Vehicle emergency linkage alarm and safe standby power system and method based on AI prediction

Abstract

The invention discloses a vehicle emergency linkage alarm and safety standby power system and method based on AI prediction, and relates to the technical field of safety warning, and the vehicle emergency linkage alarm and safety standby power method based on AI prediction comprises the following steps: fusing multi-mode sensor data of a vehicle and physiological monitoring information of a driver; the AI prediction and danger identification model based on multi-modal deep learning is constructed, potential risks are identified in advance, the danger level is dynamically adjusted by combining whether the driver is disabled, and the timeliness and accuracy of risk early warning are remarkably improved. And when the system judges that the driver is disabled or the danger level reaches a threshold value, the alarm level in the vehicle is automatically upgraded, and a safe standby power system is activated in a linkage manner, so that the safety function can be maintained and an emergency signal can be sent before a main power supply fails or at the moment of an accident, and the active warning capability and the passenger survival guarantee level of the vehicle under an extreme working condition are greatly enhanced. Therefore, the vehicle safety warning effect can be improved.

Description

Technical Field

[0001] This application relates to the field of safety warning technology, and in particular to a vehicle emergency linkage alarm and safety backup power system and method based on AI prediction. Background Technology

[0002] With the rapid development of intelligent connected vehicles and advanced driver assistance systems (ADAS), vehicle safety technology has gradually evolved from traditional passive safety to active and predictive safety. Currently, vehicle safety systems on the market generally rely on a single sensor (such as radar or camera) for environmental perception and trigger passive response mechanisms such as emergency braking and airbags when a collision is imminent.

[0003] However, such systems typically only intervene when a dangerous event has clearly occurred or is about to occur, lacking the ability to identify and classify potential risks in the early stages. Furthermore, warnings are often concentrated on in-vehicle buzzers or a single display screen, making it difficult for external road users to be informed in a timely manner. Once the vehicle's main power supply fails completely in an accident, the existing safety system will be unable to continue working, and alarm and rescue functions will be interrupted, posing a significant hazard and resulting in the poor effectiveness of current vehicle safety warnings. Summary of the Invention

[0004] The main purpose of this application is to provide a vehicle emergency linkage alarm and safety backup power system and method based on AI prediction, which aims to solve the technical problem of poor vehicle safety warning effect.

[0005] To achieve the above objectives, this application proposes an AI-predictive vehicle emergency linkage alarm and backup power system, which includes the following interconnected hardware modules:

[0006] The vehicle sensor module includes a camera, radar, inertial measurement unit, vehicle attitude sensor and vehicle control unit status interface, used to collect multimodal data in real time, wherein the multimodal data includes vehicle operating status and environmental information; The driver physiological monitoring module is used to monitor the driver's physiological data through steering wheel sensors, seat sensors and in-vehicle cameras in the vehicle; The AI ​​prediction and hazard identification module is used to predict potential hazards based on the multimodal data using a multimodal fusion deep learning model. The potential hazards include at least the risk of an impending collision, rollover, or abnormal vehicle energy system. The module also combines the physiological monitoring data to classify and dynamically adjust the hazard level and output the hazard level. The driver physiological monitoring module is also used to send a driver incapacity signal to the emergency control unit and automatically upgrade the vehicle's alarm level when the driver is detected to be unconscious, lost consciousness, or unable to perform effective vehicle operation based on the physiological monitoring data. The physiological monitoring data includes the driver's heart rate, respiration, and consciousness status, and the alarm level is determined based on the danger level.

[0007] In one embodiment, the system further includes a safety backup power module, comprising an independent battery cell, a backup power management circuit, and a power switching module, which is connected in parallel with the vehicle's main power supply via the backup power management circuit; The safety backup power module is used to switch the vehicle's power supply to backup power within a preset time window of 50ms to 200ms when the main power supply is determined to be damaged or completely disconnected. It provides emergency power to the vehicle's communication module, positioning module, automatic alarm and rescue module, and display and projection module in descending order of priority. The power supply strategy is dynamically adjusted according to the remaining power of the backup power supply and the level of danger to ensure the continuous operation of alarm, positioning and communication functions.

[0008] In one embodiment, the safety backup power module is further configured to: The system acquires the vehicle's backup power status, main power status, and environmental factors. It then uses the AI ​​prediction and hazard identification module or a risk prediction module that communicates with it to predict the risk coefficient and risk time window. Based on the hazard level change trend or prediction time window output by the AI ​​prediction and hazard identification module, the system determines to enter the pre-charging state when the risk coefficient is higher than a preset risk coefficient threshold and the risk time window is lower than a preset time window threshold, or when the voltage change rate of the backup power supply is detected to be higher than a preset change threshold and the environmental factor is a hazardous environmental factor, or when a manual pre-charging command is received. After entering the pre-charging state, the status of the backup power supply is read to determine whether it needs to be recharged. If it is needed, the charging path is controlled to open or close, and the main power supply or energy recovery path is scheduled to recharge the backup power supply. After the recharge is completed, the backup power supply is switched to the ready-to-activate state. After the backup power supply is in the ready and activated state, its internal resistance, terminal voltage and temperature are measured to confirm that the backup power supply meets the conditions for quick takeover in the event of an accident. If there is no abnormality in the backup power supply, it is checked in conjunction with the vehicle control module to replenish the backup power supply. After the replenishment is completed, it is determined that the backup power supply is ready and the vehicle's power supply is switched to the backup power supply.

[0009] In one embodiment, the system further includes a multi-display device linkage module and an AR vehicle exterior projection and multi-terminal distress prompt module; the AI ​​prediction and hazard identification module further includes an emergency control unit. The emergency control unit is used to receive the physiological monitoring data, determine the danger level, generate an emergency control command based on the danger level and the physiological monitoring data, generate a corresponding control signal based on the emergency control command, trigger different levels of display warnings, projection prompts and alarm linkage strategies through the vehicle communication bus, and send the control signal or the display warnings, projection prompts and alarm linkage strategies to the multi-display device linkage module, the automatic alarm and rescue module and the safety backup power module. The multi-display device linkage module is used to uniformly control the vehicle's instrument panel, central control screen, vehicle HUD, rear entertainment screen and window display, and form synchronous warning output of multiple terminals in the vehicle under the same hazard level control strategy. When receiving the display warning, the projection prompt and the alarm linkage strategy, it outputs warning information, distress information and accident information in a unified format with different colors and contents. The AR vehicle exterior projection and multi-terminal distress signal module is used to project warning signs onto the ground, outside the vehicle, or on the windows based on the vehicle's exterior projection lights, body LED screens, or transparent window display units. It also coordinates with rearview or side mirror projections to form multi-position collaborative displays, improving the visibility to surrounding traffic participants. The distress signal or warning information is projected onto the rearview or side mirrors. When it is determined that the driver cannot operate the steering wheel or the vehicle's central control system, an automatic alarm is triggered via voice recognition, simultaneously activating window projection, mirror projection, and exterior projection prompts. Based on weather conditions during vehicle operation, including nighttime, rain, fog, or low-visibility environments, corresponding visual and voice warnings are provided, and the display content, color, and flashing frequency are automatically adjusted according to the level of danger.

[0010] In one embodiment, the system further includes: The automatic alarm and rescue module is used to automatically dial an alarm number or send alarm and distress information to the cloud platform, alarm center, emergency contact, vehicle manufacturer or insurance company when the danger level reaches the preset maximum danger level or the driver physiological monitoring module detects that the driver is unconscious or lost consciousness, via cellular network or vehicle network.

[0011] This application also proposes an AI-predictive vehicle emergency linkage alarm and backup power method, applied to an AI-predictive vehicle emergency linkage alarm and backup power system. The method includes: Real-time acquisition of multimodal data, including vehicle operating status and environmental information; The vehicle monitors the driver's physiological data through steering wheel sensors, seat sensors, and in-vehicle cameras. Based on the multimodal data, a preset deep learning model is used to predict potential hazards. When the predicted hazard level is higher than a preset threshold, the safety backup power system is controlled to enter a pre-charging or ready-to-activate state. The hazard level is then classified and dynamically adjusted in conjunction with the physiological monitoring data, and the hazard level is output. When the driver is detected to be unconscious or unable to operate the vehicle based on the physiological monitoring data, the vehicle's alarm level is automatically upgraded. The physiological monitoring data includes the driver's heart rate, respiration, and consciousness status, and the alarm level is determined based on the danger level.

[0012] In one embodiment, the AI-based vehicle emergency linkage alarm and backup power method further includes: When it is determined that the vehicle's main power supply is damaged or completely cut off, the vehicle's power supply will be switched to the backup power supply within a preset time. Emergency power will be provided to the vehicle's communication module, positioning module, automatic alarm and rescue module and display and projection module in order of priority from high to low to ensure the continuous operation of alarm, positioning and communication functions. Non-core function modules will be shut down step by step according to the remaining power of the backup power supply.

[0013] In one embodiment, the step of switching the vehicle's power supply to a backup power supply within a preset time period when it is determined that the main power supply is damaged or completely lost includes: The system acquires the vehicle's backup power status, main power status, and environmental factors. It predicts risk coefficients and risk time windows through a preset risk prediction module, and combines the risk level change trend or prediction time window output by the AI ​​prediction and hazard identification module. When the risk coefficient is higher than a preset risk coefficient threshold and the risk time window is lower than a preset time window threshold, or when the voltage change rate of the backup power is detected to be higher than a preset change threshold and the environmental factor is a hazardous environmental factor, or when a manual pre-charging command is received, the system determines to enter the pre-charging state. After entering the pre-charging state, the status of the backup power supply is read to determine whether it needs to be recharged. If it is needed, the charging path is controlled to open or close, and the main power supply or energy recovery path is scheduled to recharge the backup power supply. After the recharge is completed, the backup power supply is switched to the ready-to-activate state. After the backup power supply is in the ready and activated state, its internal resistance, terminal voltage and temperature are measured to confirm that the backup power supply meets the conditions for quick takeover in the event of an accident. The backup power supply is checked for any abnormalities. If there are no abnormalities, the backup power supply is recharged in conjunction with the vehicle control module. After the recharge is completed, the backup power supply is confirmed to be ready, and the vehicle's power supply is switched to the backup power supply.

[0014] In one embodiment, after the step of monitoring the driver's physiological data via the steering wheel sensor, seat sensor, and in-vehicle camera, the method further includes: The system receives the physiological monitoring data, determines the danger level, generates an emergency control command based on the danger level and the physiological monitoring data, generates a corresponding control signal based on the emergency control command, triggers different levels of display warnings, projection prompts and alarm linkage strategies through the vehicle communication bus, and sends the control signal or the display warnings, projection prompts and alarm linkage strategies to the multi-display device linkage module, the automatic alarm and rescue module and the safety backup power module. The system provides unified control over the vehicle's dashboard, central control screen, in-vehicle HUD, rear entertainment screen, and window displays. Upon receiving the display warnings, projection prompts, and alarm linkage strategies, it outputs warning messages, distress messages, and accident information in a unified format with different colors and contents. Based on the vehicle's external projection lights, body LED screen, or transparent window display unit, warning signs are projected onto the ground, outside the vehicle, or on the windows. The distress message or warning message is projected onto the vehicle's rearview mirror or side mirror. When it is determined that the driver cannot operate the steering wheel or the vehicle's central control, an automatic alarm is triggered through voice recognition, and the window and mirror projections and external projection prompts are activated simultaneously. Based on the weather conditions during vehicle operation, corresponding visual and voice warnings are given, and the display content, color, and flashing frequency are automatically adjusted according to the level of danger.

[0015] In one embodiment, after the steps of predicting potential hazards based on the multimodal data using a preset deep learning model, controlling the safety backup power system to enter a pre-charging or ready-to-activate state when the predicted hazard level is higher than a preset threshold, and combining the physiological monitoring data to perform hazard level classification and dynamic adjustment, and outputting the hazard level, the method further includes: When the danger level reaches the preset maximum danger level, or when the driver is detected to be unconscious, passengers are allowed to trigger an automatic alarm process via voice recognition commands, and simultaneously activate in-vehicle and out-of-vehicle warnings and projection prompts. The system can automatically dial an alarm number or send alarm and distress information to the cloud platform, alarm center, emergency contact, vehicle manufacturer, or insurance company via cellular network or vehicle network.

[0016] One or more technical solutions proposed in this application have at least the following technical effects: This application integrates multimodal sensor data from vehicle cameras, radar, inertial measurement units, body attitude sensors, and control unit status lights with driver physiological monitoring information such as heart rate, respiration, and consciousness status to construct an AI prediction and hazard identification model based on multimodal deep learning. This model can identify potential risks such as collisions, rollovers, or energy system anomalies in advance and dynamically adjust the hazard level based on whether the driver is incapacitated, thereby significantly improving the timeliness and accuracy of risk warnings.

[0017] Furthermore, when the system determines that the driver is incapacitated or the danger level reaches a threshold, this application not only automatically upgrades the in-vehicle alarm level but also activates the safety backup power system. This ensures that critical safety functions such as the communication module, external warning lights, and door lock control remain operational until the main power fails or at the moment of an accident. It also sends emergency signals to nearby vehicles and rescue centers via external broadcasting or a network platform. This effectively solves the problem in existing technologies where alarms are interrupted due to power outages, and external road users are unable to be informed of the danger in a timely manner. It significantly enhances the vehicle's proactive warning capabilities and occupant survival protection level under extreme conditions. Therefore, this application can improve the effectiveness of vehicle safety warnings. Attached Figure Description

[0018] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0019] To more clearly illustrate the technical solutions in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, for those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0020] Figure 1 This is a schematic diagram of the overall system structure provided in Embodiment 1 of the AI-predictive vehicle emergency linkage alarm and backup power method of this application; Figure 2 This is a schematic diagram of the AI ​​prediction and hazard identification module structure provided in Embodiment 1 of the AI-based vehicle emergency linkage alarm and backup power method of this application. Figure 3 This is a block diagram of multi-terminal linkage inside and outside the vehicle provided in Embodiment 1 of the vehicle emergency linkage alarm and backup power method based on AI prediction in this application; Figure 4 This is a diagram showing the linkage between window display and mirror projection provided in Embodiment 1 of the AI-predicted vehicle emergency linkage alarm and backup power method of this application. Figure 5 The automatic alarm flowchart provided in Embodiment 1 of the vehicle emergency linkage alarm and backup power method based on AI prediction in this application; Figure 6 This is a flowchart of a voice-controlled distress call provided in Embodiment 1 of the AI-predictive vehicle emergency linkage alarm and backup power method of this application. Figure 7 This is an integrated diagram of automatic alarm and multi-terminal linkage provided in Embodiment 1 of the vehicle emergency linkage alarm and backup power method based on AI prediction in this application; Figure 8 A diagram showing the backup power switching of the AI-predictive vehicle emergency linkage alarm and safety backup power method provided in Embodiment 2 of this application; Figure 9 This diagram illustrates the data acquisition consent process involved in the AI-predicted vehicle emergency linkage alarm and backup power method in this application embodiment.

[0021] The purpose, features, and advantages of this application will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0022] It should be understood that the specific embodiments described herein are merely illustrative of the technical solutions of this application and are not intended to limit this application.

[0023] To better understand the technical solution of this application, a detailed description will be provided below in conjunction with the accompanying drawings and specific implementation methods.

[0024] It should be noted that the executing entity in this embodiment can be a computing service device with data processing, network communication, and program execution functions, such as a tablet computer, personal computer, or mobile phone, or an electronic device capable of performing the above functions, such as an AI-predictive vehicle emergency linkage alarm and safety backup power device. The following description uses an AI-predictive vehicle emergency linkage alarm and safety backup power device as an example to illustrate this embodiment and the subsequent embodiments.

[0025] With the rapid development of intelligent connected vehicles and advanced driver assistance systems (ADAS), vehicle safety technology has gradually evolved from traditional passive safety to active and predictive safety. Currently, vehicle safety systems on the market generally rely on a single sensor (such as radar or camera) for environmental perception and trigger passive response mechanisms such as emergency braking and airbags when a collision is imminent.

[0026] However, such systems typically only intervene when a dangerous event has clearly occurred or is about to occur, lacking the ability to identify and classify potential risks in the early stages. Furthermore, warnings are often concentrated on in-vehicle buzzers or a single display screen, making it difficult for external road users to be informed in a timely manner. Once the vehicle's main power supply fails completely in an accident, the existing safety system will be unable to continue working, and alarm and rescue functions will be interrupted, posing a significant hazard and resulting in the poor effectiveness of current vehicle safety warnings.

[0027] Based on this, embodiments of this application provide a vehicle emergency linkage alarm and backup power system based on AI prediction, referring to... Figure 1 The AI-based vehicle emergency alarm and backup power system includes the following interconnected hardware modules: The vehicle sensor module includes a camera, radar, inertial measurement unit, vehicle attitude sensor and vehicle control unit status interface, used to collect multimodal data in real time, wherein the multimodal data includes vehicle operating status and environmental information; It should be noted that the vehicle sensor module refers to a hardware system composed of various sensors and their interfaces integrated on the vehicle, used to perceive the vehicle's own state and the external environment. Vehicle attitude sensors detect attitude parameters such as pitch, roll, and yaw angles during vehicle operation. They are typically based on gravity sensing or work in conjunction with an inertial measurement unit (IMU) to reflect the vehicle's stability and handling status. The vehicle control unit state interface refers to the standardized data interface for communication with the vehicle's electronic control unit (ECU), used to acquire low-level vehicle control signals and operating parameters such as vehicle speed, steering angle, braking status, and throttle opening. Multimodal data refers to a heterogeneous data set collected by different types of sensors (such as vision, radar, inertial, and vehicle status sensors), covering multiple dimensions including time, space, and semantics, and possessing complementarity and redundancy. Vehicle operating status is a set of parameters describing the vehicle's dynamic behavior, including but not limited to vehicle speed, acceleration, steering angle, braking status, and engine speed. Environmental information refers to perceived data of the vehicle's external surroundings, including road structure, obstacle locations, traffic participants, and weather and lighting conditions.

[0028] Understandably, by integrating multiple heterogeneous sensors and uniformly collecting multimodal data, this embodiment achieves high coverage and robust perception capabilities for the vehicle's internal and external states. Specifically, cameras provide rich semantic information but are susceptible to interference from lighting and weather; radar can still stably measure distance and speed in harsh environments; IMU and vehicle attitude sensors accurately capture the vehicle's instantaneous dynamics; and the vehicle control unit's state interface supplements the real feedback from the underlying actuators. This embodiment effectively overcomes the limitations of a single sensor through the above architecture, significantly improving the completeness and timeliness of the raw perception data.

[0029] The driver physiological monitoring module is used to monitor the driver's physiological data through steering wheel sensors, seat sensors and in-vehicle cameras in the vehicle; It should be noted that the driver physiological monitoring module refers to the subsystem used to collect and analyze parameters related to the driver's physiological state in real time. Steering wheel sensors are sensing devices installed on the steering wheel that can detect behavioral characteristics such as the driver's grip strength, steering operation frequency, and fine-tuning amplitude, indirectly reflecting their alertness or handling stability. Seat sensors are pressure, vibration, or capacitive sensors integrated into the seat, used to monitor physiological signals such as the driver's posture, body movement, heart rate, or respiratory rate. In-vehicle cameras are dedicated cameras pointed at the driver's face or upper body, usually equipped with infrared light sources, used to capture visual physiological indicators such as eye opening and closing, head posture, yawning, and gaze direction.

[0030] Understandably, this embodiment achieves continuous monitoring of the driver's state by fusing multi-source physiological signals from the steering wheel, seat, and in-vehicle camera. Since the sensors used in this embodiment are all embedded in conventional vehicle components, the driver does not need to wear additional equipment, improving user experience and system usability. Furthermore, different sensors provide complementary information from three levels: behavioral (e.g., steering fine-tuning), physiological (e.g., heart rate, respiration), and visual (e.g., duration of eye closure), effectively overcoming the problem of single-modal susceptibility to interference or misjudgment (for example, relying solely on eye movements might misjudge closed eyes due to bright light, while combining steering wheel operations allows for cross-verification). Therefore, this embodiment significantly improves the accuracy and robustness of driver state recognition, providing a reliable basis for subsequent safety strategies such as triggering warnings, takeover requests, or emergency braking, thereby enhancing the human-centric safety and human-machine collaboration efficiency of the intelligent driving system.

[0031] The AI ​​prediction and hazard identification module is used to predict potential hazards based on the multimodal data using a multimodal fusion deep learning model. The potential hazards include at least the risk of an impending collision, rollover, or abnormal vehicle energy system. The module also combines the physiological monitoring data to classify and dynamically adjust the hazard level and output the hazard level. It should be noted that multimodal fusion deep learning models are a type of neural network architecture that can effectively fuse data from different sensor modalities (such as images, radar point clouds, IMU signals, vehicle status, etc.) at the feature level or decision level to improve the model's ability to understand complex scenarios and its prediction accuracy. Potential hazards refer to driving risk events that have not yet occurred but can be inferred from current and historical states to have a high probability of occurrence; these fall under the category of proactive safety assessments.

[0032] Vehicle energy system anomaly risks refer to potential power system failures unique to new energy vehicles, such as overheating of the power battery, sudden voltage drops, and motor malfunctions, which may lead to secondary dangers such as power interruption or fire. The hazard level is the result of quantifying and classifying the identified potential hazards according to their urgency, severity, and probability of occurrence (e.g., low, medium, high, or numerical levels), used to guide subsequent response strategies. Hazard level classification and dynamic adjustment refers to the process of initially determining the hazard level based on real-time input data (including environmental perception and driver status) and continuously updating the level over time and with the influx of new data. The structure of the AI ​​prediction and hazard identification module in this embodiment can be referred to... Figure 2 .

[0033] Understandably, this embodiment combines multimodal environmental and vehicle data with the driver's physiological state to achieve a more accurate risk assessment of potential hazards. Specifically, the multimodal fusion deep learning model can fully utilize the complementary advantages of different sensors (such as cameras identifying lane departures, radar detecting nearby obstacles, and IMU sensing skid trends), thereby improving the prediction accuracy of physical hazards such as collisions and rollovers. Furthermore, by introducing physiological monitoring data, this embodiment can determine whether the driver has the ability to respond to hazards in a timely manner (in the same situation of sudden braking ahead, if the driver is fatigued, the probability of a delayed reaction is high, and the system will increase the hazard level), avoiding false alarms or missed alarms caused by traditional systems relying solely on environmental perception. Therefore, this embodiment can significantly improve the situational adaptability and decision-making rationality of hazard identification, providing a more targeted basis for subsequent warning intensity adjustment, takeover request timing selection, or automatic intervention strategies, enhancing the active safety performance and human-machine collaborative intelligence level of the intelligent driving system.

[0034] In one embodiment, the driver physiological monitoring module is further configured to send a driver incapacity signal to the emergency control unit and automatically escalate the vehicle's alarm level when the driver is detected to be unconscious, incapacitated, or unable to perform effective vehicle operation based on the physiological monitoring data. The physiological monitoring data includes the driver's heart rate, respiration, and consciousness status, and the alarm level is determined based on the danger level.

[0035] It should be noted that the driver incapacity signal is an emergency status indicator signal issued by the driver physiological monitoring module, indicating that the driver has lost effective control of the vehicle due to unconsciousness, loss of consciousness, or severe physiological abnormalities. The emergency control unit is a subsystem in the vehicle specifically responsible for taking over vehicle control or triggering safety response mechanisms in extreme or emergency situations. It can perform operations such as automatically pulling over, decelerating, activating hazard lights, and contacting roadside assistance. The alarm level is a warning level determined based on the current level of danger and the driver's condition, used to determine the intensity of the human-machine interface prompts (such as audible, vibration, and visual warnings) or whether to activate automatic intervention measures.

[0036] Understandably, this embodiment significantly enhances the system's emergency response capability by proactively identifying driver incapacitation. Specifically, when a driver loses capacity due to a sudden health problem, even if the external environmental hazard level has not yet reached its highest level (e.g., the vehicle is still traveling at a constant speed on a straight road), a traditional system might not trigger intervention. However, this embodiment, by monitoring heart rate, respiration, and consciousness in real time, can identify corresponding risks in advance and proactively escalate the alarm level accordingly. This ensures that the system initiates emergency control in the shortest possible time, greatly improving the vehicle's autonomous safety assurance capability in the event of sudden driver incapacitation and effectively reducing the risk of secondary accidents or injuries.

[0037] In one embodiment, the system further includes a multi-display device linkage module and an AR vehicle exterior projection and multi-terminal distress prompt module; the AI ​​prediction and hazard identification module further includes an emergency control unit. The emergency control unit is used to receive the physiological monitoring data, determine the danger level, generate an emergency control command based on the danger level and the physiological monitoring data, generate a corresponding control signal based on the emergency control command, trigger different levels of display warnings, projection prompts and alarm linkage strategies through the vehicle communication bus, and send the control signal or the display warnings, projection prompts and alarm linkage strategies to the multi-display device linkage module, the automatic alarm and rescue module and the safety backup power module. The multi-display device linkage module is used to uniformly control the vehicle's instrument panel, central control screen, vehicle HUD, rear entertainment screen and window display, and form synchronous warning output of multiple terminals in the vehicle under the same hazard level control strategy. When receiving the display warning, the projection prompt and the alarm linkage strategy, it outputs warning information, distress information and accident information in a unified format with different colors and contents. The AR vehicle exterior projection and multi-terminal distress signal module is used to project warning signs onto the ground, outside the vehicle, or on the windows based on the vehicle's exterior projection lights, body LED screens, or transparent window display units. It also coordinates with rearview or side mirror projections to form multi-position collaborative displays, improving the visibility to surrounding traffic participants. The distress signal or warning information is projected onto the rearview or side mirrors. When it is determined that the driver cannot operate the steering wheel or the vehicle's central control system, an automatic alarm is triggered via voice recognition, simultaneously activating window projection, mirror projection, and exterior projection prompts. Based on weather conditions during vehicle operation, including nighttime, rain, fog, or low-visibility environments, corresponding visual and voice warnings are provided, and the display content, color, and flashing frequency are automatically adjusted according to the level of danger.

[0038] It should be noted that "window display" refers to the use of electrochromic, transparent OLED (Organic Light-Emitting Diode), or other transparent display technologies to enable information display capabilities in vehicle windows, allowing for bidirectional information transmission between the interior and exterior. Exterior projection lights are high-brightness projection devices installed on the bottom of the vehicle, front grille, or rear bumper, projecting graphics or text onto the ground or the road surface in front / behind. Transparent window display units are components based on transparent display technology, allowing the window to display digital content while maintaining light transmission, suitable for internal and external information interaction. Multi-location collaborative display refers to the simultaneous or complementary presentation of the same or related warning information at multiple external locations on the vehicle (such as the ground, side mirrors, rearview mirror, and rear), maximizing the perception probability for surrounding road users. Low-visibility environments refer to driving scenarios where visual recognition distance is significantly reduced due to weather conditions such as nighttime, rain, fog, snow, and dust storms. The process of multi-terminal linkage between the interior and exterior of this embodiment can be referred to... Figure 3 The linkage between the window display and the mirror projection can be referenced. Figure 4 The voice-activated distress call process in this embodiment can be referred to as follows. Figure 5 .

[0039] Understandably, the multi-display device linkage module in this embodiment ensures that all occupants receive consistent and clearly tiered warning information at the same time, avoiding misjudgments due to information dispersion or delay. The AR exterior projection and multi-terminal distress signal module extends the warning range to outside the vehicle, using ground projection, mirror displays, and vehicle LEDs in a coordinated manner to enable surrounding vehicles, pedestrians, and other road users to quickly identify the vehicle's emergency status even in situations with limited visibility.

[0040] Furthermore, this embodiment can dynamically optimize display parameters based on weather and visibility conditions, such as increasing the red flashing frequency in rainy or foggy weather and increasing projection brightness at night, thereby overcoming the attenuation of visual signals due to environmental interference. In the event of driver incapacitation, voice-triggered automatic alarms are activated and linked to the entire vehicle projection, enabling autonomous emergency calls without human intervention. Therefore, this embodiment enhances the synergy of in-vehicle human-machine interaction, significantly improving the effectiveness of pre-accident warnings, the timeliness of emergency calls during accidents, and the accuracy of post-accident rescue.

[0041] In one embodiment, the AI-predicted vehicle emergency linkage alarm and backup power system further includes: The automatic alarm and rescue module is used to automatically dial an alarm number or send alarm and distress information to the cloud platform, alarm center, emergency contact, vehicle manufacturer or insurance company when the danger level reaches the preset maximum danger level or the driver physiological monitoring module detects that the driver is unconscious or lost consciousness, via cellular network or vehicle network.

[0042] It should be noted that the automatic alarm and rescue module is a submodule in the system responsible for initiating external communication requests and sending distress signals and vehicle status information to relevant agencies or personnel without manual intervention when specific triggering conditions are met. The preset maximum hazard level is a predefined upper limit threshold for hazard levels by the system. Once the hazard level output by the AI ​​prediction and hazard recognition module reaches this value, it is considered that there is an extremely high risk of an accident or that a serious event has occurred. The automatic alarm process in this embodiment can be referred to... Figure 6 The integration of automatic alarms and multi-terminal linkage can be referenced. Figure 7 .

[0043] Understandably, when the danger level reaches the highest threshold, it indicates that the vehicle faces a highly probable serious accident. Even if the driver is conscious, the impact or confusion may prevent them from seeking help in time; if the driver is unconscious, they are completely incapable of calling for help. In both scenarios, this embodiment can proactively initiate an automatic alarm without human intervention, significantly shortening the time window from accident occurrence to rescue initiation. Furthermore, this embodiment utilizes both cellular and vehicle-to-everything (V2X) networks, ensuring wide-area coverage while also balancing low-latency local collaboration, improving communication reliability and thus significantly enhancing survival rates and rescue efficiency after an accident.

[0044] In summary, this embodiment integrates multimodal sensor data from vehicle cameras, radar, inertial measurement units, body attitude sensors, and control unit status lights with driver physiological monitoring information such as heart rate, respiration, and consciousness status to construct an AI prediction and hazard identification model based on multimodal deep learning. This model can identify potential risks such as collisions, rollovers, or energy system anomalies in advance and dynamically adjust the hazard level based on whether the driver is incapacitated, thereby significantly improving the timeliness and accuracy of risk warnings.

[0045] Furthermore, when the system determines that the driver is incapacitated or the danger level reaches a threshold, this application not only automatically upgrades the in-vehicle alarm level but also activates the safety backup power system. This ensures that critical safety functions such as the communication module, external warning lights, and door lock control remain operational until the main power fails or at the moment of an accident. It also sends emergency signals to nearby vehicles and rescue centers via external broadcasting or a network platform. This effectively solves the problem in existing technologies where alarms are interrupted due to power outages, and external road users are unable to be informed of the danger in a timely manner. It significantly enhances the vehicle's proactive warning capabilities and occupant survival protection level under extreme conditions. Therefore, this application can improve the effectiveness of vehicle safety warnings.

[0046] Based on the first embodiment of this application, in the second embodiment of this application, the content that is the same as or similar to that in the first embodiment described above can be referred to the above description and will not be repeated hereafter. In addition, the system further includes a safety backup power module, comprising an independent battery unit, a backup power management circuit, and a power switching module, which is connected in parallel with the vehicle's main power supply through the backup power management circuit. The safety backup power module is used to switch the vehicle's power supply to backup power within a preset time window of 50ms to 200ms when the main power supply is determined to be damaged or completely disconnected. It provides emergency power to the vehicle's communication module, positioning module, automatic alarm and rescue module, and display and projection module in descending order of priority. The power supply strategy is dynamically adjusted according to the remaining power of the backup power supply and the level of danger to ensure the continuous operation of alarm, positioning and communication functions.

[0047] It should be noted that an independent battery cell is a dedicated energy storage device physically isolated from the vehicle's main power battery or 12V low-voltage system, used to provide emergency power in the event of a main power failure. The backup power management circuit is an electronic control circuit used to monitor the status of the independent battery cell (such as voltage, charge, and health), control the charging and discharging process, and coordinate energy dispatch with the main power source. The power switching module is a high-speed electronic switch or relay component that can seamlessly switch the load from the main power source to the backup power source in a very short time after detecting a main power failure. The backup power switching process in this embodiment can be referred to... Figure 8 .

[0048] The communication module refers to a hardware unit that supports wireless communication capabilities such as cellular networks, V2X, Bluetooth, or Wi-Fi, used to enable remote data interaction between the vehicle and external environments. The positioning module is a positioning receiver used to obtain real-time vehicle location information. The power supply strategy refers to the control logic that dynamically determines whether to supply power to each power module, its power supply priority, and power allocation based on the current system status (such as remaining battery power and hazard level).

[0049] Understandably, the 50ms to 200ms fast switching window in this embodiment is far below the power-off reset threshold of most electronic modules, preventing communication or positioning modules from restarting or losing connection due to power failure, thus ensuring continuous transmission of distress signals and real-time updates of location information. By adopting a tiered priority power supply mechanism, it ensures that under limited backup energy, the most critical communication, positioning, and alarm functions for rescue are maintained first. Even when power is scarce, non-core display or projection functions can be sacrificed to extend the operating time of the core system.

[0050] Furthermore, the ability of this embodiment to dynamically adjust the power supply strategy enables the system to flexibly allocate resources according to the actual level of danger. Under the highest level of danger, it will maintain full-function operation even when the power is low. In medium- and low-risk power failure scenarios, the load can be appropriately reduced to extend standby time, thereby significantly improving the vehicle's autonomous survivability and external communication reliability in extreme accidents.

[0051] In one embodiment, the safety backup power module is further configured to: The system acquires the vehicle's backup power status, main power status, and environmental factors. It then uses the AI ​​prediction and hazard identification module or a risk prediction module that communicates with it to predict the risk coefficient and risk time window. Based on the hazard level change trend or prediction time window output by the AI ​​prediction and hazard identification module, the system determines to enter the pre-charging state when the risk coefficient is higher than a preset risk coefficient threshold and the risk time window is lower than a preset time window threshold, or when the voltage change rate of the backup power supply is detected to be higher than a preset change threshold and the environmental factor is a hazardous environmental factor, or when a manual pre-charging command is received. After entering the pre-charging state, the status of the backup power supply is read to determine whether it needs to be recharged. If it is needed, the charging path is controlled to open or close, and the main power supply or energy recovery path is scheduled to recharge the backup power supply. After the recharge is completed, the backup power supply is switched to the ready-to-activate state. After the backup power supply is in the ready and activated state, its internal resistance, terminal voltage and temperature are measured to confirm that the backup power supply meets the conditions for quick takeover in the event of an accident. If there is no abnormality in the backup power supply, it is checked in conjunction with the vehicle control module to replenish the backup power supply. After the replenishment is completed, it is determined that the backup power supply is ready and the vehicle's power supply is switched to the backup power supply.

[0052] It should be noted that environmental factors refer to external or operating conditions that affect the safety of a vehicle's power system, such as collision acceleration, water immersion signals, high temperatures, battery pack deformation, and road roughness levels. The risk coefficient is a numerical indicator that quantifies the likelihood of a serious accident or power failure occurring within a certain period of time. The risk time window refers to the time interval from the current moment to the predicted occurrence of the accident, used to assess the urgency of the risk.

[0053] The preset risk coefficient threshold is a critical value for risk coefficients pre-set by the system. Exceeding this value indicates a high probability of power supply or vehicle safety risk. The preset time window threshold is the upper limit of the risk time window set by the system. If an accident is predicted to occur within this time, it is considered an emergency. The voltage change rate refers to the magnitude of voltage change at the backup power supply terminal per unit time (e.g., dV / dt). Abnormal increases or decreases may indicate internal short circuits, aging, or sudden load changes. Hazardous environmental factors refer to environmental conditions marked as high-risk by the system, such as detecting that the vehicle is in a skidding, rollover, wading, or high-temperature cabin state.

[0054] Manual pre-charging commands are active commands issued by the driver via the central control interface or voice, requesting the system to prepare for replenishing and activating the backup power supply in advance. The pre-charging state is a preparation phase for the backup power supply; in this state, the system assesses whether replenishment is needed and initiates the charging process. The charging path is a controllable power transmission path connecting the main power supply or energy recovery system (such as regenerative braking) to the backup power supply. The energy recovery path refers to the channel through which kinetic energy is converted into electrical energy and returned to the power system during vehicle deceleration or braking, and can be used to replenish the backup power supply. The ready-to-activate state is the standby state of the backup power supply after completing self-checks and replenishment, indicating that it has met all the electrical and thermodynamic conditions for rapid switchover. The vehicle control module (VCU) is the vehicle's central control system, responsible for coordinating the work of various subsystems, including power management, drive control, and brake distribution.

[0055] Understandably, traditional backup power supplies typically activate passively only after the main power supply fails. However, if the backup battery is low on charge or has hidden faults (such as increased internal resistance or low-temperature performance degradation), it cannot effectively support emergency functions. This embodiment, by predicting high-risk events in advance and monitoring abnormal dynamics of the power supply itself, proactively enters the pre-charging and activation preparation process before the actual accident occurs, thereby ensuring that the backup power supply is available at critical moments.

[0056] Furthermore, this embodiment empowers the driver with the ability to proactively upgrade the system's safety level when a potential hazard is anticipated through a manual pre-charging command, enhancing the flexibility of human-machine collaboration. Moreover, by verifying key parameters such as internal resistance, voltage, and temperature before readiness, and coordinating with the vehicle control module for confirmation, the risk of secondary failure due to forcibly taking over a faulty battery is avoided, improving the startup success rate and power supply stability of the backup power supply under extreme conditions.

[0057] In summary, this embodiment can rapidly switch to the backup power supply within 50ms to 200ms when the main power supply fails or is interrupted, prioritizing emergency power supply for critical functions such as communication, positioning, alarms, and display projection. Furthermore, by integrating vehicle power status, environmental factors, and AI-predicted risk coefficients and time windows, the system can automatically enter a pre-charging state before a high-risk event occurs, assess the backup battery's charge level, schedule the main power supply or energy recovery path for recharging, and perform a self-check by measuring internal resistance, terminal voltage, and temperature. Once no abnormalities are confirmed, the system, in conjunction with the vehicle control module, puts the backup power supply into a ready-to-activate state, ensuring it meets the conditions for rapid and reliable takeover in an accident.

[0058] This embodiment ensures the continued operation of the distress channel after the main power supply fails through millisecond-level power switching and a tiered power supply strategy. Furthermore, the proactive pre-charging mechanism based on multi-source risk perception avoids the failure of emergency functions due to insufficient power or hidden faults. In high-risk scenarios such as collisions, rollovers, or water wading, even if the main power supply is instantly destroyed, the system can still maintain core safety functions by relying on a pre-activated and intact backup power supply, thereby ensuring uninterrupted positioning, communication, and automatic alarms, and improving the effectiveness of vehicle safety warnings.

[0059] Based on the first and second embodiments of this application, the executing entity of the AI-predictive vehicle emergency linkage alarm and safety backup power method of this application can be an AI-predictive vehicle emergency linkage alarm and safety backup power device. The AI-predictive vehicle emergency linkage alarm and safety backup power device belongs to the above-mentioned AI-predictive vehicle emergency linkage alarm and safety backup power system. The specific execution steps of the AI-predictive vehicle emergency linkage alarm and safety backup power method are basically the same as the implementation methods of the above-mentioned hardware modules, and can be referred to the above description, which will not be repeated hereafter.

[0060] It should be noted that the above examples are only for understanding this application and do not constitute a limitation on the vehicle emergency linkage alarm and safety backup power method based on AI prediction in this application. Any simple modifications based on this technical concept are within the protection scope of this application.

[0061] All user-related data involved in this application was obtained with the user's permission or consent, as per [reference]. Figure 9 In other words, when this application is applied to a specific product or technology, user permission is required to acquire and process the relevant data, and the processing of the relevant data must comply with the relevant laws, regulations and regulatory standards of the relevant countries and regions.

[0062] The above description is only a part of the embodiments of this application and does not limit the scope of protection of this application. All equivalent structural transformations made under the technical concept of this application and using the content of this application specification and drawings, or direct / indirect applications in other related technical fields, are included in the patent protection scope of this application.

Claims

1. A vehicle emergency linkage alarm and backup power system based on AI prediction, characterized in that, The AI-based vehicle emergency alarm and backup power system includes the following interconnected hardware modules: The vehicle sensor module includes a camera, radar, inertial measurement unit, vehicle attitude sensor and vehicle control unit status interface, used to collect multimodal data in real time, wherein the multimodal data includes vehicle operating status and environmental information; The driver physiological monitoring module is used to monitor the driver's physiological data through steering wheel sensors, seat sensors and in-vehicle cameras in the vehicle; The AI ​​prediction and hazard identification module is used to predict potential hazards based on the multimodal data using a multimodal fusion deep learning model. The potential hazards include at least the risk of an impending collision, rollover, or abnormal vehicle energy system. The module also combines the physiological monitoring data to classify and dynamically adjust the hazard level and output the hazard level. The driver physiological monitoring module is also used to send a driver incapacity signal to the emergency control unit and automatically upgrade the vehicle's alarm level when the driver is detected to be unconscious, lost consciousness, or unable to perform effective vehicle operation based on the physiological monitoring data. The physiological monitoring data includes the driver's heart rate, respiration, and consciousness status, and the alarm level is determined based on the danger level.

2. The AI-predicted vehicle emergency linkage alarm and backup power system as described in claim 1, characterized in that, The system also includes a safety backup power module, which includes an independent battery unit, a backup power management circuit and a power switching module, and is connected in parallel with the vehicle's main power supply through the backup power management circuit; The safety backup power module is used to switch the vehicle's power supply to backup power within a preset time window of 50ms to 200ms when the main power supply is determined to be damaged or completely disconnected. It provides emergency power to the vehicle's communication module, positioning module, automatic alarm and rescue module, and display and projection module in descending order of priority. The power supply strategy is dynamically adjusted according to the remaining power of the backup power supply and the level of danger to ensure the continuous operation of alarm, positioning and communication functions.

3. The AI-predicted vehicle emergency linkage alarm and backup power system as described in claim 2, characterized in that, The safety backup power module is also used for: The system acquires the vehicle's backup power status, main power status, and environmental factors. It then uses the AI ​​prediction and hazard identification module or a risk prediction module that communicates with it to predict the risk coefficient and risk time window. Combining the hazard level change trend or prediction time window output by the AI ​​prediction and hazard identification module, the system determines to enter the pre-charging state when the risk coefficient is higher than a preset risk coefficient threshold and the risk time window is lower than a preset time window threshold, or when the voltage change rate of the backup power supply is detected to be higher than a preset change threshold and the environmental factor is a hazardous environmental factor, or when a manual pre-charging command is received. After entering the pre-charging state, the status of the backup power supply is read to determine whether it needs to be recharged. If it is needed, the charging path is controlled to open or close, and the main power supply or energy recovery path is scheduled to recharge the backup power supply. After the recharge is completed, the backup power supply is switched to the ready-to-activate state. After the backup power supply is in the ready and activated state, its internal resistance, terminal voltage and temperature are measured to confirm that the backup power supply meets the conditions for quick takeover in the event of an accident. If there is no abnormality in the backup power supply, it is checked in conjunction with the vehicle control module to replenish the backup power supply. After the replenishment is completed, it is determined that the backup power supply is ready and the vehicle's power supply is switched to the backup power supply.

4. The AI-predicted vehicle emergency linkage alarm and backup power system as described in claim 1, characterized in that, The system also includes a multi-display device linkage module and an AR vehicle exterior projection and multi-terminal distress prompt module. The AI ​​prediction and hazard identification module also includes an emergency control unit. The emergency control unit is used to receive the physiological monitoring data, determine the danger level, generate an emergency control command based on the danger level and the physiological monitoring data, generate a corresponding control signal based on the emergency control command, trigger different levels of display warnings, projection prompts and alarm linkage strategies through the vehicle communication bus, and send the control signal or the display warnings, projection prompts and alarm linkage strategies to the multi-display device linkage module, the automatic alarm and rescue module and the safety backup power module. The multi-display device linkage module is used to uniformly control the vehicle's instrument panel, central control screen, vehicle HUD, rear entertainment screen and window display, and form synchronous warning output of multiple terminals in the vehicle under the same hazard level control strategy. When receiving the display warning, the projection prompt and the alarm linkage strategy, it outputs warning information, distress information and accident information in a unified format with different colors and contents. The AR vehicle exterior projection and multi-terminal distress signal module is used to project warning signs onto the ground, outside the vehicle, or on the windows based on the vehicle's exterior projection lights, body LED screens, or transparent window display units. It also coordinates with rearview or side mirror projections to form multi-position collaborative displays, improving the visibility to surrounding traffic participants. The distress signal or warning information is projected onto the rearview or side mirrors. When it is determined that the driver cannot operate the steering wheel or the vehicle's central control system, an automatic alarm is triggered via voice recognition, simultaneously activating window projection, mirror projection, and exterior projection prompts. Based on weather conditions during vehicle operation, including nighttime, rain, fog, or low-visibility environments, corresponding visual and voice warnings are provided, and the display content, color, and flashing frequency are automatically adjusted according to the level of danger.

5. The AI-predicted vehicle emergency linkage alarm and backup power system as described in claim 1, characterized in that, The system also includes: The automatic alarm and rescue module is used to automatically dial an alarm number or send alarm and distress information to the cloud platform, alarm center, emergency contact, vehicle manufacturer or insurance company when the danger level reaches the preset maximum danger level or the driver physiological monitoring module detects that the driver is unconscious or lost consciousness, via cellular network or vehicle network.

6. A vehicle emergency linkage alarm and backup power method based on AI prediction, applied to an AI prediction vehicle emergency linkage alarm and backup power system, characterized in that, The method includes: Real-time acquisition of multimodal data, including vehicle operating status and environmental information; The vehicle monitors the driver's physiological data through steering wheel sensors, seat sensors, and in-vehicle cameras. Based on the multimodal data, a preset deep learning model is used to predict potential hazards. When the predicted hazard level is higher than a preset threshold, the safety backup power system is controlled to enter a pre-charging or ready-to-activate state. The hazard level is then classified and dynamically adjusted in conjunction with the physiological monitoring data, and the hazard level is output. When the driver is detected to be unconscious or unable to operate the vehicle based on the physiological monitoring data, the vehicle's alarm level is automatically upgraded. The physiological monitoring data includes the driver's heart rate, respiration, and consciousness status, and the alarm level is determined based on the danger level.

7. The method as described in claim 6, characterized in that, The AI-based vehicle emergency linkage alarm and backup power method also includes: When it is determined that the vehicle's main power supply is damaged or completely cut off, the vehicle's power supply will be switched to the backup power supply within a preset time. Emergency power will be provided to the vehicle's communication module, positioning module, automatic alarm and rescue module and display and projection module in order of priority from high to low to ensure the continuous operation of alarm, positioning and communication functions. Non-core function modules will be shut down step by step according to the remaining power of the backup power supply.

8. The method as described in claim 7, characterized in that, The step of switching the vehicle's power supply to a backup power supply within a preset time when it is determined that the main power supply is damaged or completely lost includes: The system acquires the vehicle's backup power status, main power status, and environmental factors. It predicts risk coefficients and risk time windows through a preset risk prediction module, and combines the risk level change trend or prediction time window output by the AI ​​prediction and hazard identification module. When the risk coefficient is higher than a preset risk coefficient threshold and the risk time window is lower than a preset time window threshold, or when the voltage change rate of the backup power is detected to be higher than a preset change threshold and the environmental factor is a hazardous environmental factor, or when a manual pre-charging command is received, the system determines to enter the pre-charging state. After entering the pre-charging state, the status of the backup power supply is read to determine whether it needs to be recharged. If it is needed, the charging path is controlled to open or close, and the main power supply or energy recovery path is scheduled to recharge the backup power supply. After the recharge is completed, the backup power supply is switched to the ready-to-activate state. After the backup power supply is in the ready and activated state, its internal resistance, terminal voltage and temperature are measured to confirm that the backup power supply meets the conditions for quick takeover in the event of an accident. It is also determined whether the backup power supply is abnormal. If there is no abnormality, the backup power supply is recharged in conjunction with the vehicle control module. After the recharge is completed, the backup power supply is confirmed to be ready and the vehicle's power supply is switched to the backup power supply.

9. The method as described in claim 6, characterized in that, Following the step of monitoring the driver's physiological data using steering wheel sensors, seat sensors, and in-vehicle cameras, the method further includes: The system receives the physiological monitoring data, determines the danger level, generates an emergency control command based on the danger level and the physiological monitoring data, generates a corresponding control signal based on the emergency control command, triggers different levels of display warnings, projection prompts and alarm linkage strategies through the vehicle communication bus, and sends the control signal or the display warnings, projection prompts and alarm linkage strategies to the multi-display device linkage module, the automatic alarm and rescue module and the safety backup power module. The system provides unified control over the vehicle's dashboard, central control screen, in-vehicle HUD, rear entertainment screen, and window displays. Upon receiving the display warnings, projection prompts, and alarm linkage strategies, it outputs warning messages, distress messages, and accident information in a unified format with different colors and contents. Based on the vehicle's external projection lights, body LED screen, or transparent window display unit, warning signs are projected onto the ground, outside the vehicle, or on the windows. The distress message or warning message is projected onto the vehicle's rearview mirror or side mirror. When it is determined that the driver cannot operate the steering wheel or the vehicle's central control, an automatic alarm is triggered through voice recognition, and the window and mirror projections and external projection prompts are activated simultaneously. Based on the weather conditions during vehicle operation, corresponding visual and voice warnings are given, and the display content, color, and flashing frequency are automatically adjusted according to the level of danger.

10. The method as described in claim 6, characterized in that, Following the steps of predicting potential hazards using a preset deep learning model based on the multimodal data, controlling the safety backup power system to enter a pre-charging or ready-to-activate state when the predicted hazard level is higher than a preset threshold, and combining the physiological monitoring data to classify and dynamically adjust the hazard level, and then outputting the hazard level, the system further includes: When the danger level reaches the preset maximum danger level, or when the driver is detected to be unconscious, passengers are allowed to trigger an automatic alarm process via voice recognition commands, and simultaneously activate in-vehicle and out-of-vehicle warnings and projection prompts. The system can automatically dial an alarm number or send alarm and distress information to the cloud platform, alarm center, emergency contact, vehicle manufacturer, or insurance company via cellular network or vehicle network.

Images