a decrease in comfort or environmental conditions change, power consumption is appropriately increased to ensure a good user experience. Through this mechanism, the system can achieve intelligent energy-saving control while ensuring vehicle functionality and user experience, effectively preventing battery depletion.

[0108] Considering the impact of high-power devices on battery capacity in new energy vehicles, when the system detects that the vehicle is about to enter a low-battery mode and is about to enter a low-temperature parking scenario at night, it will automatically prioritize the power consumption of all electrical devices in the vehicle. Assuming that the in-vehicle entertainment system and electric seat heaters have high power consumption and are not immediately necessary in the current environment, the system identifies them as target devices, generates control commands, and requests the entertainment system to enter standby mode and temporarily disable the seat heating function.

[0109] Control commands are sent to the target device via an encrypted in-vehicle communication module. Upon receiving the command, the device executes it and reports the control result back to the system. The system continuously monitors the execution effect to ensure effective power saving while guaranteeing the normal operation of critical vehicle functions. During this process, the system may also adjust control commands in a timely manner based on user preferences and environmental changes. For example, if the user manually activates the seat heating, the system will prioritize user comfort and appropriately adjust the power consumption control strategies of other devices.

[0110] Optionally, in response to the battery being at risk of being depleted as detected by the low-power detection, the method further includes: obtaining the vehicle's current location information and / or vehicle navigation information; determining the target service point's service point information based on the current location information and / or vehicle navigation information, wherein the target service point is used to provide battery repair services; and generating a navigation path to the target service point in response to receiving a confirmation instruction for the service point information.

[0111] The aforementioned current location information refers to the geographic coordinates of the vehicle's current location, which is usually obtained through global positioning systems such as GPS or BeiDou, and is used to pinpoint the vehicle's specific location.

[0112] The aforementioned vehicle navigation information includes not only the current location information, but may also include the destination, estimated arrival time, driving route, etc. It is a set of information input by the user in the vehicle navigation system or mobile application for planning driving routes.

[0113] The aforementioned target service outlets refer to locations recommended by the system based on the vehicle's current condition and user needs, which can provide repair services for the vehicle's battery. These include, but are not limited to, auto repair shops, battery service centers, and 4S car dealerships, which possess professional technology and equipment to handle battery depletion issues.

[0114] The above-mentioned service outlet information is detailed information about the target service outlet, including address, business hours, service type, user reviews, etc., providing users with the necessary information to select and visit the service outlet.

[0115] The aforementioned confirmation command is a command sent by the user to the system via the in-vehicle screen, mobile app, or voice interaction after receiving the target service point information recommended by the system, confirming that they are going to the service point and initiating the navigation planning process.

[0116] The navigation route mentioned above refers to the driving route from the vehicle's current location to the target service point. It is calculated by the vehicle navigation system or mobile application based on algorithms such as real-time traffic conditions, shortest distance, or least time, to provide navigation guidance for users to reach the service point.

[0117] In one optional embodiment, the system acquires the vehicle's location data in real time through the vehicle's built-in GPS positioning system. Simultaneously, if the user has already set a destination or route in the in-vehicle navigation system or mobile application, the system obtains this navigation information as a reference for determining service outlets. Combining the vehicle's current location information and navigation information, the system uses big data analytics and machine learning techniques to intelligently filter out service outlets that are closest to the vehicle, have good service ratings, and can provide battery repair services. This process may also involve querying real-time data such as the service outlet's opening hours and queue status to ensure that the recommended outlets can respond to user needs promptly.

[0118] Once the user confirms the recommended service point via the in-car screen, mobile app, or voice interaction, the system uses APIs from map services such as Google Maps and Baidu Maps, combined with real-time traffic conditions, to generate a recommended navigation route from the vehicle's current location to the service point. The user can view the route details on the in-car navigation system or mobile app and select to begin navigation.

[0119] By intelligently recommending service outlets and generating navigation routes, the system provides users with convenient battery repair solutions. When the system detects a risk of battery depletion, it not only issues a warning but also proactively recommends nearby service outlets and plans navigation routes to help users quickly locate and reach a repair shop, effectively preventing further battery damage and improving vehicle availability and safety. Furthermore, by integrating real-time traffic data and user reviews, the system's recommended service outlets are more reliable and convenient, enhancing user trust and satisfaction with the entire prevention and repair system.

[0120] Through an integrated high-precision navigation module, the system acquires the vehicle's precise location information in real time, including latitude, longitude, altitude, and speed. Simultaneously, it utilizes the vehicle's navigation system to obtain navigation information such as the current route, destination, and estimated arrival time. This process ensures the real-time nature and accuracy of location and route data, providing a foundation for intelligent matching of subsequent service points.

[0121] Based on the acquired current location and vehicle navigation information, the system uses geospatial analysis techniques, such as spatial indexing (Quadtree, R-tree) and nearest neighbor search algorithms, to quickly filter service points near the vehicle or along the route from a pre-set service point database. Taking into account factors such as service point qualifications, user reviews, service area, and waiting time, a multi-objective improvement algorithm (such as genetic algorithm and particle swarm optimization) is used to evaluate and rank the service points, ultimately determining the target service point—the one that can provide battery repair services and best meets the user's needs.

[0122] The system presents the filtered target service point information in a visual interface (such as map markers or list display) on the in-vehicle screen or the user's mobile app, including the service point name, address, service content, and estimated waiting time. After the user confirms the target service point via the in-vehicle touchscreen or mobile phone, the system receives the confirmation command and uses a route planning algorithm (such as Dijkstra's algorithm) combined with real-time traffic conditions to generate a navigation route from the current vehicle location to the target service point. The system updates the route information in real time to ensure the accuracy and timeliness of the navigation.

[0123] This application can promptly detect battery depletion risks and further provide intelligent service point matching and navigation route planning, greatly facilitating users' access to repair services when facing battery problems and effectively improving user experience and vehicle maintenance efficiency. For example, when a battery depletion risk is detected, the system automatically filters highly rated service points with battery repair capabilities near the user's current location. After the user confirms, the system immediately plans a better route to guide the user to the location quickly, solving the problem of finding suitable repair resources in emergency situations. It also improves the entire repair service process, demonstrating the efficiency and convenience of intelligent vehicle maintenance management.

[0124] Optionally, the method further includes: acquiring sample voltage data, sample current data, and sample temperature data of the first sample battery; constructing first sample state data and sample state detection results of the first sample battery based on the sample voltage data, sample current data, and sample temperature data; inputting the first sample state data into a first initial prediction model, using the first initial prediction model to perform state detection on the first sample battery, and obtaining a predicted state detection result; and adjusting the model parameters of the first initial prediction model based on the sample state detection result and the predicted state detection result to obtain a first prediction model.

[0125] The sample batteries mentioned above refer to a specific set of batteries used for model training and validation. These can be batteries from a laboratory environment or representative vehicle batteries from real-world application scenarios. These batteries need to be carefully selected to cover different battery types, lifespans, and operating conditions, thereby ensuring the universality and accuracy of the predictive model.

[0126] The sample voltage data, sample current data, and sample temperature data mentioned above represent the voltage, current, and temperature values ​​recorded by the sample battery at different time points, respectively. These data serve as the basic inputs for constructing the battery state model. This type of data is typically collected using high-precision sensors and stored in a database as the basis for model training.

[0127] The aforementioned first sample state data refers to the state description that integrates the voltage, current, and temperature data of the sample battery. It not only reflects the working state of the battery at a certain moment, but may also contain multi-dimensional information such as the degree of battery aging and health status. It is a key data form for the input of the prediction model.

[0128] The above-mentioned sample status detection results are the actual results obtained by professional technicians or existing detection technologies after diagnosing the sample battery status, including battery capacity, health status, and whether there are any safety hazards, and are used to verify the accuracy of the prediction model.

[0129] The initial prediction model described above is a foundational model built during the early stages of model training. It may be based on regression or classification algorithms in machine learning or neural network architectures in deep learning, and is used to make preliminary predictions about battery status. The model parameters are not fully adjusted during the initial training phase and need to be improved iteratively using sample data.

[0130] The aforementioned predicted state detection result is a battery state estimate obtained by predicting and analyzing the first sample state data using the first initial prediction model. It may include information such as the classification of battery health status and the predicted value of remaining power.

[0131] The aforementioned model parameter adjustment refers to adjusting the parameters of the first initial prediction model based on the error between the sample state detection result and the predicted state detection result, using improved algorithms such as backpropagation and gradient descent, in order to improve the accuracy and stability of the model prediction.

[0132] First, the system periodically collects voltage, current, and temperature data from sample batteries using high-precision sensors such as voltage, current, and temperature sensors, forming the first sample state dataset. After collection, this data undergoes preprocessing methods such as data cleaning, outlier detection, and standardization to ensure data quality and lay the foundation for model training. The first sample state data is then input into the first initial prediction model. The model learns patterns and rules from the data to generate predicted state detection results. Subsequently, the prediction results are compared with sample state detection results obtained by professional technicians or existing detection technologies to calculate prediction errors, such as mean squared error and classification accuracy, to evaluate model performance.

[0133] Based on the prediction error, the system employs gradient descent, stochastic gradient descent, or other improved algorithms to adjust the parameters of the initial prediction model, minimizing the gap between the prediction and the actual detection results. This process may require multiple iterations until the model achieves the expected prediction accuracy. After parameter adjustment and verification, an improved first prediction model is obtained, which can be used for intelligent prediction of real-time vehicle battery status. Deploying this model on a big data platform or in-vehicle system enables real-time prediction of battery status and early warning of potential battery depletion risks based on actual battery voltage, current, and temperature data.

[0134] By building and improving models, accurate predictions of vehicle battery status can be achieved, helping to identify battery risks early and avoid sudden battery depletion failures. The model is trained based on a large amount of real-world sample data, and its prediction accuracy is continuously improved by adjusting parameters, making early warning information more reliable. This enhances the capabilities of the vehicle's intelligent management system and improves driving safety and convenience for users.

[0135] The system first selected a group of batteries from multiple new energy vehicles participating in the experiment as the first sample batteries. These batteries covered different brands, models, years of use, and operating environments, ensuring diversity in model training. Subsequently, through high-precision sensors inside the vehicles, the system collected a large amount of voltage, current, and temperature data from the sample batteries, forming a rich dataset of the first sample state. This data, after preprocessing, was input into the initially constructed prediction model for training. The model adopted a multi-layer neural network architecture, capable of handling complex nonlinear relationships. By comparing the battery state predicted by the model with the actual detection results of technical experts, the system used a gradient descent algorithm to adjust the model parameters, improving prediction accuracy. After multiple rounds of iterative improvements, the model can accurately predict key information such as battery health status and remaining charge, demonstrating excellent performance in prediction accuracy metrics. Finally, the improved first prediction model was deployed in the vehicle's intelligent diagnostic system. By monitoring the vehicle's battery status in real time, it provides early warnings of potential battery depletion risks, guiding users to take preventative measures and significantly improving the reliability of battery management and vehicle use.

[0136] By implementing the aforementioned technical features, the system can not only monitor vehicle battery status based on real-time data, but also utilize intelligent models for in-depth analysis and prediction, identifying trends in battery health decline in advance and preventing vehicle malfunctions caused by sudden battery depletion. This prediction method based on big data and machine learning significantly improves the accuracy and timeliness of early warnings, reducing inconvenience and safety risks for users due to battery problems. It represents a significant technological innovation in the field of intelligent vehicles and has a substantial positive impact on enhancing the user experience and market competitiveness of new energy vehicles.

[0137] In one optional embodiment, high-precision voltage, current, and temperature sensors are installed on the first sample battery to monitor and record voltage, current, and temperature data in real time during battery operation. Data stream processing technology is used to collect sensor data in real time, ensuring data continuity and integrity. After data preprocessing, such as denoising, normalization, and feature selection, a first sample state dataset is constructed, containing multi-dimensional time-series data of battery voltage, current, and temperature. A battery testing device is used to perform in-depth testing on the first sample battery to obtain accurate detection results of the battery's health status, including but not limited to key parameters such as battery capacity, internal resistance, and state of charge. These accurate detection results constitute the sample state detection results and are used as label data during model training.

[0138] A first initial prediction model is constructed. The model structure can be a deep learning model, such as a Long Short-Term Memory network, a gated recurrent unit, or other neural network structures suitable for processing time series data. The preprocessed first sample state data is input into the model, and supervised learning methods are used, with the sample state detection results as the ground truth labels, to train the model's ability to predict battery states. During training, techniques such as cross-validation and grid search are used to continuously adjust model parameters (such as learning rate, number of hidden layers, and activation function type) until the error between the predicted state detection results and the sample state detection results reaches an acceptable range, resulting in an improved first prediction model.

[0139] To ensure the model's generalization ability and prediction accuracy, a separate set of sample data is input into the first prediction model for validation, evaluating the model's performance on unseen data. If the model's prediction error on the validation set meets the set criteria, it can be deployed to real-world application scenarios for online monitoring and early warning of battery status. Furthermore, the system will continuously collect actual operational data and, through an online learning mechanism, periodically update the model parameters to continuously improve model performance, ensuring its adaptability to changes in battery characteristics and the usage environment.

[0140] By further implementing the aforementioned technical features, this application can learn the patterns of battery state changes from the historical operating data of the first sample battery, constructing a first prediction model with high prediction accuracy. This model can not only accurately diagnose the current state of the battery but also predict potential future power depletion risks, providing strong technical support for intelligent battery management. For example, the model can predict the trends of battery voltage and current changes under specific temperature conditions, detect signs of battery aging in advance, achieve precise early warning, and effectively prevent problems such as vehicle starting failure caused by battery malfunctions, thereby improving the reliability of new energy vehicles and the user experience.

[0141] Optionally, the method further includes: acquiring the sample battery internal resistance, sample electrolyte state, sample driving data, and sample environment data of the second sample battery, wherein the sample driving data is the driving data of the vehicle where the second sample battery is located, and the sample environment data is the environmental data of the vehicle where the second sample battery is located; constructing second sample state data and sample depletion detection results of the second sample battery based on the sample battery internal resistance, sample electrolyte state, sample driving data, and sample environment data; inputting the second sample state data into a second initial prediction model, using the second initial prediction model to perform depletion detection on the second sample battery, and obtaining a predicted depletion detection result; adjusting the model parameters of the second initial prediction model based on the sample depletion detection result and the predicted depletion detection result, and obtaining a second prediction model.

[0142] The second sample battery mentioned above refers to a set of representative battery samples used to build and improve the battery depletion prediction model. These batteries can be laboratory test batteries or vehicle batteries in actual operation, used for model training and validation to ensure the accuracy and generalization ability of the model.

[0143] The aforementioned sample battery internal resistance refers to the internal resistance value of the second sample battery under different operating conditions obtained through battery internal resistance testing equipment. Internal resistance is an important parameter reflecting the battery's health status and power transmission efficiency, and its changing trend can indirectly reflect the battery's risk of depletion.

[0144] The aforementioned sample electrolyte state represents key attributes such as electrolyte composition, density, and temperature. By using electrolyte state monitoring equipment, such as optical sensors and ultrasonic sensors, various state data of the sample battery electrolyte are collected periodically to analyze the correlation between electrolyte state changes and the risk of power depletion.

[0145] The aforementioned sample driving data refers to the driving status information of the vehicle where the second sample battery is located during actual use, including mileage, speed, route, number of starts, idling time, etc., which are collected in real time through on-board sensors and GPS system to assess the impact of different driving conditions on battery depletion.

[0146] The aforementioned sample environment data covers the operating conditions of the vehicle containing the second sample battery under different environments, such as temperature, humidity, and air pressure. This data is collected through external environmental monitoring equipment, such as temperature sensors and hygrometers, to analyze the impact of environmental changes on battery lifespan and the risk of battery depletion.

[0147] The aforementioned second initial prediction model is a preliminary model built on machine learning or deep learning algorithms to predict the risk of battery depletion. The model parameters were not fully adjusted in the early stages of training and need to be improved and adjusted by comparing the sample depletion detection results with the predicted depletion detection results.

[0148] The aforementioned predicted battery depletion detection results refer to the battery depletion risk prediction results obtained after analyzing the second sample state data through the second initial prediction model, such as high, medium, and low risk levels or specific depletion probability values.

[0149] The system integrates high-precision battery internal resistance testing equipment, electrolyte state monitoring equipment, and vehicle-mounted sensors and a navigation system to capture real-time data on the internal resistance, electrolyte state, driving data, and environmental data of a second sample battery. After data acquisition, preprocessing steps such as data cleaning, outlier detection, and feature selection filter invalid or abnormal data to construct a second sample state dataset, ensuring the quality and accuracy of the model training data. The collected sample battery internal resistance, electrolyte state, driving data, and environmental data are integrated to construct second sample state data, including battery internal resistance time series, electrolyte density change records, driving condition statistics, and environmental condition monitoring values, comprehensively reflecting the battery's operating status and external environmental conditions. Using professional personnel or existing testing technologies, in-depth testing of the second sample battery is conducted to obtain the actual battery depletion status, including whether it is depleted, the degree of depletion, and possible causes of depletion, which serves as label data for model training to evaluate the model's predictive accuracy. The second sample state data is input into the second initial prediction model for training. The model may adopt a deep neural network architecture, such as a convolutional neural network or a recurrent neural network. Through multiple rounds of iteration and improved algorithms (such as gradient descent and Adam improver), the model parameters are continuously adjusted to ensure that the error between the predicted power depletion detection result and the sample power depletion detection result is minimized, thus obtaining the second prediction model with improved performance.

[0150] To verify the effectiveness of the second prediction model, independent second sample state data can be input into the model to evaluate its accuracy in predicting battery depletion risk. After verification, the second prediction model can accurately predict battery depletion risk and can be deployed in the vehicle's intelligent diagnostic system for real-time monitoring and early warning of battery depletion, improving vehicle safety and usage efficiency.

[0151] This application can construct a second prediction model based on multi-dimensional data to predict the risk of battery depletion. It can effectively identify and predict the possible depletion of batteries under specific operating conditions, provide users with timely early warning information and preventive measures, significantly improve the intelligent management and user experience of new energy vehicles, and avoid driving safety problems and vehicle downtime caused by battery depletion. It is an important supplement and improvement to the intelligent vehicle prevention and maintenance system.

[0152] This application primarily addresses the issue of vehicle battery depletion by constructing an intelligent diagnostic and prevention system based on big data analytics to improve vehicle reliability and convenience. The application includes core modules such as multi-dimensional data acquisition, big data preprocessing, deep modeling analysis, and diagnostic warning execution. Through an onboard sensor network system, it collects real-time battery status, electrical equipment operating parameters, vehicle driving data, and environmental information at high frequency and precision, uploading this data to a cloud platform via 5G encrypted communication technology. In the cloud, multiple algorithms are used to perform data cleaning, standardization, dimensionality reduction, and feature engineering processing. Combined with deep association rule mining algorithms, multi-level classification prediction models, and reinforcement learning time series analysis techniques, the system accurately identifies the causes of battery depletion and predicts risks in advance. When a risk is triggered, the system pushes warning information through multiple terminals (in-vehicle screen / mobile terminal / voice interaction), clearly informing users of the risk level and cause. Simultaneously, it automatically and remotely manages high-power, non-essential equipment to reduce static power consumption and intelligently matches suitable repair resources based on the cloud database to complete service appointments. This application achieves full-process automation and intelligence in battery depletion diagnosis, risk warning, and proactive prevention, significantly improving the overall efficiency of solving vehicle battery depletion problems.

[0153] This application describes the process from three aspects: data collection, big data platform construction, and diagnosis, early warning, and prevention.

[0154] The deployment of the data acquisition system includes the installation of battery sensors, the modification of electrical equipment circuits, the installation of driving status monitoring equipment, the deployment of environmental data acquisition modules, and the selection and configuration of communication modules.

[0155] In battery sensor installation, high-precision voltage, current, temperature, and internal resistance sensors can be embedded into the Battery Management System (BMS) during vehicle manufacturing or aftermarket retrofitting. For voltage sensors, a customized interface ensures a secure connection to the battery's positive and negative terminals and casing. Gold-plated interfaces and special sealing processes effectively prevent interface oxidation and electrolyte leakage, ensuring stable data transmission. Filtering circuits and anti-interference components are added to the sensor circuitry to reduce the impact of electromagnetic interference from onboard electrical equipment on the data. Furthermore, an independent backup power supply is provided for the sensors to ensure continuous operation and uninterrupted data acquisition when the vehicle's main power supply fails.

[0156] In the electrical equipment circuit upgrade, a novel micro data acquisition and power monitoring module was developed and integrated for each vehicle's electrical equipment to achieve intelligent circuit upgrades. Taking vehicle air conditioning as a typical application, a customized module was embedded in its control circuit board. Utilizing an improved Controller Area Network (CAN) bus communication protocol, high-precision monitoring of the equipment's operating status was achieved, recording air conditioning start-up and shutdown times, real-time cumulative operating time, and tracking power fluctuations with millisecond accuracy. This module adopts a low-power architecture design to minimize its own energy consumption and avoid additional load on the vehicle's electrical system. For high-end equipment with complex protocols, such as intelligent driver assistance systems, a specially developed adapter interface module was created to overcome communication barriers, deeply integrate into the system, and accurately collect core operating data such as camera operating time and radar module power consumption, comprehensively quantifying their impact on the battery. Furthermore, the upgraded electrical equipment circuit has a built-in intelligent diagnostic unit. Once a device fault, communication anomaly, or abnormal data fluctuation is detected, it immediately sends fault codes and status information to the vehicle's data center, helping maintenance personnel efficiently locate and resolve problems.

[0157] The driving status monitoring equipment is installed on the vehicle using a high-precision Global Positioning System (GPS) module, an accelerometer, and a gyroscope. The GPS module is positioned near the roof antenna to ensure good signal reception, enabling centimeter-level accuracy in mileage measurement and degree-level accuracy in real-time speed and direction acquisition. The accelerometer and gyroscope are fixed to the vehicle chassis or a rigid body component using dedicated shock-absorbing brackets to minimize the impact of driving vibrations on measurement accuracy. Utilizing a deep learning-based sensor fusion algorithm, the data collected by the equipment is combined with real-time map data. By comprehensively analyzing acceleration, speed, direction data, and map slope information, the system accurately identifies driving conditions such as uphill, downhill, and sharp turns. After installation, the equipment undergoes rigorous calibration testing to ensure the accuracy and consistency of data from all sensors. Regular maintenance and updates are performed to adapt to different usage environments and vehicle technology development needs.

[0158] The environmental data acquisition module is deployed in a planned layout within low-interference, unobstructed areas outside the vehicle body. The exterior temperature sensor, employing a waterproof and dustproof probe, is installed below the front bumper to effectively avoid interference from engine cooling and ensure accurate ambient temperature measurement. Humidity and light intensity sensors are mounted on the side of the roof rack to guarantee accurate data acquisition. An atmospheric pressure sensor is positioned on the side of the vehicle near the tires to effectively monitor air pressure fluctuations caused by changes in vehicle height. A high-sensitivity electromagnetic ambient intensity sensor based on the fluxgate principle is installed under the vehicle, away from metal components, to capture subtle changes in electromagnetic signals. Additionally, acid rain and alkalinity detection devices, as well as monitoring devices for the concentrations of harmful chemicals such as sulfur dioxide and nitrogen oxides, are added near the vehicle's air intake to collect key environmental data affecting battery performance in real time. A regular calibration and maintenance mechanism is established to ensure the accuracy and reliability of data acquisition. A dedicated data fusion algorithm has been developed to deeply integrate and analyze multi-source environmental sensor data, providing detailed and accurate environmental parameters for subsequent big data processing.

[0159] For communication module selection and configuration, vehicle-mounted communication modules supporting 5G and above communication standards are chosen to ensure data transmission quality through their high reliability and low latency. Based on the actual vehicle usage scenarios, parameters such as communication frequency band and transmission power are dynamically adjusted. For example, in urban environments with strong signal interference, signal strength stability is ensured by improving transmission power and selecting anti-interference frequency bands. Quantum key distribution encryption technology is employed to perform high-strength encryption processing on transmitted data, effectively preventing data leakage and tampering risks. Simultaneously, a real-time communication link monitoring system is established to continuously monitor key indicators such as signal strength and bit error rate. Upon detecting a communication anomaly, it automatically switches to backup links such as satellite communication to maintain data transmission. Furthermore, the communication module undergoes regular software upgrades to adapt to evolving communication technologies and changing data transmission requirements. Comprehensive compatibility and stability testing is conducted before the vehicle leaves the factory to ensure the communication module operates flawlessly in conjunction with other vehicle systems.

[0160] The construction of a big data platform includes cloud architecture building, algorithm library deployment and improvement, and visualization tool integration.

[0161] A cloud-based big data processing platform is built, employing distributed storage technology to securely store massive amounts of data. Through a multi-node redundancy backup mechanism, critical data is distributed across different physical nodes, effectively mitigating the risk of data loss due to single points of failure. A distributed computing framework is adopted to achieve parallel and efficient data processing; Kubernetes (Kubernetes) is used for cluster management and dynamic resource scheduling, flexibly allocating computing and storage resources based on task priority and resource availability, significantly improving data processing efficiency. The platform architecture is designed with scalability in mind; as the number of vehicles and the amount of data increase, compute nodes and storage devices can be easily added to the Kubernetes cluster, achieving linear scaling of platform performance. Simultaneously, regular performance improvements and security audits are conducted to ensure stable system operation and controllable data security.

[0162] For algorithm library deployment and improvement, a fully functional algorithm system is built within the big data platform, integrating core algorithm modules such as data cleaning, feature engineering, deep association analysis, multi-level classification prediction, and reinforcement learning time series prediction. A modular, plug-in architecture design is adopted to support flexible algorithm invocation and rapid upgrades, facilitating adaptation to diverse project needs. Algorithm performance is improved through parallel computing and distributed processing technologies; for example, data block parallelism significantly improves data cleaning and preprocessing efficiency. Innovative achievements such as the Transformer architecture are regularly integrated into the algorithm library, such as upgrading the reinforcement learning time series prediction module to further improve prediction accuracy. A full-process algorithm evaluation mechanism is established, dynamically adjusting algorithm parameters and model structure based on historical data simulation tests and real-world application feedback to enhance the algorithm's generalization ability and scenario adaptability. During deployment, unified data interface specifications and format standards ensure efficient collaboration and seamless data flow between algorithm modules.

[0163] For visualization tool integration, data visualization tools are introduced to transform platform analysis results into intuitive and easy-to-understand charts and graphs, helping users quickly interpret the key information behind the data. The visualization interface focuses on presenting core data such as battery power change trends, dynamic evolution of low-power risk levels, and power consumption distribution of various electrical devices. For example, line charts accurately show battery voltage fluctuations over time, and bar charts visually compare vehicle energy consumption differences under different road conditions. The system supports personalized customization, allowing users to flexibly choose display indicators and chart types according to their needs. A data interaction module has also been developed, allowing users to instantly obtain detailed data and in-depth analysis results by clicking on chart elements. For instance, clicking on the battery power curve will display battery parameters and low-power risk factors for the corresponding time period. During the integration process, the simplicity and ease of use of the interface design were highly valued. By improving the interaction logic and visual presentation, it is ensured that users with different technical backgrounds can use it conveniently.

[0164] The deployment of the power loss diagnosis, early warning and prevention system includes setting up the diagnosis and early warning triggering system, building a system for accurate cause tracing, implementing a multi-channel early warning push system, and deploying a remote energy-saving control and intelligent appointment system for maintenance services.

[0165] For the real-time diagnostic and early warning triggering system setup, an edge computing module is integrated into the vehicle terminal. This module automatically activates upon vehicle startup, collecting key data such as battery voltage, current, and the operating status of major electrical equipment in real time. Using a pre-trained simple model, it quickly analyzes the data. If a suspected low-battery risk is detected, such as a sudden drop in battery voltage exceeding a set threshold, detailed data is immediately uploaded to the cloud-based big data platform via the vehicle communication module. The cloud platform efficiently preprocesses and performs deep feature engineering on the data, then simultaneously performs deep analysis at the millisecond level using deep association rule mining, multi-level classification prediction, and reinforcement learning time series prediction models. If a strong combination of complex factors strongly correlated with low-battery status is detected, the vehicle is determined to be at a low-battery risk level, or the predicted battery level will rapidly decrease, the system triggers a low-battery warning within seconds. Simultaneously, a dynamic adjustment mechanism for the warning threshold is established, automatically adjusting the threshold based on vehicle usage history and environmental changes to improve the accuracy and timeliness of the warnings.

[0166] A precise cause tracing system was built, employing advanced technologies such as fault tree analysis based on Bayesian networks. Upon detecting a power loss event, the system immediately traces the root cause of the power loss precisely from massive amounts of data based on in-depth analysis. For example, by analyzing the operating status data and power change curves of electrical equipment, combined with changes in battery parameters, it accurately pinpoints a short circuit or leakage problem in a tiny component within the equipment. A fault cause knowledge base was built, organizing and storing various causes of power loss, corresponding fault symptoms, and diagnostic methods. If existing algorithms cannot accurately trace the cause of power loss, the system automatically searches the fault case library and analyzes similar cases. Simultaneously, leveraging the knowledge reasoning capabilities of an expert system composed of vehicle engineers and data analysts, complex faults can be manually analyzed. Experts can remotely connect to the platform for in-depth judgment of faults that the system cannot resolve. During the system's development, the fault cause knowledge base and case library were continuously updated and improved to enhance the system's diagnostic capabilities for various causes of power loss. Furthermore, a visual fault tracing interface was developed to intuitively display the tracing process and results to users, facilitating understanding and appropriate action.

[0167] A multi-channel early warning push system has been implemented, integrating an in-vehicle communication module, vehicle dashboard, and mobile application. It employs diverse methods such as pop-ups, voice broadcasts, and vibration alerts to deliver real-time and accurate low battery warning information. Warning content is presented in a tiered manner, clearly indicating risk levels such as mild and moderate, and simultaneously pushing specific causes, such as "abnormal operation of the in-vehicle entertainment system causing continuous power consumption" or "loose alternator belt causing decreased charging efficiency." Furthermore, dynamic line graphs display battery power trends, and heatmaps visually present the risk evolution process, enhancing information visualization. Based on personalized data such as driver habits and usage scenarios, the system intelligently matches customized response strategies. For example, it recommends regular deep charging for drivers who frequently drive short distances and pushes energy-saving headlight settings suggestions to nighttime drivers. During development, multiple rounds of stress testing were conducted to simulate different usage scenarios, continuously improving the push logic and content presentation to ensure the timeliness, accuracy, and readability of warning information delivery.

[0168] The deployment of a remote energy-saving control and intelligent appointment system for maintenance services involves developing an intelligent remote energy-saving control system for remotely operated electrical devices such as ambient lighting, car audio, and air conditioning. When the system detects that the vehicle is at risk of running out of power, it immediately activates an energy-saving strategy based on a multi-objective improvement algorithm, automatically and remotely shutting down unnecessary high-energy-consuming devices. For example, when the battery level is below a safe threshold, it automatically turns off ambient lighting, lowers the audio volume to mute, and switches the air conditioning to low-power mode. Simultaneously, an intelligent charging strategy improvement module is built, dynamically adjusting charging time and power parameters based on battery health status, real-time grid price fluctuations, and the owner's historical driving habits. This enables automatic charging during off-peak hours at night and precise control of charging power based on remaining battery power and the next day's travel needs, significantly reducing charging costs while ensuring a full charge. For hardware issues such as charging system malfunctions, a trusted IoT platform built using blockchain technology intelligently matches nearby certified vehicle repair stations and automatically pushes detailed repair work orders containing fault codes, fault locations, cause analysis, and real-time vehicle location. The system integrates big data analytics and user feedback to provide car owners with visualized decision-making support, including service station ratings, estimated repair costs, and project timelines. During system deployment, a two-way real-time communication mechanism ensures synchronized information updates with service stations and tracks repair progress throughout the entire process, promptly sending repair status updates to car owners and comprehensively improving service response efficiency and user experience.

[0169] Figure 2 This is a schematic diagram of a vehicle battery depletion prediction system according to an embodiment of this application, as shown below. Figure 2The deployment of the data acquisition system can include battery sensor integration, electrical equipment circuit modification, installation of driving status monitoring equipment, deployment of environmental data acquisition modules, selection and configuration of communication modules, and construction and algorithm deployment of a big data platform. The construction and algorithm deployment of the big data platform includes cloud architecture construction, algorithm library deployment and improvement, integration of visualization tools, and deployment of a power shortage diagnosis, early warning and prevention execution system. The deployment of the power shortage diagnosis, early warning and prevention execution system includes setting up a real-time diagnosis and early warning triggering system, building a system for accurate cause tracing, implementing a multi-channel early warning push system, and deploying a remote energy-saving control and intelligent appointment system for maintenance services.

[0170] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data used for analysis, data stored, data displayed, etc.) involved in this application are all information and data authorized by the user or fully authorized by all parties. Furthermore, the collection, use and processing of the relevant data must comply with the relevant laws, regulations and standards of the relevant countries and regions, and corresponding operation portals are provided for users to choose to authorize or refuse.

[0171] Figure 3 This is a schematic diagram of a vehicle battery depletion detection device according to an embodiment of this application, as shown below. Figure 3 As shown, the device includes the following: acquisition module 302, status detection module 304, power depletion detection module 306, and generation module 308.

[0172] The system comprises the following modules: a data acquisition module, which, in response to receiving a low battery detection command from the vehicle, acquires battery status data, power consumption data of electrical devices in the vehicle, vehicle driving status data, and environmental data of the vehicle's environment; a status detection module, which uses a first prediction model to perform status detection on the battery based on the low battery detection conditions and battery status data, and obtains a status detection result indicating whether the battery meets the low battery detection conditions; a low battery detection module, which, in response to the status detection result indicating that the battery meets the low battery detection conditions, uses a second prediction model to perform low battery detection on the battery based on power consumption data, driving status data, and environmental data, and obtains a low battery detection result indicating whether the battery is at risk of low battery status; and a generation module, which, in response to the low battery detection result indicating that the battery is at risk of low battery status, generates a low battery warning message based on the battery status data, power consumption data, driving status data, and environmental data.

[0173] Optionally, the generation module is used to locate the risk of battery depletion based on battery status data, power consumption data, driving status data, and environmental data, and determine the risk type corresponding to the risk of battery depletion; quantify the risk of battery depletion based on the risk type to obtain the risk level of the battery; and generate a battery depletion warning message based on the risk level and risk type.

[0174] Optionally, the generation module is used to acquire vehicle scenario data, wherein the vehicle scenario data is used to represent vehicle usage data under different scenarios; based on the vehicle scenario data and risk type, the low battery warning scenario is determined; and based on the low battery warning scenario, risk type, and risk level, low battery warning information is generated.

[0175] Optionally, the device is used to sort the power consumption of multiple devices corresponding to the vehicle to obtain a power consumption sorting result, wherein different devices are used to implement different functions of the vehicle; determine at least one target device based on the power consumption sorting result, and generate control instructions for at least one target device; send the control instructions to at least one target device, and receive control results from at least one target device, wherein the control results are used to indicate whether at least one target device has been successfully turned off.

[0176] Optionally, the device is used to acquire the vehicle's current location information and / or vehicle navigation information; based on the current location information and / or vehicle navigation information, determine the network information of the target service network, wherein the target service network is used to provide battery repair services; and in response to receiving a confirmation instruction for the network information, generate a navigation path for the target service network.

[0177] Optionally, the device is used to acquire sample voltage data, sample current data, and sample temperature data of the first sample battery; based on the sample voltage data, sample current data, and sample temperature data, construct first sample state data and sample state detection results of the first sample battery; input the first sample state data into a first initial prediction model, use the first initial prediction model to perform state detection on the first sample battery, and obtain a predicted state detection result; based on the sample state detection result and the predicted state detection result, adjust the model parameters of the first initial prediction model to obtain a first prediction model.

[0178] Optionally, the device is used to acquire the sample battery internal resistance, sample electrolyte state, sample driving data, and sample environment data of the second sample battery, wherein the sample driving data is the driving data of the vehicle where the second sample battery is located, and the sample environment data is the environmental data of the vehicle where the second sample battery is located; based on the sample battery internal resistance, sample electrolyte state, sample driving data, and sample environment data, second sample state data and sample depletion detection results of the second sample battery are constructed; the second sample state data is input into the second initial prediction model, and the second initial prediction model is used to perform depletion detection on the second sample battery to obtain the predicted depletion detection result; based on the sample depletion detection result and the predicted depletion detection result, the model parameters of the second initial prediction model are adjusted to obtain the second prediction model.

[0179] Figure 4 This is a schematic diagram of a vehicle battery depletion detection system according to an embodiment of this application, as shown below. Figure 4 As shown, the system includes: a data acquisition module 402, a first detection module 404, a second detection module 406, and a notification module 408.

[0180] The system includes: a data acquisition module 402, which, in response to receiving a low-power detection command from the vehicle, acquires battery status data, power consumption data of electrical devices in the vehicle, driving status data, and environmental data of the vehicle's environment; a first detection module 404, deployed locally in the vehicle 410, uses a first prediction model to perform battery status detection based on low-power detection conditions and battery status data, obtaining a status detection result indicating whether the battery meets the low-power detection conditions; in response to the status detection result indicating that the battery meets the low-power detection conditions, it sends power consumption data, driving status data, and environmental data to the second detection module; a second detection module 406, deployed in the cloud 412 corresponding to the vehicle, uses a second prediction model to perform low-power detection on the battery based on power consumption data, driving status data, and environmental data, obtaining a low-power detection result indicating whether the battery is at risk of low power; and a warning module 408, which, in response to the low-power detection result indicating that the battery is at risk of low power, generates a low-power warning message based on battery status data, power consumption data, driving status data, and environmental data.

[0181] Embodiments of this application also provide a vehicle, including: a memory storing an executable program; and a processor for running the program, wherein the program executes the methods described in various embodiments of this application when it runs.

[0182] Embodiments of this application also provide a computer-readable storage medium including a stored executable program, wherein, when the executable program is running, it controls the device where the computer-readable storage medium is located to perform the methods of various embodiments of this application.

[0183] Embodiments of this application also provide a computer program product, including a computer program that, when executed by a processor, implements the methods of various embodiments of this application.

[0184] Embodiments of this application also provide a computer program product, including a non-volatile computer-readable storage medium for storing a computer program that, when executed by a processor, implements the methods in various embodiments of this application.

[0185] Embodiments of this application also provide a computer program that, when executed by a processor, implements the methods described in the various embodiments of this application.

[0186] In the above embodiments of this application, the descriptions of each embodiment have different focuses. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments.

[0187] In the several embodiments provided in this application, it should be understood that the disclosed technical content can be implemented in other ways. The device embodiments described above are merely illustrative; for example, the division of units can be a logical functional division, and in actual implementation, there may be other division methods. For instance, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the displayed or discussed mutual coupling, direct coupling, or communication connection may be through some interfaces; the indirect coupling or communication connection between units or modules may be electrical or other forms.

[0188] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple units. Some or all of the units can be selected to achieve the purpose of this embodiment according to actual needs.

[0189] Furthermore, the functional units in the various embodiments of this application can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0190] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of this application. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, read-only memory (ROM), random access memory (RAM), portable hard drives, magnetic disks, or optical disks.

[0191] The above description is only a preferred embodiment of this application. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of this application, and these improvements and modifications should also be considered within the scope of protection of this application.

Claims

1. A method for detecting battery depletion in a vehicle, characterized in that, include: In response to receiving a low battery detection command from the vehicle, the system collects battery status data of the vehicle's battery, power consumption data of the vehicle's electrical devices, driving status data of the vehicle, and environmental data of the environment in which the vehicle is located. The battery is state-detected using a first prediction model based on the low-charge detection conditions and the battery state data, and a state detection result is obtained, wherein the state detection result is used to indicate whether the battery meets the low-charge detection conditions. In response to the state detection result indicating that the battery meets the low-power detection condition, a second prediction model is used to perform low-power detection on the battery based on the power consumption data, the driving state data, and the environmental data to obtain a low-power detection result, wherein the low-power detection result is used to indicate whether the battery is at risk of low power. In response to the low battery detection result indicating that the battery is at risk of low battery, a low battery warning message is generated based on the battery status data, the power consumption data, the driving status data, and the environmental data.

2. The battery depletion detection method for a vehicle according to claim 1, characterized in that, Based on the battery status data, power consumption data, driving status data, and environmental data, a low battery warning message is generated, including: Based on the battery status data, the power consumption data, the driving status data, and the environmental data, fault location is performed on the risk of low battery power, and the risk type corresponding to the risk of low battery power is determined. The risk of battery depletion is quantified based on the risk type to obtain the risk level of the battery; The power depletion warning message is generated based on the risk level and the risk type.

3. The battery depletion detection method for a vehicle according to claim 2, characterized in that, Based on the risk level and the risk type, the power depletion warning information is generated, including: Obtain vehicle scene data for the vehicle, wherein the vehicle scene data is used to represent the vehicle's usage data under different scenarios; Based on the vehicle scenario data and the risk type, a low battery warning scenario is determined; The low battery warning information is generated based on the low battery warning scenario, the risk type, and the risk level.

4. The method for detecting battery depletion in a vehicle according to claim 1, characterized in that, The method further includes: The power consumption of multiple devices corresponding to the vehicle is sorted to obtain a power consumption sorting result, wherein different devices are used to implement different functions of the vehicle; Based on the power consumption ranking results, at least one target device is determined, and control instructions for the at least one target device are generated; The control command is sent to the at least one target device, and the control result of the at least one target device is received, wherein the control result is used to indicate whether the at least one target device has been successfully shut down.

5. The method for detecting battery depletion in a vehicle according to claim 1, characterized in that, In response to the result of the low-charge detection indicating that the battery has the risk of low-charge, the method further includes: Obtain the current location information and / or vehicle navigation information of the vehicle; Based on the current location information and / or the vehicle navigation information, the network information of the target service point is determined, wherein the target service point is used to provide repair services for the battery; In response to receiving a confirmation command for the network information, a navigation path for the target service network is generated.

6. The battery depletion detection method for a vehicle according to any one of claims 1 to 5, characterized in that, The method further includes: Obtain the sample voltage data, sample current data, and sample temperature data of the first sample battery; Based on the sample voltage data, the sample current data, and the sample temperature data, the first sample state data of the first sample battery and the sample state detection result of the first sample state data are constructed. The first sample state data is input into the first initial prediction model, and the state of the first sample battery is detected using the first initial prediction model to obtain the predicted state detection result. Based on the sample state detection results and the predicted state detection results, the model parameters of the first initial prediction model are adjusted to obtain the first prediction model.

7. The battery depletion detection method for a vehicle according to any one of claims 1 to 5, characterized in that, The method further includes: The sample battery internal resistance, sample electrolyte state, sample driving data, and sample environment data of the second sample battery are obtained, wherein the sample driving data is the driving data of the vehicle in which the second sample battery is located, and the sample environment data is the environmental data of the vehicle in which the second sample battery is located. Based on the internal resistance of the sample battery, the state of the sample electrolyte, the sample driving data, and the sample environmental data, the second sample state data of the second sample battery and the sample depletion detection result of the second sample state data are constructed. The second sample state data is input into the second initial prediction model, and the second initial prediction model is used to detect the low charge of the second sample battery to obtain the predicted low charge detection result. Based on the sample power depletion detection results and the predicted power depletion detection results, the model parameters of the second initial prediction model are adjusted to obtain the second prediction model.

8. A vehicle battery depletion detection system, characterized in that, include: The data acquisition module, in response to receiving a low battery detection command from the vehicle, acquires battery status data of the vehicle's battery, power consumption data of the vehicle's electrical devices, driving status data of the vehicle, and environmental data of the environment in which the vehicle is located. The first detection module, deployed locally in the vehicle, uses a first prediction model to perform state detection on the battery based on the low-power detection conditions and the battery state data, and obtains a state detection result, wherein the state detection result is used to indicate whether the battery meets the low-power detection conditions; in response to the state detection result indicating that the battery meets the low-power detection conditions, the power consumption data, the driving state data, and the environmental data are sent to the second detection module. The second detection module is deployed in the cloud corresponding to the vehicle. It uses a second prediction model to detect the battery's low charge based on the power consumption data, the driving status data, and the environmental data, and obtains a low charge detection result. The low charge detection result is used to indicate whether the battery is at risk of low charge. The prompting module is used to generate a low battery prompt message in response to the low battery detection result indicating that the battery is at risk of low battery, based on the battery status data, the power consumption data, the driving status data, and the environmental data.

9. A vehicle, characterized in that, include: Memory, which stores executable programs; A processor for running the program, wherein the program, when running, performs the method according to any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium includes a stored executable program, wherein, when the executable program is executed, it controls the device on which the storage medium is located to perform the method according to any one of claims 1 to 7.

Images