Vehicle scenario management mode switching method and system

Abstract

This invention discloses a method for switching vehicle scenario management modes: The system includes a status monitoring and mode trigger condition judgment step. After power-on, the system continuously monitors the vehicle status and occupant intentions, dynamically deciding whether to enter a rest mode or a sentry mode. The rest mode trigger condition is that the following conditions are met simultaneously: the vehicle is in P gear and under parking brake; the capacitive seat sensor detects that the occupant is in place and the backrest angle is tilted back beyond a set threshold angle; and millimeter-wave radar detects stable vital signs. The sentry mode trigger condition is that the following conditions are met simultaneously: the vehicle is in P gear and under parking brake; all doors are locked and the occupant has left their seat or the user has locked the vehicle. The advantage of this invention is that, in terms of mode coordination, the mode switching latency based on a multimodal decision engine is reduced to 0.38 ± 0.07 seconds.

Description

Technical Field

[0001] This invention relates to the field of automotive electronic control technology, and in particular to a scenario management mode switching technology for a vehicle in a parked state. Background Technology

[0002] Current vehicle scenario control technology is mainly divided into two major directions:

[0003] The first category is external security systems. The mainstream solution uses a surround-view camera array to continuously scan the vehicle's surroundings, combined with millimeter-wave radar to detect moving targets (patent number US20190283739A1). When abnormal proximity or vibration is detected, the system automatically starts high-definition recording and triggers a flashing warning light. Some improved solutions incorporate a voiceprint recognition engine (patent number KR102020008789A), which can identify specific threat audio such as broken glass or metal impacts. These systems require the image processing unit and multiple sensors to operate continuously, with typical measured power consumption reaching 200-300W (based on energy consumption test data from mid-to-large-sized electric SUVs), and rely on a high-voltage electronic and electrical architecture for continuous power supply. Recent advancements include dynamic sensitivity adjustment based on GPS geofencing (patent number US20210122374A1), but this does not solve the problem of excessively high basic energy consumption.

[0004] The second type is the occupant rest system. The mainstream technology relies on seat pressure distribution sensors to identify human posture (patent number JP2020157822A), which then triggers an electric adjustment mechanism to adjust the seat angle to a 110°-130° resting position, simultaneously controlling the air conditioning to maintain a constant temperature. An advanced solution integrates a timed wake-up function (patent number DE102019215278A1), triggering gradual seat vibration and increased air conditioning fan speed at preset time intervals. Another solution involves mobile terminal linkage (patent number US20210094421A1), allowing users to remotely switch operating modes, but this requires manual operation and involves communication delays. These systems generally lack an integrated environmental safety monitoring module and do not have an active response mechanism to changes in air quality in a closed vehicle cabin.

[0005] The existing technology system has three core bottlenecks:

[0006] 1. In terms of functional coordination, the security system and the rest system adopt a rigidly mutually exclusive architecture (such as forcibly shutting down the cabin environment control module when external monitoring is activated), which prevents occupants from resting safely in a risky environment. Conversely, mainstream rest systems lack intelligent wake-up mechanisms, and third-party crash tests show that the alarm transmission success rate is insufficient when external threats occur. The mobile terminal switching solution has an 8-12 second response delay due to network latency, which is difficult to meet the needs of handling sudden security incidents.

[0007] 2. In terms of life safety, real-time monitoring of CO concentration is generally lacking in the idling scenario of fuel vehicles. Actual tests show that the CO concentration in a closed car can reach a dangerous level of 70 ppm in 30 minutes. Moreover, the existing solutions rely on mechanical timed wake-up mechanisms. According to clinical research data from Volume 56 of the publicly available literature "Sleep Medicine", such solutions have a protection efficacy of less than 20% against risks such as sleep apnea.

[0008] 3. In terms of energy management, the security system adopts a fixed power consumption model. Basic monitoring consumes 250W of power (12kWh in 48 hours). During event response, the peak power consumption soars to 400W, causing the low-voltage battery SOC value to drop to the depletion warning line (<20%) within 24 hours. More seriously, most related patents have not established a dynamic matching mechanism between threat level and energy consumption, and cannot call different power systems according to the risk level.

[0009] Therefore, this application aims to provide a vehicle scenario management mode switching method and system that can intelligently coordinate different vehicle scenario modes, ensure the safety of occupants and achieve efficient energy management, so as to overcome the above-mentioned shortcomings of the prior art and provide a safer, more efficient and intelligent vehicle scenario management solution. Summary of the Invention

[0010] The technical problem to be solved by the present invention is to realize a vehicle scenario management mode switching method and system that can reduce the response delay during switching and achieve coordinated control of environmental risks and human body status.

[0011] To achieve the above objectives, the technical solution adopted by this invention is: a method for switching vehicle scenario management modes.

[0012] Status monitoring and mode trigger condition judgment steps: After the system is powered on, it continuously monitors the vehicle status and the occupants' intentions to dynamically decide whether to enter the rest mode or the sentry mode.

[0013] The nap mode is triggered when the following conditions are met simultaneously:

[0014] The vehicle is in P gear and the parking brake is engaged.

[0015] The capacitive seat sensor detected that the occupant was in place and the backrest angle was tilted back beyond a set threshold angle.

[0016] Millimeter-wave radar detected stable vital signs;

[0017] The Sentinel mode is triggered when the following conditions are met simultaneously:

[0018] The vehicle is in P gear and the parking brake is engaged.

[0019] All doors are locked and occupants are off their seats or the user has locked the vehicle.

[0020] It also includes: Environmental safety assessment and model final decision-making steps: After the dynamic decision is determined, the dynamic weight fusion center is launched to conduct an environmental safety assessment;

[0021] The final decision-making step in the environmental safety assessment and model involves real-time acquisition of information from vehicle external cameras and sensors, calculation of the safety factor, and calculation model.

[0022]

[0023] in The biometric safety score is based on occupant status assessment. The environmental safety score is based on an assessment of external threats. The vehicle condition safety score is based on an assessment of the vehicle's own condition. : is a dynamic weighting coefficient, and satisfies The dynamic weighting coefficients are dynamically adjusted using a preset fuzzy logic rule base of the current dominant mode.

[0024] The execution process of the rest mode is as follows:

[0025] When the nap mode is activated, the system executes the following closed-loop control:

[0026] Cabin environment optimization: Adjust the interior temperature to the set value, switch to external air circulation mode, turn the fan speed to low, and open the windows for ventilation;

[0027] Continuous monitoring of vital signs: Continuously monitor the occupants' respiratory rate, heart rate, body movement, and interior air quality;

[0028] Safety monitoring and risk assessment: Real-time fusion of vital signs and environmental data, and operation of risk assessment models;

[0029] Abnormal Intervention: When the set abnormal intervention conditions are met, execute the set intervention command;

[0030] Smart wake-up: Execute the wake-up function when the wake-up conditions are met.

[0031] In the aforementioned cabin environment optimization, window ventilation involves controlling the window motor to raise the window, leaving only a 2-4cm gap for ventilation.

[0032] During the continuous monitoring of vital signs, millimeter-wave radar monitors the occupants' respiratory rate and heart rate, capacitive seat sensors monitor body movement, and CO concentration sensors monitor the air quality inside the vehicle.

[0033] Abnormal interventions include interventions for abnormal breathing and interventions for excessive CO levels;

[0034] Intervention for abnormal breathing: If the respiratory rate is consistently below 6 breaths / minute or above 30 breaths / minute, the system determines it to be abnormal. First, the processor 201 will control the seat vibration motor 304 to provide a level 1 (mild) vibration reminder; if this is ineffective, it will be upgraded to level 2 (strong) vibration; if there is still no response, the audible and visual alarm 303 will be triggered and a warning message will be displayed on the vehicle's infotainment screen.

[0035] CO Exceedance Intervention: If the CO concentration exceeds 50 ppm, the system will immediately execute the highest priority response: control the window motor 302 to lower all windows and control the air conditioning controller 301 to open the maximum airflow external circulation to quickly dilute the harmful gas.

[0036] In the intelligent wake-up mode, the wake-up conditions include reaching a set time, sensing that the user has naturally woken up, or exiting the nap mode by manually tapping the screen. Upon intelligent wake-up, the air conditioning and windows will gradually return to the user's preferred driving state.

[0037] The Sentinel mode execution process is as follows:

[0038] When Sentinel mode is activated, the system executes the following closed-loop control:

[0039] Multi-level alert monitoring: Entering a low-power monitoring state, controlling the surround-view camera, ultrasonic radar and other sensors to periodically wake up and scan. The multi-level alert monitoring has multiple monitoring levels.

[0040] Dynamic energy management: Dynamically adjust power consumption according to the monitoring level;

[0041] Remote interaction and evidence collection: Users can selectively view the vehicle's surroundings in real time via a mobile app, conduct voice communication with the site, and download complete surveillance footage as evidence.

[0042] The multi-level alert monitoring system has three monitoring levels, including:

[0043] Level 1: When no anomalies are detected, the system keeps the core sensors and processor running at low power, continuously recording and storing the data locally;

[0044] Level 2: When the ultrasonic radar detects an object continuously approaching the vehicle within 1.5 meters, or the camera detects a person loitering for an extended period of time, the system upgrades to Level 2, triggering an audible and visual alarm to perform a strobe flash and a high-pitched siren for deterrence, and sending a real-time alert and on-site images to the vehicle owner's APP via the network.

[0045] Level 3: When the sensor detects a violent situation, the system immediately upgrades to the highest level, triggers the loudest alarm, uploads video data of one minute before and after the event to the cloud, and can automatically dial the preset emergency contact number according to user settings;

[0046] In the aforementioned dynamic energy management, at level 1, the system uses a low-voltage battery for power supply.

[0047] At level 2 or 3, the system switches to power battery supply and monitors the SOC of the power battery and low-voltage battery in real time. If the power level is lower than the safety threshold, the system will send an alarm to the user in advance and gracefully reduce the monitoring frequency or enter hibernation.

[0048] In the multi-level alert monitoring, the violent information status includes violent impact and glass breakage. The violent impact is when the speedometer senses a violent vibration, and the glass breakage is when the audio sensor identifies a specific high-frequency sound pattern.

[0049] The vehicle's scenario management mode switching system consists of a perception layer, a decision-making layer, and an execution layer.

[0050] The perception layer is used to collect raw data and transmit the collected data to the decision layer;

[0051] The decision-making layer outputs control signals to the execution layer based on the signals input from the perception layer.

[0052] The execution layer is used to receive control signals from the decision layer and perform specific operations;

[0053] The system executes the scenario management mode switching method for the vehicle as described in any one of claims 1-7.

[0054] The perception layer includes millimeter-wave radar, capacitive seat sensor, CO concentration sensor, surround view camera, and vehicle CAN bus;

[0055] The decision layer is a multimodal decision processor, which is equipped with an SPI interface for connecting millimeter-wave radar, an I2C interface for connecting capacitive seat sensors, an ADC interface for connecting capacitive seat sensors, a MIPI interface for connecting surround-view cameras, and peripheral pins for connecting the vehicle's CAN bus.

[0056] The execution layer includes an air conditioning controller, window motors, and an audible and visual alarm.

[0057] The millimeter-wave radar is installed in the roof of the vehicle, with its detection direction directed toward the driver's chest area;

[0058] The capacitive seat sensor is equipped with sensing electrodes, which are embedded in the seat back and seat cushion in a dot matrix pressure distribution manner to detect the occupant's sitting posture, position status and body movement.

[0059] The CO concentration sensor is located near the rearview mirror inside the vehicle compartment and is used to monitor the carbon monoxide concentration inside the vehicle in real time.

[0060] The surround-view cameras are arranged around the vehicle to provide surround-view video streams;

[0061] The vehicle CAN bus is used to acquire vehicle status information in real time, including vehicle speed information, gear information, door lock status information, and electronic parking brake status information.

[0062] The air conditioner controller is used to control the air volume, internal / external circulation mode, and temperature of the air conditioner.

[0063] The window motor is used to drive the window to rise and fall.

[0064] The sound and light alarm includes a loudspeaker and an LED light strip.

[0065] The advantages of this invention are as follows: In terms of mode coordination, the mode switching latency based on the multimodal decision engine is reduced to 0.38±0.07 seconds (test conditions: 25℃±3℃ environment, CANoe15.0 simulation platform, average of 100 cycles), which is more than 20 times faster than the existing technology, and the resource occupancy rate of the local alert mode is reduced to 18.5%; In terms of life safety protection, the idling airtight test of fuel vehicles (GB / T 28370 standard) shows that the response time for CO exceeding the standard (>50ppm) is ≤2.8 seconds (sensor: Alphasense CO-B4), and the detection rate of apnea reaches 100% (comparison device: Philips Alice 6 medical respiratory monitor, sample size n=50); In terms of energy efficiency optimization, referring to SAE The J2982 standard was tested for 48 hours (ambient temperature 25℃, initial SOC 80%), and the basic monitoring energy consumption was only 0.6kWh (12kWh for existing technology), achieving an energy saving rate of 95%, while the event response energy consumption was reduced to 0.05kWh / event (a reduction of 87.5%). Attached Figure Description

[0066] The following is a brief explanation of the content represented by each figure in this specification:

[0067] Figure 1 This is a schematic diagram of the overall system architecture of this application;

[0068] Figure 2 This is a flowchart illustrating the dynamic mode collaboration process of this application;

[0069] Figure 3 This is the multimodal perception fusion logic diagram of this application;

[0070] Figure 4 This is a schematic diagram of the threat assessment decision tree for this application;

[0071] Figure 5 This is a timing diagram for the security response control of this application. Detailed Implementation

[0072] The following description, with reference to the accompanying drawings, details the specific implementation of the present invention, including the shape and structure of each component, the relative positions and connections between the parts, the function and working principle of each part, the manufacturing process, and the operation and use methods, to help those skilled in the art to have a more complete, accurate, and in-depth understanding of the inventive concept and technical solution of the present invention.

[0073] This invention is a multimodal scenario management system and method for automobiles. It monitors the driver's state using millimeter-wave radar, capacitive seat sensors, and other biosensors, and calculates a dynamic safety factor based on an environmental threat model. This enables intelligent switching between rest mode, sentry mode, and local alert mode. It solves the problem of the inability to operate external threat monitoring and occupant rest functions in parallel, reduces response latency during mode switching, and achieves coordinated control of environmental risks and human condition. Simultaneously, it addresses the shortcomings of CO concentration monitoring and non-contact vital sign detection in idling scenarios of fuel-powered vehicles, improves protection against the risks of deep sleep, optimizes the energy consumption architecture of sentry mode, and establishes a dynamic matching mechanism between threat level and energy supply to avoid the risk of low-voltage battery depletion.

[0074] This invention achieves a vehicle scenario management mode switching method and system that reduces response latency during switching and enables coordinated control of environmental risks and human body status through the following core technological breakthroughs.

[0075] 1. Multimodal dynamic decision-making mechanism: Based on three-dimensional perception data of biological signs, environmental threats, and vehicle status, a safety factor quantification model (0.0-1.0) is constructed. Through the decision tree algorithm, the system realizes zero-delay autonomous switching between rest mode, sentry mode, and local alert mode (switching time <0.5 seconds).

[0076] 2. Mobile convergence point mechanism: Deploy non-contact millimeter-wave radar respiratory monitoring modules (accuracy ±0.5 times / minute) and electrochemical CO concentration sensors to establish a three-tiered response chain of "monitoring-early warning-intervention":

[0077] 3. Threat-Driven Dynamic Energy Architecture: The dynamic energy architecture utilizes different energy supply systems based on the threat level.

[0078] When the threat level is Level 1, the power supply system is a low-voltage battery with a power consumption limit of 50W and a functional range of basic radar scanning.

[0079] When the threat level is level 2, the power supply system is the main power supply of the power battery, with a power consumption limit of 200W, and the function range is all sensors + audible and visual alarms.

[0080] When the threat level is level 3, the power supply system is a backup power source for the power battery, with a power consumption limit of 2000W, and the functional scope is V2X broadcasting + high-voltage defense equipment.

[0081] 4. Localized alertness sub-mode: Maintaining selective perception ability while the occupant is at rest:

[0082] Voiceprint whitelist filtering: retains key audio such as children's cries and cries for help;

[0083] Pre-stored license plate recognition: No alarm is triggered when an authorized vehicle approaches;

[0084] Biometric authentication: Seatbelt pretensioning when a non-owner touches the driver's side window;

[0085] See Vehicle Intelligent Context Management System Figure 1 The hardware architecture of this system is divided into a perception layer, a decision-making layer, and an execution layer. The perception layer, decision-making layer, and execution layer are described in detail below:

[0086] The perception layer is responsible for collecting various types of raw data, including: biosignal detection, environmental perception, and vehicle status. Biosignal detection transmits the collected biological signals to the decision layer, environmental perception transmits the collected environmental signals to the decision layer, and vehicle status transmits vehicle status data to the decision layer.

[0087] The biometric detection includes millimeter-wave radar 101 and capacitive seat sensor 102; environmental perception includes CO concentration sensor 103 and surround view camera 104; and vehicle status is mainly obtained by CAN bus to acquire vehicle status data information.

[0088] Millimeter-wave radar 101: The millimeter-wave radar 101 is used to determine whether there are people inside the vehicle. It preferably uses Infineon's BGT60LTR11AIP 60GHz radar chip, positioned in the vehicle's ceiling facing the driver's chest area. It is used for non-contact detection of the micro-Doppler effect caused by breathing, outputting raw I / Q data signals. Its power supply pin is connected to the vehicle's +12V power supply, and its data output pin is connected to the decision processor 201 via the SPI bus.

[0089] Capacitive seat sensor 102: The capacitive seat sensor 102 is used to detect whether there is a person in the rear seat; preferably, it uses TI's FDC1004 capacitive digital converter chip, whose sensing electrodes are embedded in the seat back and seat cushion to form a 256-dot matrix pressure distribution map for detecting the occupant's sitting posture, presence status, and body movement. Its I2C communication interface is connected to the decision processor 201.

[0090] CO concentration sensor 103: It adopts the CO-B4 electrochemical sensor of Alphasense (UK), which is placed near the rearview mirror in the car. Its analog voltage output pin is connected to the ADC0 analog input channel of the decision processor 201 to monitor the carbon monoxide concentration in the car in real time, with a range of 0-2000ppm.

[0091] Surround-view camera 104: Employing an ON Semiconductor AR0234AT CMOS image sensor, it is positioned around the vehicle to provide a surround-view video stream. Its MIPI CSI-2 interface connects to the image processing unit of the decision processor 201.

[0092] Vehicle CAN bus 105: Directly connected to the CAN controller (such as MCP2515) of the decision processor 201, used to acquire vehicle status information such as vehicle speed, gear, door lock status, and electronic parking brake status in real time.

[0093] A Global Positioning System (GPS) and an Inertial Measurement Unit (IMU) can also be configured. GPS provides vehicle position information; the IMU provides vehicle attitude angle information; and the Electronic Stability System (ESS) provides vehicle driving status information. The results from the millimeter-wave radar 101 and the capacitive seat sensor 102 are only for reference, with high fault tolerance and false alarm rate. However, the results obtained by combining GPS, IMU, and ESP are generally considered accurate, so the decision-making module should assign appropriate weights during the comprehensive evaluation. The preferred solution is that the perception layer includes millimeter-wave radar, capacitive seat sensor, GPS, IMU, and ESS.

[0094] The decision-making layer is the core of the system. It connects to and acquires data information from the perception layer, namely, acquiring vital sign detection data from millimeter-wave radar 101 and capacitive seat sensor 102, environmental perception data from CO concentration sensor 103 and surround view camera 104, and vehicle status data information acquired by CAN bus.

[0095] The core of the decision-making layer is a multimodal decision processor 201. This processor is preferably an NXP S32G274A automotive-grade processor. This processor integrates multi-core ARM Cortex-A53 and Cortex-M7 cores and has a built-in hardware security module. The SPI interface of the millimeter-wave radar 101, the I2C interface of the capacitive seat sensor 102, the ADC interface of the CO concentration sensor 103, the MIPI interface of the surround-view camera 104, and the CAN bus controller are all connected to the corresponding peripheral pins of the processor 201. The processor 201 runs a Linux RTOS real-time operating system, responsible for running all the algorithms of this invention. The decision-making layer also includes a dynamic weight fusion center.

[0096] The execution layer is responsible for receiving instructions from the decision-making layer and executing specific operations. The execution layer connects to and obtains control signals issued by the decision-making layer, and controls the corresponding front-end mechanisms to execute control signals. The control signals include control instructions and trigger signals. The control instructions are output to the cabin environment control, and the trigger signals are sent to the active security system. The execution layer of the cabin environment control includes the air conditioning controller 301 and the window motor 302. The execution layer of the active security system includes the sound and light alarm.

[0097] Air conditioning controller 301: Connected to decision processor 201 via LIN bus. Receives instructions from processor 201 to control air conditioning airflow, internal / external circulation mode, and temperature.

[0098] Window motor 302: Its controller is connected to the decision processor 201 via a CAN bus. It receives window raising and lowering commands from the processor 201.

[0099] The audible and visual alarm 303 includes a tweeter (such as Pioneer's TS-B101PRO) and an LED light strip. Its control terminal is connected to the decision processor 201 via a GPIO pin, and the processor 201 directly outputs a PWM wave to control its sound frequency and light flashing mode.

[0100] For scenario initialization and dynamic decision-making, see [link to relevant documentation]. Figure 2 The control methods for scenario mode initialization and dynamic decision-making include: judgment of parking status, judgment of biometric detection, judgment of environmental threat assessment, and then, based on the environmental threat assessment, activation of rest mode, mandatory sentry mode and local alert mode are executed within minutes.

[0101] After the system is powered on, it continuously monitors the vehicle status and the occupants' intentions, and dynamically decides whether to enter rest mode or sentry mode.

[0102] S201: Status monitoring and mode trigger condition judgment:

[0103] Nap mode trigger: When the vehicle is in P gear and the electronic parking brake is engaged, the capacitive seat sensor 102 detects that the occupant is in place and the backrest angle is tilted back beyond the set threshold (for example, the threshold is set to 55 degrees, which is no longer suitable for use in driving mode, and can only be used in driving mode), and the millimeter-wave radar 101 detects stable vital signs, the system determines that the user intends to take a nap and prepares to activate the nap mode.

[0104] Sentry mode is triggered when the vehicle is in P gear and the electronic parking brake is engaged, but the system detects that all doors are locked and the occupants have left their seats (the pressure of the capacitive seat sensor 102 has disappeared), or when the user explicitly issues a lock command via the smart key / mobile app, the system prepares to activate Sentry mode.

[0105] S202: Environmental Safety Assessment and Model Final Decision:

[0106] After the aforementioned basic triggering conditions are met, the decision processor 201 initiates the dynamic weight fusion center to conduct an environmental security assessment. This algorithm analyzes data from external sensors such as the surround-view camera 104 and ultrasonic radar in real time, calculating a security factor S between 0.0 and 1.0. The calculation model for this factor is as follows:

[0107] ;

[0108] in The biometric safety score is based on occupant status assessment. The environmental safety score is based on an assessment of external threats. The vehicle condition safety score is based on an assessment of the vehicle's own condition. : is a dynamic weighting coefficient, and satisfies The weights are not fixed, but dynamically adjusted using a fuzzy logic rule base based on the current dominant pattern. For example, when the occupants are in position... The weight increases; when passengers leave their seats, and The weighting has increased.

[0109] This is derived from a simple binary decision table:

[0110] The patient is in an in-situ state (P_occupant) of 1, a normal breathing state (S_breath) of 1, and a backward posture state (P_posture) of 1. At this time, the biosafety score (S_bio) is 1.0, indicating that the system response is in an ideal resting state.

[0111] The patient is in an in-situ state (P_occupant) of 1 (in-situ), a normal breathing state (S_breath) of 1 (normal), and an upright posture state (P_posture) of 0 (upright). At this time, the biosafety score (S_bio) is 0.6, indicating that the system response is suspected to be resting but not relaxed.

[0112] The system is in an in-situ state (P_occupant) of 1 (in-situ), a breathing state (S_breath) of 0 (abnormal), and a posture state (P_posture) of X (arbitrary). At this time, the biosafety score (S_bio) is 0.1, indicating that the system response is abnormal breathing, which is a high risk.

[0113] In the occupant state (P_occupant) 0 (out of position), the breathing state (S_breath) X (arbitrary), and the posture state (P_posture) X (arbitrary), the biosafety score (S_bio) is 0.0, indicating that the system response is that the occupant has left their seat.

[0114] In summary, this can be represented in a table as follows:

[0115]

[0116] A weighted summation method is typically used, with appropriate weights assigned to different features. The formula is:

[0117] The input features (all of which need to be normalized to [0,1]) are:

[0118] F_dist: Distance to the nearest moving object / pedestrian (the closer the object / pedestrian, the lower the value).

[0119] F_sound: Intensity of abnormal ambient sounds (decibels or confidence level for specific voiceprint recognition; the more abnormal the sound, the lower the value).

[0120] F_weather: Weather severity (judged based on a combination of factors including wind speed and precipitation; the more severe the weather, the lower the value).

[0121] Weighting (totaling 1):

[0122] W_dist=0.4 (Distance is the direct threat and has the highest weight)

[0123] W_sound=0.3 (Abnormal sounds are an important auxiliary judgment)

[0124] W_weather=0.3 (Weather changes slowly, usually representing background risk)

[0125] The calculation of S_veh typically employs the minimum value method (the barrel principle), meaning that the weakest link determines the overall safety of the vehicle. The formula is:

[0126]

[0127] in:

[0128] S_co: Safety score for CO concentration. 1.0 when C_co < 50 ppm; 0.0 when C_co > 100 ppm; linear mapping for intermediate values.

[0129] S_soc: State of Charge safety score. It is 1.0 when SOC_battery > 70%; it is 0.0 when SOC_battery < 30%; linearly mapped for intermediate values.

[0130] S_doors: Door lock safety score. It is 1.0 when all doors are locked, otherwise 0.0.

[0131] If S ≥ 0.8, it indicates that the environment is very safe, and the processor 201 finally activates the nap mode.

[0132] If S ≤ 0.3, it indicates that there is a high risk in the environment. The processor 201 will forcibly activate the sentry mode and can remind the occupants through the acoustic and optical alarm 303.

[0133] If 0.3 < S < 0.8, it indicates that there is uncertainty in the environment. The processor 201 will activate the local alert mode (a hybrid mode).

[0134] The specific execution process of the nap mode (corresponding to steps S201 - S205)

[0135] When the nap mode is activated, the system performs the following closed-loop control:

[0136] S301: Cabin environment optimization: The processor 201 sends instructions to the air-conditioning controller 301 to adjust the vehicle interior temperature to the set value (such as 24°C), switch to the external circulation mode, and adjust the wind speed to low gear to ensure air circulation. At the same time, it can control the window motor 302 to raise the window leaving only a 2 - 4 cm gap for ventilation. The width of the gap ventilation can be set as needed, but it should ensure the safety inside the vehicle, generally not exceeding 4 cm, and preferably set a 3 cm gap for ventilation;

[0137] S302: Continuous vital sign monitoring: The millimeter-wave radar 101 continuously monitors the breathing frequency and heart rate of the occupants. During this test process, the control method used is the existing control strategy, and the capacitive seat sensor 102 monitors body movements. The CO concentration sensor 103 continuously monitors the air quality inside the vehicle, and all data is transmitted to the processor 201 in real time.

[0138] S303: Safety monitoring and risk assessment: The core function of the processor 21 - dynamic monitoring and assessment. It fuses vital signs and environmental data in real time and runs a risk assessment model.

[0139] S304: Abnormal intervention:

[0140] Intervention for abnormal breathing: If the respiratory rate is consistently below 6 breaths / minute or above 30 breaths / minute, the system determines it to be abnormal. First, the processor 201 will control the seat vibration motor 304 to provide a level 1 (mild) vibration reminder; if this is ineffective, it will be upgraded to level 2 (strong) vibration; if there is still no response, the audible and visual alarm 303 will be triggered and a warning message will be displayed on the vehicle's infotainment screen.

[0141] CO Exceedance Intervention: If the CO concentration exceeds 50 ppm, the system will immediately execute the highest priority response: control the window motor 302 to lower all windows and control the air conditioning controller 301 to open the maximum airflow external circulation to quickly dilute the harmful gas.

[0142] S305: Intelligent Wake-up: Users can exit nap mode by setting a preset time, waking up naturally (when the system detects significant body movement and seating adjustments), or manually tapping the screen. When the system exits, the air conditioning and windows will gradually return to the user's preferred driving settings to avoid sudden temperature changes.

[0143] The specific execution flow of Sentinel mode (corresponding steps S301-S303)

[0144] When Sentinel mode is activated, the system executes the following closed-loop control:

[0145] S401: Multi-level surveillance monitoring: The system enters a low-power monitoring state. Surround view cameras 104, ultrasonic radar, and other sensors periodically wake up to perform scans.

[0146] Level 1 (Basic Monitoring): When no anomalies are detected, the system only keeps the core sensors and processor running at low power, continuously recording and storing the video locally, with power consumption controlled within 50W.

[0147] Level 2 (Warning and Deterrence): When the ultrasonic radar detects an object continuously approaching the vehicle within 1.5 meters, or the camera detects someone loitering for an extended period, the system upgrades to Level 2. The audible and visual alarm 303 is triggered, executing a strobe light and a high-pitched siren (110dB) for deterrence, and sending a real-time alert and on-site images to the owner's app via the 4G network.

[0148] Level 3 (Active Defense and Alarm): When the sensors detect a violent impact (such as a violent vibration sensed by the accelerometer) or broken glass (the audio sensor identifies a specific high-frequency sound pattern), the system immediately escalates to the highest level. It triggers a maximum-volume alarm, simultaneously uploads one minute of video data before and after the event to the cloud, and can automatically dial preset emergency contact numbers according to user settings.

[0149] S402: Dynamic Energy Management: Energy consumption control in Sentinel Mode is crucial. The system dynamically adjusts power consumption based on the threat level.

[0150] At Level 1, the system uses a low-voltage battery for power and strictly limits power consumption to prevent the vehicle from running out of power.

[0151] When trigger level 2 or 3, the system will automatically switch to power battery to ensure full-power operation of security functions. Simultaneously, the system will monitor the SOC (State of Charge) of both the power battery and the low-voltage battery in real time. If the charge level falls below a safe threshold, it will send an alert to the user in advance and gracefully reduce the monitoring frequency or enter sleep mode, prioritizing vehicle starting capability.

[0152] S403: Remote Interaction and Evidence Collection: Users can remotely view the vehicle's surroundings in real time via a mobile app, conduct voice communication with the site, and download complete surveillance footage as evidence.

[0153] This application achieves three major breakthroughs compared to existing technologies: In terms of mode coordination, the mode switching latency based on the multimodal decision engine is reduced to 0.38±0.07 seconds (test conditions: 25℃±3℃ environment, CANoe15.0 simulation platform, average of 100 cycles), which is more than 20 times faster than existing technologies, and the resource occupancy rate of local alert mode is reduced to 18.5%; In terms of life safety protection, the idling closed-loop test of fuel vehicles (GB / T 28370 standard) shows that the response time for CO exceeding the standard (>50ppm) is ≤2.8 seconds (sensor: Alphasense CO-B4), and the detection rate of apnea reaches 100% (comparison device: Philips Alice 6 medical respiratory monitor, sample size n=50); In terms of energy efficiency optimization, referring to SAE... The J2982 standard was tested for 48 hours (ambient temperature 25℃, initial SOC 80%), and the basic monitoring energy consumption was only 0.6kWh (12kWh for existing technology), achieving an energy saving rate of 95%, while the event response energy consumption was reduced to 0.05kWh / event (a reduction of 87.5%).

[0154] Based on the above-mentioned vehicle scenario management mode switching method and system, the following are examples:

[0155] The P-gear parking mode trigger judgment step: Subsequent judgments are triggered only when the vehicle is in P gear and under parking brake conditions. This ensures that after parking, the user can turn off the engine, remove the key, and the vehicle is under parking brake conditions, requiring no further user intervention. This prevents the user from accidentally triggering the system's mode switching logic when the vehicle is not parked. Millimeter-wave radar vital sign detection step: Using a millimeter-wave radar installed on the roof, the system determines if there are any occupants inside the vehicle. If occupants are present, the process continues; if not, this step is retried upon the next entry into the vehicle. Capacitive seat sensor detection step: Using capacitive sensors on the rear seat surface, the system detects if anyone is sitting in the seat. If occupants are present, the process continues; if not, this step is retried upon the next entry into the vehicle. Determining if the driver has left the vehicle step: Based on the vital sign signals and seat sensor signals obtained in the above steps, the system determines the user's intended behavior. If it indicates the driver has left the vehicle, the next step is executed; otherwise, the system waits for the next trigger. Steps to obtain vehicle ground clearance: The ground clearance is obtained using the RPS sensor located under the door sill. This sensor has higher reliability and accuracy than traditional height sensors and is unaffected by vehicle height. Steps to determine if the nap mode activation conditions are met: Based on the signals obtained in the above steps, determine if the following conditions are met simultaneously;

[0156] The millimeter-wave radar did not detect any signs of life.

[0157] The capacitive seat sensor detected that the passenger was in place and the backrest angle was reclined more than 55 degrees.

[0158] The height above the ground is greater than 0.25m (for parking in ground-level or underground garages);

[0159] When all of the above conditions are met, it can be determined that the intention of this behavior is to activate the nap mode.

[0160] Steps to activate nap mode: When the conditions for enabling nap mode are met, the system will activate nap mode.

[0161] The execution of the nap mode includes, but is not limited to, the following:

[0162] Turn on the air conditioner and switch to external circulation; turn on multimedia playback; enable various active services of the intelligent assistant, such as remote control of the in-vehicle screen and audio system; for models equipped with a dashcam, allow the dashcam to record and capture images, and switch to NTC mode; for models equipped with a 360-degree panoramic view, start the 360-degree panoramic process and display the surrounding view of the vehicle; for models equipped with a V2X communication module, establish a connection with the terminal application, allow updates to navigation map data packages from the cloud, and obtain weather information and rest recommendations from the cloud; for models equipped with intelligent vehicle control, enable active safety-related ADAS functions such as lane keeping assist, brake assist, and collision avoidance assist. Determine if the user is leaving the vehicle: Again determine the user's intention. If it is a scenario of leaving the vehicle, proceed to the next step; otherwise, wait for the next trigger. Execute the rest mode termination operation: When the rest mode activation conditions are no longer met, the system stops the rest mode. When stopping the rest mode, turn off passive and active services such as multimedia playback. Combined with the above implementation methods, the beneficial effects of this invention are:

[0163] Intelligent collaboration and seamless switching between modes have been achieved: the safety coefficient S calculated by the dynamic weight fusion center enables automatic and precise switching between rest mode and sentry mode, solving the problem of functional fragmentation in existing technologies and reducing response latency from seconds to milliseconds.

[0164] A closed-loop system for life safety protection has been established: In the rest mode, the linkage between non-contact radar monitoring and CO concentration sensor forms a full-chain protection of "monitoring-early warning-intervention (ventilation / wake-up)-rescue", minimizing the safety risks of idling and resting fuel vehicles.

[0165] Excellent energy efficiency is achieved: the multi-level energy management strategy in Sentinel mode enables the system’s basic power consumption to be stably controlled below 50W, which reduces energy consumption by more than 90% compared to the traditional continuous full power consumption solution, fundamentally eliminating the risk of vehicle battery depletion, while ensuring full response in high-threat situations.

[0166] In summary, this invention provides a safe, efficient, and intelligent automotive scenario management solution through the innovative design combining hardware and software. Those skilled in the art can select suitable automotive-grade chips and sensors based on the above description, and write corresponding embedded software algorithms to implement the technical solution of this invention.

[0167] The present invention has been described above by way of example with reference to the accompanying drawings. Obviously, the specific implementation of the present invention is not limited to the above-described manner. Any non-substantial improvements made using the inventive concept and technical solution of the present invention, or the direct application of the inventive concept and technical solution of the present invention to other occasions without modification, are all within the protection scope of the present invention.

Claims

1. A method for switching vehicle scenario management modes, characterized in that: Status monitoring and mode trigger condition judgment steps: After the system is powered on, it continuously monitors the vehicle status and the occupants' intentions to dynamically decide whether to enter the rest mode or the sentry mode. The nap mode is triggered when the following conditions are met simultaneously: The vehicle is in P gear and the parking brake is engaged. The capacitive seat sensor detected that the occupant was in place and the backrest angle was tilted back beyond a set threshold angle. Millimeter-wave radar detected stable vital signs; The Sentinel mode is triggered when the following conditions are met simultaneously: The vehicle is in P gear and the parking brake is engaged. All doors are locked and occupants are off their seats or the user has locked the vehicle.

2. The vehicle scenario management mode switching method according to claim 1, characterized in that: Also includes: Environmental safety assessment and model final decision-making steps: After the dynamic decision is determined, the dynamic weight fusion center is activated to conduct the environmental safety assessment; The final decision-making step in the environmental safety assessment and model involves real-time acquisition of information from vehicle external cameras and sensors, calculation of the safety factor, and calculation model. ； in The biometric safety score is based on occupant status assessment. The environmental safety score is based on an assessment of external threats. The vehicle condition safety score is based on an assessment of the vehicle's own condition. : is a dynamic weighting coefficient, and satisfies The dynamic weighting coefficients are dynamically adjusted using a preset fuzzy logic rule base of the current dominant mode.

3. The vehicle scenario management mode switching method according to claim 1 or 2, characterized in that: The execution flow of the rest mode is as follows: When the nap mode is activated, the system executes the following closed-loop control: Cabin environment optimization: Adjust the interior temperature to the set value, switch to external air circulation mode, turn the fan speed to low, and open the windows for ventilation; Continuous monitoring of vital signs: Continuously monitor the occupants' respiratory rate, heart rate, body movement, and interior air quality; Safety monitoring and risk assessment: Real-time fusion of vital signs and environmental data, and operation of risk assessment models; Abnormal Intervention: When the set abnormal intervention conditions are met, execute the set intervention command; Smart wake-up: Execute the wake-up function when the wake-up conditions are met.

4. The vehicle scenario management mode switching method according to claim 3, characterized in that: In the aforementioned cabin environment optimization, window ventilation involves controlling the window motor to raise the window, leaving only a 2-4cm gap for ventilation. During the continuous monitoring of vital signs, millimeter-wave radar monitors the occupants' respiratory rate and heart rate, capacitive seat sensors monitor body movement, and CO concentration sensors monitor the air quality inside the vehicle. Abnormal interventions include interventions for abnormal breathing and interventions for excessive CO levels; Intervention for abnormal breathing: If the respiratory rate is detected to be consistently below 6 breaths / minute or above 30 breaths / minute, the system will determine it as abnormal. First, the processor 201 will control the seat vibration motor 304 to provide a level one vibration reminder. If this is ineffective, it will be upgraded to level two vibration. If there is still no response, the sound and light alarm 303 will be triggered and a warning message will be displayed on the vehicle screen. CO Exceedance Intervention: If the CO concentration exceeds 50ppm, the system will immediately execute the highest priority response: control the window motor 302 to lower all windows and control the air conditioning controller 301 to open the maximum airflow external circulation to quickly dilute the harmful gas; In the intelligent wake-up, the wake-up conditions include reaching a set time, sensing that the user has woken up naturally, or exiting the nap mode by clicking the screen automatically or manually. When intelligently woke up, the air conditioning and windows will slowly return to the user's usual driving state.

5. The method for switching vehicle scenario management modes according to claim 1, 2, or 4, characterized in that: The Sentinel mode execution process is as follows: When Sentinel mode is activated, the system executes the following closed-loop control: Multi-level alert monitoring: Entering a low-power monitoring state, the surround-view camera and ultrasonic radar sensor are periodically woken up to scan. The multi-level alert monitoring has multiple monitoring levels. Dynamic energy management: Dynamically adjust power consumption according to the monitoring level; Remote interaction and evidence collection: Users can selectively view the vehicle's surroundings in real time via a mobile app, conduct voice communication with the site, and download complete surveillance footage as evidence.

6. The vehicle scenario management mode switching method according to claim 5, characterized in that: The multi-level alert monitoring system has three monitoring levels, including: Level 1: When no anomalies are detected, the system keeps the core sensors and processor running at low power, continuously recording and storing the data locally; Level 2: When the ultrasonic radar detects an object continuously approaching the vehicle within 1.5 meters, or the camera detects a person loitering for an extended period of time, the system upgrades to Level 2, triggering an audible and visual alarm to perform a strobe flash and a high-pitched siren for deterrence, and sending a real-time alert and on-site images to the vehicle owner's APP via the network. Level 3: When the sensor detects a violent situation, the system immediately upgrades to the highest level, triggers the loudest alarm, uploads video data of one minute before and after the event to the cloud, and can automatically dial the preset emergency contact number according to user settings; In the aforementioned dynamic energy management, at level 1, the system uses a low-voltage battery for power supply. At level 2 or 3, the system switches to power battery supply and monitors the SOC of the power battery and low-voltage battery in real time. If the power level is lower than the safety threshold, the system will send an alarm to the user in advance and gracefully reduce the monitoring frequency or enter hibernation.

7. The vehicle scenario management mode switching method according to claim 6, characterized in that: In the multi-level alert monitoring, the violent information status includes violent impact and glass breakage. The violent impact is when the speedometer senses a violent vibration, and the glass breakage is when the audio sensor identifies a specific high-frequency sound pattern.

8. A vehicle scenario management mode switching system, which consists of a perception layer, a decision-making layer, and an execution layer; The perception layer is used to collect raw data and transmit the collected data to the decision layer; The decision-making layer outputs control signals to the execution layer based on the signals input from the perception layer. The execution layer is used to receive control signals from the decision layer and perform specific operations; Its features are: The system executes the scenario management mode switching method for the vehicle as described in any one of claims 1-7.

9. The vehicle scenario management mode switching system according to claim 8, characterized in that: The perception layer includes millimeter-wave radar, capacitive seat sensor, CO concentration sensor, surround view camera, and vehicle CAN bus; The decision layer is a multimodal decision processor, which is equipped with an SPI interface for connecting millimeter-wave radar, an I2C interface for connecting capacitive seat sensors, an ADC interface for connecting capacitive seat sensors, a MIPI interface for connecting surround-view cameras, and peripheral pins for connecting the vehicle's CAN bus. The execution layer includes an air conditioning controller, window motors, and an audible and visual alarm.

10. The vehicle scenario management mode switching system according to claim 8, characterized in that: The millimeter-wave radar is installed in the roof of the vehicle, with its detection direction directed toward the driver's chest area; The capacitive seat sensor is equipped with sensing electrodes, which are embedded in the seat back and seat cushion in a dot matrix pressure distribution manner to detect the occupant's sitting posture, position status and body movement. The CO concentration sensor is located near the rearview mirror inside the vehicle compartment and is used to monitor the carbon monoxide concentration inside the vehicle in real time. The surround-view cameras are arranged around the vehicle to provide surround-view video streams; The vehicle CAN bus is used to acquire vehicle status information in real time, including vehicle speed information, gear information, door lock status information, and electronic parking brake status information. The air conditioner controller is used to control the air volume, internal / external circulation mode, and temperature of the air conditioner. The window motor is used to drive the window to rise and fall. The sound and light alarm includes a loudspeaker and an LED light strip.

Images