Emergency rescue method and device based on driver health monitoring, vehicle and medium

Abstract

The invention relates to an emergency rescue method and device based on driver health monitoring, a vehicle and a medium. The method comprises the following steps: acquiring real-time health state data and a current driving scene type of a driver in a vehicle driving process, and performing weight distribution on each monitoring dimension of the real-time health state data based on the current driving scene type to obtain a monitoring data group corresponding to the current driving scene type; determining a current alarm level under the condition that the driver is judged to be in a preset abnormal state based on the monitoring data group, and selecting a corresponding auxiliary driving strategy and a corresponding emergency rescue strategy based on the current alarm level and the current driving scene type; the current vehicle is controlled to run based on the corresponding auxiliary driving strategy, and / or emergency rescue is carried out on the driver based on the corresponding emergency rescue strategy. Therefore, the problem that traffic accidents occur due to the fact that the abnormal state of the driver is not monitored in time in the prior art is solved, and driving safety is improved.

Description

Technical Field

[0001] This application relates to the field of intelligent transportation and autonomous driving technologies, and particularly relates to an emergency rescue method, device, vehicle, and medium based on driver health monitoring. Background Art

[0002] With the rapid development of autonomous driving technologies, the vehicle industry is entering a brand-new intelligent era. The popularization of intelligent technologies not only greatly improves the convenience of driving but also is expected to fundamentally solve traffic accident problems caused by human operation errors. However, while autonomous driving technologies are developing rapidly, it is also necessary to face up to their deficiencies in dealing with sudden health conditions of drivers. Emergency conditions such as heart attacks and fainting are often sudden and unpredictable. Once they occur, the driver will instantly lose consciousness or be unable to effectively control the vehicle, thus increasing the risk of traffic accidents.

[0003] In related technologies, the health status of the driver is monitored in real time through wearable devices worn on the driver.

[0004] However, this method is highly dependent on wearable devices. The driver may not be monitored in a timely manner for abnormal conditions due to reasons such as forgetting to wear the device or device failure, which urgently needs to be solved. Summary of the Invention

[0005] This application provides an emergency rescue method, device, vehicle, and medium based on driver health monitoring to solve the problem of traffic accidents caused by untimely monitoring of abnormal driver states in the prior art and improve driving safety.

[0006] To achieve the above object, the first aspect embodiment of this application proposes an emergency rescue method based on driver health monitoring, including the following steps:

[0007] Obtain the real-time health status data of the driver during the driving process and the current driving scene type, and allocate weights to each monitoring dimension of the real-time health status data based on the current driving scene type to obtain a monitoring data group corresponding to the current driving scene type;

[0008] Judge whether the driver is in a preset abnormal state based on the monitoring data group;

[0009] When the driver is in the preset abnormal state, determine the current warning level, and select a corresponding assisted driving strategy and a corresponding emergency rescue strategy based on the current warning level and the current driving scene type;

[0010] Controlling the current vehicle to travel based on the corresponding assisted driving strategy and / or implementing emergency rescue for the driver based on the corresponding emergency rescue strategy.

[0011] According to an embodiment of the present application, the weight assignment for each monitoring dimension of the real-time health status data based on the current driving scenario type includes:

[0012] Setting different initial priority weight coefficients for each monitoring dimension of the health status data based on the driving scenario type and preset trigger conditions, where the preset trigger conditions correspond to the driving scenario type;

[0013] Judging whether the real-time health status data triggers the preset conflict resolution rule based on the current driving scenario type;

[0014] If the real-time health status data triggers the preset conflict resolution rule, adjusting the initial priority weight coefficients of each monitoring dimension based on the current driving scenario type and the real-time health status data to obtain new priority weight coefficients; otherwise, keeping the initial priority weight coefficients of each monitoring dimension unchanged.

[0015] According to an embodiment of the present application, the judging whether the driver is in a preset abnormal state based on the monitoring data group includes:

[0016] When the real-time health status data triggers the preset conflict resolution rule, calculating the current comprehensive confidence level based on the new priority weight coefficients of each monitoring dimension of the monitoring data group and the preset conflict resolution rule, and judging whether the driver is in the preset abnormal state based on the current comprehensive confidence level;

[0017] Or, when the real-time health status data does not trigger the preset conflict resolution rule, calculating the current comprehensive confidence level based on the initial priority weight coefficients of each monitoring dimension of the monitoring data group, and judging whether the driver is in the preset abnormal state based on the current comprehensive confidence level.

[0018] According to an embodiment of the present application, when the driver is in the preset abnormal state, determining the current alarm level includes:

[0019] Determining a grading response threshold interval according to the current comprehensive confidence level;

[0020] Determining the current alarm level based on the grading response threshold interval.

[0021] According to an embodiment of the present application, before obtaining the health status data of the current driver, it further includes:

[0022] Determine whether the current vehicle is in the driving stage and whether the face of the current driver is in the preset recognition area;

[0023] When the current vehicle is in the driving stage and the face of the current driver is in the preset recognition area, obtain the health status data of the current driver.

[0024] According to the emergency rescue method based on driver health monitoring proposed in the embodiments of the present application, by obtaining the real-time health status data of the driver during the driving process and the current driving scene type, weight allocation can be performed on each monitoring dimension of the real-time health status data based on the current driving scene type, so as to obtain a monitoring data group corresponding to the current driving scene type; then, when it is determined that the driver is in a preset abnormal state based on the monitoring data group, determine the current warning level, and select the corresponding assisted driving strategy and the corresponding emergency rescue strategy based on the current warning level and the current driving scene type; control the current vehicle to drive based on the corresponding assisted driving strategy, and / or perform emergency rescue on the driver based on the corresponding emergency rescue strategy. Thus, the problem of traffic accidents caused by the untimely monitoring of the abnormal state of the driver in the prior art is solved, and driving safety is improved.

[0025] To achieve the above object, an emergency rescue device based on driver health monitoring proposed in the second aspect of the present application includes:

[0026] An obtaining module, configured to obtain the real-time health status data of the driver during the driving process and the current driving scene type, and perform weight allocation on each monitoring dimension of the real-time health status data based on the current driving scene type to obtain a monitoring data group corresponding to the current driving scene type;

[0027] A judging module, configured to judge whether the driver is in a preset abnormal state based on the monitoring data group;

[0028] A selecting module, configured to determine the current warning level when the driver is in the preset abnormal state, and select the corresponding assisted driving strategy and the corresponding emergency rescue strategy based on the current warning level and the current driving scene type;

[0029] A processing module, configured to control the current vehicle to drive based on the corresponding assisted driving strategy, and / or perform emergency rescue on the driver based on the corresponding emergency rescue strategy.

[0030] According to an embodiment of the present application, the obtaining module is specifically configured to:

[0031] Set different initial priority weight coefficients for each monitoring dimension of the health status data based on the driving scenario type and preset trigger conditions, where the preset trigger conditions correspond to the driving scenario type;

[0032] Based on the current driving scenario type, determine whether the real-time health status data triggers the preset conflict resolution rule;

[0033] If the real-time health status data triggers the preset conflict resolution rule, adjust the initial priority weight coefficients of each monitoring dimension based on the current driving scenario type and the real-time health status data to obtain new priority weight coefficients; otherwise, keep the initial priority weight coefficients of each monitoring dimension unchanged.

[0034] According to an embodiment of the present application, the determining module is specifically configured to:

[0035] When the real-time health status data triggers the preset conflict resolution rule, calculate the current comprehensive confidence level based on the new priority weight coefficients of each monitoring dimension of the monitoring data group and the preset conflict resolution rule, and determine whether the driver is in the preset abnormal state based on the current comprehensive confidence level;

[0036] Or, when the real-time health status data does not trigger the preset conflict resolution rule, calculate the current comprehensive confidence level based on the initial priority weight coefficients of each monitoring dimension of the monitoring data group, and determine whether the driver is in the preset abnormal state based on the current comprehensive confidence level.

[0037] According to an embodiment of the present application, the selecting module is specifically configured to:

[0038] Determine the hierarchical response threshold interval according to the current comprehensive confidence level;

[0039] Determine the current alarm level based on the hierarchical response threshold interval.

[0040] According to an embodiment of the present application, before obtaining the health status data of the current driver, the obtaining module is further configured to:

[0041] Determine whether the current vehicle is in the driving stage and whether the face of the current driver is in the preset recognition area;

[0042] When the current vehicle is in the driving stage and the face of the current driver is in the preset recognition area, obtain the health status data of the current driver.

[0043] An emergency rescue device based on driver health monitoring proposed according to an embodiment of the present application can, by acquiring real-time health status data of a driver during the process of driving a vehicle and the current driving scenario type, assign weights to each monitoring dimension of the real-time health status data based on the current driving scenario type, so as to obtain a monitoring data set corresponding to the current driving scenario type; then, when it is determined that the driver is in a preset abnormal state based on the monitoring data set, determine the current alarm level, and select a corresponding assisted driving strategy and a corresponding emergency rescue strategy based on the current alarm level and the current driving scenario type; control the current vehicle to drive based on the corresponding assisted driving strategy, and / or perform emergency rescue on the driver based on the corresponding emergency rescue strategy. Thereby, the problem in the prior art that traffic accidents occur due to untimely monitoring of the abnormal state of the driver is solved, and driving safety is improved.

[0044] To achieve the above object, an embodiment of the third aspect of the present application proposes a vehicle, including: a memory, a processor, and a computer program stored on the memory and executable on the processor, and the processor executes the program to implement the emergency rescue method based on driver health monitoring as described in the above embodiment.

[0045] To achieve the above object, an embodiment of the fourth aspect of the present application proposes a computer-readable storage medium, on which a computer program is stored, and the program is executed by a processor to be used to implement the emergency rescue method based on driver health monitoring as described in the above embodiment.

[0046] Additional aspects and advantages of the present application will be given in part in the following description, become apparent in part from the following description, or be learned through the practice of the present application. Description of the Drawings

[0047] The above and / or additional aspects and advantages of the present application will become apparent and be readily understood from the following description of the embodiments in conjunction with the drawings, where:

[0048] Figure 1 It is a flowchart of an emergency rescue method based on driver health monitoring provided according to an embodiment of the present application;

[0049] Figure 2 It is a block diagram of an emergency rescue device based on driver health monitoring provided according to an embodiment of the present application;

[0050] Figure 3 It is a structural diagram of a vehicle provided according to an embodiment of the present application. Detailed Embodiments

[0051] Embodiments of the present application will be described in detail below. Examples of the embodiments are shown in the accompanying drawings, where like or similar reference numerals denote like or similar elements or elements having like or similar functions throughout. The embodiments described below by referring to the accompanying drawings are exemplary and are intended to explain the present application, and should not be construed as limiting the present application.

[0052] The emergency rescue method, device, vehicle and medium based on driver health monitoring proposed according to the embodiments of the present application will be described below with reference to the accompanying drawings.

[0053] Figure 1 It is a flowchart of an emergency rescue method based on driver health monitoring according to an embodiment of the present application.

[0054] Before introducing the emergency rescue method based on driver health monitoring proposed in the embodiments of the present application, the autonomous driving emergency rescue system involved in this method will be introduced first. This system integrates advanced biometric technology and Internet of Things technology. Through the built-in Driver Monitoring System (DMS for short) and Steering Wheel Physiological Monitoring Module (SWPM for short), it realizes comprehensive and real-time monitoring of the driver's physiological indicators (such as heart rate, blood oxygen saturation, body temperature, etc.) and behavioral states (such as facial expressions, head postures, gesture movements, etc.). Among them, the DMS module includes facial recognition and attention monitoring, identity recognition, FIR (Far Infrared Radiation) camera body temperature monitoring, behavior monitoring, etc. Facial recognition and attention monitoring uses a high-precision camera and deep learning algorithm to real-time monitor the driver's facial expressions, blink frequency and head posture to determine whether the driver is in a distracted or fatigued state; identity recognition combines facial recognition technology to quickly verify the driver's identity to ensure the safety and personalized settings of vehicle use; FIR camera body temperature monitoring integrates an FIR camera, which can non-contact measure the driver's body temperature and real-time monitor abnormal body temperature to prevent disease transmission and heat stroke risk; behavior monitoring analyzes the driver's gestures and limb movements to identify abnormal driving behaviors such as suddenly deviating from the lane and sudden braking. The SWPM module includes a photoplethysmography (PPG for short), a pressure sensor array, a temperature sensor, and an electrocardiogram (ECG for short). The PPG can measure the change in the volume of the driver's hand blood vessels with the heartbeat through an LED (Light Emitting Diode) light source and a photosensitive element, and accurately calculate the heart rate and blood oxygen saturation; the pressure sensor array is distributed on the surface of the steering wheel, which can monitor the pressure distribution when the driver holds the steering wheel and evaluate their tension and fatigue state; the temperature sensor is used to monitor the driver's hand skin temperature and analyze the driver's body temperature regulation state in combination with the ambient temperature to warn of abnormal conditions such as fever or shivering; the ECG can directly collect the heart electrical signal through a contact electrode to provide detailed heart health information such as heart rate and heart rhythm.

[0055] The following will elaborate in detail on the emergency rescue method based on driver health monitoring proposed in the embodiments of the present application in combination with this autonomous driving emergency rescue system.

[0056] Exemplarily, as Figure 1 shown, the emergency rescue method based on driver health monitoring includes the following steps:

[0057] In step S101, obtain the real-time health status data of the driver during the vehicle driving process and the current driving scenario type, and allocate weights to each monitoring dimension of the real-time health status data based on the current driving scenario type to obtain a monitoring data set corresponding to the current driving scenario type.

[0058] Specifically, when the vehicle starts, the autonomous driving emergency rescue system also starts and enters the monitoring state. At the same time, the system can be initialized according to preset parameters and configurations, including the setting of emergency contacts, the contact information of the rescue center, and vehicle scenario information, etc. After the system starts and the parameters are set, the DMS module and the SWPM module can obtain the real-time health status data of the driver during the vehicle driving process. The DMS module mainly covers facial expression confidence (0 - 1, 1 indicates full concentration, 0 indicates full distraction. For example, a continuous facial expression confidence < 0.3 may trigger a fatigue warning), blink frequency (the number of blinks per unit time, normally about 0.2 - 0.3 Hz (12 - 18 times per minute). If the blink frequency > 0.5 Hz, it may be nervousness, and if the blink frequency < 0.1 Hz, it may be drowsiness), head deflection angle (the angle by which the head deviates from the due front measured by a 3D camera. If the head deflection angle continuously > 30° for more than 5 seconds, it is determined as distracted driving), body temperature (measured non-contact by a far-infrared camera. If it is detected that > 37.5 °C lasts for 10 minutes, a high-temperature warning is activated), abnormal gesture coding (such as G1 - G5, defining specific gestures corresponding to abnormal states. For example, G3 indicates that the driver covers the chest with one hand (suspected heart discomfort), etc.); the SWPM module mainly covers PPG heart rate (the heart rate measured by photoplethysmography, the normal range is 60 - 100 bpm. For example, a sudden > 140 bpm may indicate a panic attack), blood oxygen saturation (reflecting the oxygen-carrying capacity of the blood, normally > 95%. If the blood oxygen saturation < 90% lasts for 1 minute, an oxygen deficiency alarm is triggered), grip force distribution entropy value (0 - 100, quantifying the force fluctuation of holding the steering wheel. During normal driving, it is about 20 - 40. For example, an entropy value > 70 indicates muscle tension (such as the grip during sudden pain)), hand temperature (the normal hand skin temperature is usually in the range of 28 - 34 °C, but it is greatly affected by environmental temperature, blood circulation and other factors, and can be comprehensively judged in combination with other factors such as heart rate), ECG R - R interval (the time difference between adjacent R waves in an electrocardiogram, the normal fluctuation is 50 - 200 ms. If the difference between three consecutive R - R intervals > 300 ms, it indicates a risk of arrhythmia). After analysis, the real-time health status data is divided into three monitoring dimensions: physiological indicators, behavior monitoring, and facial conditions.

[0059] In addition, by obtaining real-time vehicle speed, GPS (Global Positioning System) positioning data, and ambient light sensor data (i.e., light intensity), the driving scenario can be divided in real time to obtain the current driving scenario type. It should be noted that in the embodiments of the present application, the driving scenario types include, but are not limited to, the following: highway cruise, congested section, night driving, and emergency takeover. Based on the sliding window technology, the determination result of the current driving scenario type can be updated once per second.

[0060] To further improve the accuracy and practicality of monitoring, the embodiments of the present application also propose a multi-modal data fusion and dynamic weight allocation mechanism, which can dynamically adjust the weight of each monitoring dimension of the real-time health status data according to the current driving scenario type to further obtain the most accurate monitoring result (i.e., the monitoring data group corresponding to the current driving scenario type).

[0061] For ease of understanding, the following details how to allocate weights to each monitoring dimension of the real-time health status data based on the current driving scenario type.

[0062] As a possible implementation manner, in some embodiments, allocating weights to each monitoring dimension of the real-time health status data based on the current driving scenario type includes: setting different initial priority weight coefficients for each monitoring dimension of the health status data based on the driving scenario type and a preset trigger condition, where the preset trigger condition corresponds to the driving scenario type; determining whether the real-time health status data triggers a preset conflict resolution rule based on the current driving scenario type; if the real-time health status data triggers the preset conflict resolution rule, then adjusting the initial priority weight coefficient of each monitoring dimension based on the current driving scenario type and the real-time health status data to obtain a new priority weight coefficient, otherwise maintaining the initial priority weight coefficient of each monitoring dimension unchanged.

[0063] Specifically, after obtaining the real-time health status data and the current driving scenario type, the real-time health status data can be preprocessed first, including time alignment, normalization, and outlier filtering. In the embodiments of the present application, a sliding window technology with a 5-second window and a 1-second step is used to align multi-source data. For example, if the DMS module collects facial data every second and the SWPM module collects heart rate data every 0.5 seconds, the system will package all the data within every 5 seconds into an analysis unit, and the window slides 1 second each time (i.e., there is a 4-second overlap between adjacent windows), so as to ensure synchronous analysis of physiological data and behavioral data in the time dimension. Normalization (i.e., Z-score normalization) can be performed on data with different dimensions (such as heart rate in bpm and temperature in °C). For example, assuming that the average heart rate of a certain driver in history is 75 bpm and the standard deviation is 10, when the current measured value is 85 bpm, Z-score = (85 - 75) / 10 = 1; if the average hand temperature is 36 °C and the standard deviation is 0.5, when the measured value is 36.5 °C, Z-score = (36.5 - 36) / 0.5 = 1. Thus, it is convenient for subsequent algorithms to uniformly process. A dynamic formula based on age is used to set the filtering threshold. For example, the maximum heart rate threshold for a 30-year-old driver = 220 - 30×0.8 = 196 bpm. If the PPG sensor suddenly detects a heart rate of 200 bpm, it is determined as an outlier and discarded. The system can continuously update the historical data (such as the heart rate distribution in the recent 30 days) and dynamically adjust the standard deviation multiple threshold (such as triggering filtering when exceeding the mean ±3σ). Through the combined effect of these three operations, the problems of time synchronization, dimension unification, and noise interference of multi-source heterogeneous data can be solved, providing high-quality data input for subsequent fusion decision-making.

[0064] It should be noted that in the embodiments of the present application, different initial priority weight coefficients can be assigned to each monitoring dimension of the health status data in advance according to different driving scenario types and the corresponding preset trigger conditions. As shown in Table 1, these weight coefficients indicate which health indicators are more important and need to be given priority attention under specific circumstances.

[0065] Table 1

[0066] Driving scenario type Preset trigger condition Priority weight assignment (α1:α2:α3) High-speed cruise Vehicle speed > 90 km / h and lane lines are stable Physiological indicators: Behavior monitoring: Facial = 5:3:2 Congested section Vehicle speed < 20 km / h and frequent start / stop Behavior monitoring: Physiological indicators: Facial = 4:4:2 Night driving Illumination < 10 lux and time > 20:00 Facial: Physiological indicators: Behavior monitoring = 6:2:2 Emergency takeover The system detects a collision risk > L3 level Behavior monitoring: Physiological indicators: Facial = 5:4:1

[0067] Furthermore, based on the preprocessed real-time health status data, multi-source data can be fused using fuzzy logic and the Dempster-Shafer (D-S) theory. Fuzzy logic is a logical system for dealing with uncertainty and fuzzy information. It allows the membership degree of things to take values between completely belonging and completely not belonging, rather than a simple binary opposition (yes or no). The D-S theory is a mathematical method for dealing with uncertain information. It represents the degree of support of evidence for a certain proposition by establishing a belief function. In the embodiments of this application, the evidence bodies are defined as follows: E1 (physiological abnormality): PPG heart rate variability > 50 ms + hand temperature gradient > 0.5 °C / s; E2 (behavioral abnormality): grip strength entropy value > 70 + 3 lane departures within 10 seconds; E3 (facial abnormality): blink frequency < 8 times / min + pupil diameter < 3 mm for 20 s.

[0068] Next, the system can determine whether the preprocessed real-time health status data triggers a preset conflict resolution rule based on the current driving scenario type, where the preset conflict resolution rule is a mechanism that can be dynamically adjusted according to different driving scenario types. The preset conflict resolution rule is as follows: When the vehicle is in a high-speed cruise scenario, if both E1 and E3 are detected to be triggered simultaneously, the confidence level is increased by 30%; when the vehicle is in a night driving scenario, if E2 is detected to be triggered alone, the body temperature data can be combined for verification. For example, when the hand temperature < 35 °C, the weight of the E2 evidence is reduced by 50% to avoid false triggering of an emergency response.

[0069] By analyzing the preprocessed real-time health status data (checking whether E1 and / or E2 and / or E3 are triggered), when it is determined that the preprocessed real-time health status data triggers the preset conflict resolution rule, the initial priority weight coefficient of each monitoring dimension can be adjusted in combination with the current driving scenario type and the real-time health status data to obtain a new priority weight coefficient, where the calculation formula for the priority weight coefficient of each monitoring dimension is as follows:

[0070] W i = α i ×(1 + β × Sensor_Confidence) × (γ × Historical_Risk);

[0071] where, W i is the new priority weight coefficient of each monitoring dimension, α iα is the initial priority weight coefficient for each monitoring dimension, β is the sensor confidence compensation factor (e.g., when the PPG signal quality > 80%, β = 0.2), Sensor_Confidence is the sensor confidence, γ is the amplification factor of the driver's historical risk level (e.g., if there are 3 instances of fatigue driving within 3 months, then γ = 1.5), and Historical_Risk is the driver's historical risk level.

[0072] If the preprocessed real-time health status data does not trigger the preset conflict resolution rule, then keep the initial priority weight coefficient of each monitoring dimension of the preprocessed real-time health status data unchanged.

[0073] In step S102, based on the monitoring data set, determine whether the driver is in a preset abnormal state.

[0074] Specifically, after obtaining the monitoring data set corresponding to the current driving scenario type, the system can determine whether the driver is in a preset abnormal state or whether the driver has potential health risks based on this monitoring data set.

[0075] As a possible implementation, in some embodiments, determining whether the driver is in a preset abnormal state based on the monitoring data set includes: when the real-time health status data triggers the preset conflict resolution rule, calculate the current comprehensive confidence based on the new priority weight coefficient of each monitoring dimension of the monitoring data set and the preset conflict resolution rule, and determine whether the driver is in a preset abnormal state based on the current comprehensive confidence; or, when the real-time health status data does not trigger the preset conflict resolution rule, calculate the current comprehensive confidence based on the initial priority weight coefficient of each monitoring dimension of the monitoring data set, and determine whether the driver is in a preset abnormal state based on the current comprehensive confidence.

[0076] Specifically, when the preprocessed real-time health status data triggers the preset conflict resolution rule, the system can calculate to obtain the current comprehensive confidence according to the new priority weight coefficient of each monitoring dimension in the monitoring data set and the preset conflict resolution rule. Subsequently, based on this comprehensive confidence, the system can determine whether the driver is in a preset abnormal state. Conversely, when the preprocessed real-time health status data does not trigger the preset conflict resolution rule, the system can calculate the current comprehensive confidence based on the initial priority weight coefficient of each monitoring dimension in the monitoring data set and determine whether the driver is in a preset abnormal state accordingly.

[0077] It should be noted that in the embodiments of the present application, when the comprehensive confidence is greater than or equal to 60%, it is determined whether the driver is in a preset abnormal state.

[0078] In step S103, when the driver is in a preset abnormal state, determine the current warning level, and select corresponding assisted driving strategies and corresponding emergency rescue strategies based on the current warning level and the current driving scenario type.

[0079] That is to say, when abnormal conditions are detected in the driver's physiological indicators or behavior states, such as a too high heart rate, a decreased blood oxygen saturation, a too high body temperature, driver fatigue or distraction, sudden fainting of the driver, etc., the system can fuse multi-sensor data, calculate the comprehensive confidence level according to the current driving scenario of the vehicle, and determine the current warning level based on the comprehensive confidence level to trigger corresponding response strategies (i.e., assisted driving strategies and emergency rescue strategies), including vehicle control assistance, intelligent route planning, emergency rescue signal sending, etc. In addition, the heterogeneous network collaborative communication technology can be used to save the driver's health status information for joint analysis by medical institutions and traffic police.

[0080] Among them, the heterogeneous network collaborative communication technology adopts a multi-mode communication redundancy design, including a main channel, a backup channel, and data lightweighting. The main channel can upload structured abnormal data packets (including encrypted physiological waveform segments, pupil image timestamps) through C-V2X (Cellular Vehicle-to-Everything, a cellular network-based vehicle wireless communication technology); the backup channel can use the Beidou RDSS (Radio Determination Satellite Service) short message to transmit key text information (driver ID + GPS coordinates + abnormal type code); data lightweighting is used to perform compressive sensing reconstruction on ECG signals (compression ratio > 10:1) to adapt to the narrowband communication environment.

[0081] Intelligently package the driver's health status information, automatically generate an electronic health record (including PPG / ECG trend charts, body temperature heat maps) that complies with the HL7 FHIR standard, and embed vehicle OBD-II (On-Board Diagnostics II, the second-generation vehicle automatic diagnostic system) fault codes and operation logs for joint analysis by medical institutions and traffic police.

[0082] As a possible implementation method, in some embodiments, when the driver is in a preset abnormal state, determining the current warning level includes: determining a hierarchical response threshold interval according to the current comprehensive confidence level; determining the current warning level based on the hierarchical response threshold interval.

[0083] Specifically, the current comprehensive confidence level calculated based on the monitoring data set can determine the grading response threshold interval it belongs to. Furthermore, the current alarm level can be determined according to the locked grading response threshold interval, and the corresponding response strategy can be triggered. Among them, the relationship between the grading response threshold interval and the alarm level is shown in Chart 2:

[0084] Table 2

[0085] Graded response threshold range Alarm level (corresponding response actions) 60％-75％ Level 1 alarm (HUD prompt + seat vibration) 75％-90％ Level 2 alarm (Automatic speed reduction + contact emergency contact) ≥90％ Level 3 alarm (Full-automatic driving takeover + hospital navigation)

[0086] The following elaborates on the assisted driving strategies in the graded alarm response in combination with Table 2.

[0087] Level 1 alarm [60%-75%): In terms of vehicle control: Maintain the driver's operation authority, but apply reverse torque through the steer-by-wire system to assist in correcting deviations; In terms of route planning: Highlight the nearest rest area in the HUD (Head Up Display) projection, pre-load the navigation path (do not actively switch), and the seat vibrates simultaneously.

[0088] Level 2 alarm [75%-90%): In terms of vehicle control: Gradually take over control, and start the "semi-automatic driving mode" (longitudinal control is taken over, and lateral control requires driver confirmation); In terms of path planning: Combine the driver's real-time physiological data (such as the blood oxygen decline rate) to dynamically adjust the priority of the target hospital (distance priority → treatment capacity priority).

[0089] Level 3 alarm (≥90%): In terms of vehicle control: Take over full control and safety protection, activate the defibrillation protection coating of the steering wheel ECG electrode to prevent the risk of electric shock when the driver suddenly has a cardiac arrest; In terms of path optimization: Automatically select a collision avoidance strategy (lane keeping / emergency stop / ramps in escape lanes) according to the collision risk level (L1-L5).

[0090] For the emergency rescue strategy, when the driving scenario type is a high-speed cruise scenario, it is preferred to plan an escape lane rather than an emergency stop to avoid secondary accidents; when the driving scenario type is a night driving scenario, broadcast a distress signal to all vehicles within 2 km through V2X (Vehicle to Everything), triggering the double flash warning of surrounding vehicles; when the driving scenario type is a congested road section scenario, link with the traffic signal control system to open a "green wave" traffic lane for rescue vehicles.

[0091] In step S104, control the current vehicle to drive based on the corresponding assisted driving strategy, and / or conduct emergency rescue for the driver based on the corresponding emergency rescue strategy.

[0092] Specifically, when a driver's health problem is detected, the driving state of the current vehicle is controlled by implementing corresponding assisted driving strategies to ensure the safety and stability of the vehicle under various road and traffic conditions. In addition, by implementing corresponding emergency rescue strategies, measures can be quickly taken to assist the driver when a driver's health problem or an emergency situation of the vehicle is detected, so as to ensure the safety of the driver.

[0093] Further, in some embodiments, before obtaining the health status data of the current driver, it further includes: determining whether the current vehicle is in the driving stage and whether the face of the current driver is in a preset recognition area; when the current vehicle is in the driving stage and the face of the current driver is in the preset recognition area, obtaining the health status data of the current driver.

[0094] It can be understood that before obtaining the health status data of the current driver, the system can perform a series of inspection steps, including: first, determining whether the current vehicle is in the driving stage, which is to ensure that the real-time health status data obtained in the embodiments of the present application is the real-time reflection of the driver during driving; second, checking whether the face of the current driver is within the preset recognition area, which is to ensure the accuracy of face recognition and the effectiveness of data collection. Only when both of these conditions are met, that is, the current vehicle is in the driving stage and the face of the current driver is in the preset recognition area, can the system continue to execute and obtain the real-time health status data of the current driver, so as to ensure timely and accurate monitoring of the driver's health status and improve driving safety.

[0095] In addition, the embodiments of the present application also integrate functions such as biofeedback-driven environmental regulation, human-machine co-driving collaborative interface, and post-rescue effect management. These functions can not only provide timely health monitoring and rescue support for the driver in case of emergency, but also reduce the driver's sense of tension and anxiety through intelligent adjustment of the vehicle interior environment, mood soothing technology, etc., and at the same time realize event backtracking and system self-optimization, providing strong support for the driver's recovery.

[0096] Specifically, biofeedback-driven environmental regulation includes the following aspects: (1) Physiological-environmental coupling control, which dynamically adjusts the seat and seat belt based on the entropy value of grip force distribution and heart rate variability (HRV). For example, when the driver is in a tense state (entropy value > 70), the lateral wing support force (simulating a hugging feeling) and the graded release of seat belt pre-tensioning force can be increased; when the driver is incapacitated (HRV < 20 ms), the seat is automatically flattened to the recovery position (15° head-down and feet-up), and the medical device interface is unlocked. (2) Multimodal emotion intervention can be combined with facial micro-expression recognition (such as the frequency of frowning) and voice emotion analysis (the jitter rate of fundamental frequency). For example, when the driver's anxiety index > the preset threshold, linalool odor can be released (sedative effect) + the seat surface temperature is raised to 38°C (simulating human contact); when the driver is confused, a directional sound field can be activated (focusing the alarm sound on the driver's head area to avoid scaring passengers). The system automatically adjusts the posture and angle of the seat and the position and tightness of the seat belt according to the driver's physical condition and comfort needs to ensure that the driver can maintain a relatively comfortable state and receive maximum protection in an emergency; in addition, corresponding measures can be taken to relieve the driver's emotions according to different emotional states.

[0097] The human-machine co-driving collaborative interface includes the following aspects: (1) Intention prediction and fault tolerance control. When the system detects that the driver attempts to forcibly regain control of the vehicle, a pupil light reflex test can be performed through the DMS module. If the reaction is slow, the takeover state is maintained; if the reflex is normal but there is a behavior conflict, the "cooperative driving mode" can be activated (allowing the driver to correct the path but restricting the acceleration < 0.3g). (2) Cross-modal interaction protocol, which sets an emergency voice command set by defining a finite state machine to respond to specific keywords (such as "help" corresponding to a level-three alarm response, "dizzy" corresponding to a level-two alarm response, etc.). (3) Tactile feedback coding, which transmits the road condition risk level through the vibration frequency of the steering wheel (such as 5 Hz indicating traffic congestion ahead, 10 Hz indicating an ambulance approaching, etc.). The system can automatically activate the human-machine co-driving mode, judge whether the self-driving system needs to fully take over the vehicle or the degree of taking over the vehicle by interpreting the driving behavior, so as to ensure the safety of vehicle driving.

[0098] The post-rescue effect management includes the following aspects: (1) Event backtracking and system self-optimization. By storing multi-modal data (including marginal cases without triggering alarms) for 30 minutes before and after an emergency event, it is used to optimize the γ coefficient (historical risk factor) in the dynamic weight allocation formula; training a Generative Adversarial Network (GAN) to simulate sensor failure situations in extreme scenarios. (2) Driver rehabilitation support. Generate a "Health Event Report" (including heart rate turbulence maps and stress response time series), and store it on the blockchain for insurance companies to call; link with the in-vehicle entertainment system to automatically avoid the accident section during subsequent driving and provide progressive exposure therapy training. In addition, when professional rescue personnel arrive at the scene and complete the rescue task, the system can receive a confirmation message of the end of the rescue and automatically return to the standby state, waiting for the next emergency rescue task.

[0099] According to the emergency rescue method based on driver health monitoring proposed in the embodiments of the present application, by obtaining the real-time health status data of the driver during the driving process and the current driving scene type, the weight of each monitoring dimension of the real-time health status data can be allocated based on the current driving scene type, so as to obtain a monitoring data set corresponding to the current driving scene type; then, when it is determined that the driver is in a preset abnormal state based on the monitoring data set, the current alarm level is determined, and the corresponding assisted driving strategy and the corresponding emergency rescue strategy are selected based on the current alarm level and the current driving scene type; the current vehicle is controlled to drive based on the corresponding assisted driving strategy, and / or the driver is given an emergency rescue based on the corresponding emergency rescue strategy. Thereby, the problem of traffic accidents caused by the untimely monitoring of the abnormal state of the driver in the prior art is solved, and the driving safety is improved.

[0100] Next, refer to the drawings to describe the emergency rescue device based on driver health monitoring proposed in the embodiments of the present application.

[0101] Figure 2 It is a block diagram of an emergency rescue device based on driver health monitoring according to an embodiment of the present application.

[0102] As Figure 2 shown, the emergency rescue device 10 based on driver health monitoring includes: an acquisition module 100, a judgment module 200, a selection module 300, and a processing module 400.

[0103] Among them, the acquisition module 100 is used to obtain the real-time health status data of the driver during the driving process and the current driving scene type, and allocate weights to each monitoring dimension of the real-time health status data based on the current driving scene type, so as to obtain a monitoring data set corresponding to the current driving scene type;

[0104] A judgment module 200, configured to judge whether a driver is in a preset abnormal state based on a monitoring data group;

[0105] A selection module 300, configured to determine a current alarm level in the case that the driver is in a preset abnormal state, and select a corresponding assisted driving strategy and a corresponding emergency rescue strategy based on the current alarm level and the current driving scenario type;

[0106] A processing module 400, configured to control the current vehicle to travel based on the corresponding assisted driving strategy, and / or perform emergency rescue on the driver based on the corresponding emergency rescue strategy.

[0107] Further, in some embodiments, the obtaining module 100 is specifically configured to:

[0108] Set different initial priority weight coefficients for each monitoring dimension of the health status data based on the driving scenario type and a preset trigger condition, where the preset trigger condition corresponds to the driving scenario type;

[0109] Judge whether the real-time health status data triggers a preset conflict resolution rule based on the current driving scenario type;

[0110] If the real-time health status data triggers the preset conflict resolution rule, adjust the initial priority weight coefficient of each monitoring dimension based on the current driving scenario type and the real-time health status data to obtain a new priority weight coefficient, otherwise keep the initial priority weight coefficient of each monitoring dimension unchanged.

[0111] Further, in some embodiments, the judgment module 200 is specifically configured to:

[0112] In the case that the real-time health status data triggers the preset conflict resolution rule, calculate a current comprehensive confidence level based on the new priority weight coefficient of each monitoring dimension of the monitoring data group and the preset conflict resolution rule, and judge whether the driver is in a preset abnormal state based on the current comprehensive confidence level;

[0113] Or, in the case that the real-time health status data does not trigger the preset conflict resolution rule, calculate a current comprehensive confidence level based on the initial priority weight coefficient of each monitoring dimension of the monitoring data group, and judge whether the driver is in a preset abnormal state based on the current comprehensive confidence level.

[0114] Further, in some embodiments, the selection module 300 is specifically configured to:

[0115] Determine a hierarchical response threshold interval according to the current comprehensive confidence level;

[0116] Determine the current alarm level based on the hierarchical response threshold interval.

[0117] Further, in some embodiments, before obtaining the health status data of the current driver, the obtaining module 100 is further configured to:

[0118] Determine whether the current vehicle is in a driving stage and whether the face of the current driver is in a preset recognition area;

[0119] When the current vehicle is in a driving stage and the face of the current driver is in a preset recognition area, obtain the health status data of the current driver.

[0120] It should be noted that the foregoing explanation of the embodiments of the emergency rescue method based on driver health monitoring also applies to the emergency rescue device based on driver health monitoring of this embodiment, and will not be elaborated here.

[0121] According to the emergency rescue device based on driver health monitoring proposed in the embodiments of the present application, by obtaining the real-time health status data of the driver during the vehicle driving process and the current driving scene type, the weight can be assigned to each monitoring dimension of the real-time health status data based on the current driving scene type, so as to obtain a monitoring data group corresponding to the current driving scene type; then, when it is determined that the driver is in a preset abnormal state based on the monitoring data group, determine the current alarm level, and select the corresponding assisted driving strategy and the corresponding emergency rescue strategy based on the current alarm level and the current driving scene type; control the current vehicle to drive based on the corresponding assisted driving strategy, and / or perform emergency rescue on the driver based on the corresponding emergency rescue strategy. Thereby, the problem of traffic accidents caused by the untimely monitoring of the abnormal state of the driver in the prior art is solved, and the driving safety is improved.

[0122] Figure 3 The structural schematic diagram of the vehicle provided by the embodiments of the present application. The vehicle may include:

[0123] A memory 301, a processor 302, and a computer program stored on the memory 301 and executable on the processor 302.

[0124] When the processor 302 executes the program, it implements the emergency rescue method based on driver health monitoring provided in the above embodiments.

[0125] Further, the vehicle further includes:

[0126] A communication interface 303 for communication between the memory 301 and the processor 302.

[0127] The memory 301 is used to store a computer program executable on the processor 302.

[0128] The memory 301 may include a high-speed RAM (Random Access Memory) memory and may also include a non-volatile memory, such as at least one disk memory.

[0129] If the memory 301, the processor 302, and the communication interface 303 are implemented independently, the communication interface 303, the memory 301, and the processor 302 can be interconnected through a bus and communicate with each other. The bus can be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of simplicity of representation, Figure 3 only a thick line is used to represent it in the figure, but it does not mean that there is only one bus or one type of bus.

[0130] Optionally, in a specific implementation, if the memory 301, the processor 302, and the communication interface 303 are integrated on a chip, the memory 301, the processor 302, and the communication interface 303 can communicate with each other through an internal interface.

[0131] The processor 302 may be a CPU (Central Processing Unit), or an ASIC (Application Specific Integrated Circuit), or one or more integrated circuits configured to implement the embodiments of the present application.

[0132] The embodiments of the present application also provide a computer-readable storage medium, on which a computer program is stored, and when the program is executed by a processor, the emergency rescue method based on driver health monitoring as described above is implemented.

[0133] In addition, the terms "first" and "second" are only used for descriptive purposes and cannot be construed as indicating or implying relative importance or implicitly specifying the quantity of the indicated technical features. Thus, the features defined with "first" and "second" may explicitly or implicitly include at least one of such features. In the description of the present application, the meaning of "a plurality" is at least two, such as two, three, etc., unless otherwise specifically defined.

[0134] In the description of this specification, the descriptions referring to the terms "one embodiment", "some embodiments", "example", "specific example", or "some examples", etc., mean that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms are not necessarily directed to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described may be combined in any one or more embodiments or examples in a suitable manner. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0135] Although the embodiments of the present application have been shown and described above, it can be understood that the above embodiments are exemplary and should not be construed as limiting the present application. Those of ordinary skill in the art can make changes, modifications, substitutions, and variations to the above embodiments within the scope of the present application.

Claims

1. An emergency rescue method based on driver health monitoring, characterized in that: The following steps are involved: Acquire the real-time health status data and the current driving scene type of the driver during driving the vehicle, and assign weights to each monitoring dimension of the real-time health status data based on the current driving scene type to obtain a monitoring data group corresponding to the current driving scene type; Determining whether the driver is in a preset abnormal state based on the monitoring data group; When the driver is in the preset abnormal state, determining a current warning level, and selecting a corresponding auxiliary driving strategy and a corresponding emergency rescue strategy based on the current warning level and the current driving scene type; The current vehicle driving is controlled based on the corresponding assisted driving strategy, and / or emergency rescue is performed on the driver based on the corresponding emergency rescue strategy.

2. The method according to claim 1, characterized in that The weighting of each monitoring dimension of the real-time health status data based on the current driving scenario type includes: Based on the driving scenario type and the preset trigger condition, setting a different initial priority weight coefficient for each monitoring dimension of the health status data, wherein the preset trigger condition corresponds to the driving scenario type; Determining whether the real-time health status data triggers a preset conflict resolution rule based on the current driving scene type; If the real-time health status data triggers the preset conflict resolution rule, the initial priority weight coefficient of each monitoring dimension is adjusted based on the current driving scene type and the real-time health status data to obtain a new priority weight coefficient; otherwise, the initial priority weight coefficient of each monitoring dimension is maintained unchanged.

3. The method according to claim 2, characterized in that The determining whether the driver is in a preset abnormal state based on the monitoring data group includes: In the case where the real-time health status data triggers the preset conflict resolution rule, a current comprehensive confidence is calculated based on a new priority weight coefficient of each monitoring dimension of the monitoring data group and the preset conflict resolution rule, and whether the driver is in the preset abnormal state is determined based on the current comprehensive confidence; Or, when the real-time health status data does not trigger the preset conflict resolution rules, the current comprehensive confidence is calculated based on the initial priority weight coefficient of each monitoring dimension of the monitoring data group, and whether the driver is in the preset abnormal state is determined based on the current comprehensive confidence.

4. The method according to claim 3, characterized in that When the driver is in the preset abnormal state, determining the current warning level includes: Determining a graded response threshold interval according to the current comprehensive confidence; The current alarm level is determined based on the graded response threshold interval.

5. The method according to claim 1, characterized in that Before obtaining the health status data of the current driver, the method further includes: Determine whether the current vehicle is in the driving stage and whether the current driver's face is in the preset recognition area; When the current vehicle is in the driving stage and the face of the current driver is in the preset recognition area, the health status data of the current driver is obtained.

6. An emergency rescue device based on driver health monitoring, characterized in that: include: An acquisition module is used to obtain the real-time health status data and the current driving scene type of the driver during driving the vehicle, and to assign weights to each monitoring dimension of the real-time health status data based on the current driving scene type to obtain a monitoring data group corresponding to the current driving scene type; A judgment module, used for judging whether the driver is in a preset abnormal state based on the monitoring data group; A selection module, configured to determine a current warning level when the driver is in the preset abnormal state, and select a corresponding auxiliary driving strategy and a corresponding emergency rescue strategy based on the current warning level and the current driving scene type; A processing module is used to control the current vehicle driving based on the corresponding auxiliary driving strategy, and / or to perform emergency rescue for the driver based on the corresponding emergency rescue strategy.

7. The device according to claim 6, characterized in that The obtaining module is specifically used for: Based on the driving scenario type and the preset trigger condition, setting a different initial priority weight coefficient for each monitoring dimension of the health status data, wherein the preset trigger condition corresponds to the driving scenario type; Determining whether the real-time health status data triggers a preset conflict resolution rule based on the current driving scene type; If the real-time health status data triggers the preset conflict resolution rule, the initial priority weight coefficient of each monitoring dimension is adjusted based on the current driving scene type and the real-time health status data to obtain a new priority weight coefficient; otherwise, the initial priority weight coefficient of each monitoring dimension is maintained unchanged.

8. The device according to claim 7, characterized in that The judgment module is specifically used for: In the case where the real-time health status data triggers the preset conflict resolution rule, a current comprehensive confidence is calculated based on a new priority weight coefficient of each monitoring dimension of the monitoring data group and the preset conflict resolution rule, and whether the driver is in the preset abnormal state is determined based on the current comprehensive confidence; Or, when the real-time health status data does not trigger the preset conflict resolution rules, the current comprehensive confidence is calculated based on the initial priority weight coefficient of each monitoring dimension of the monitoring data group, and whether the driver is in the preset abnormal state is determined based on the current comprehensive confidence.

9. A vehicle, characterized in that: include: A memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the emergency rescue method based on driver health monitoring as described in any one of claims 1 to 5.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The program is executed by a processor to implement the emergency rescue method based on driver health monitoring as described in any one of claims 1-5.

Images