Vehicle rest mode adaptive adjustment method, device, equipment and medium

Abstract

This application provides a method, apparatus, device, and medium for adaptive adjustment of vehicle rest mode, relating to the field of vehicle technology. The method includes: determining the user's identity in the vehicle when the rest mode is activated, and acquiring user state perception data and environmental perception data; generating initial adjustment parameters based on a pre-stored historical preference model corresponding to the user's identity, the deviation between the current environmental perception data and pre-stored historical environmental data, and performing adjustment according to the initial adjustment parameters; during the operation of the rest mode, monitoring the changes in preset types of perception data, and when the change in any perception data exceeds a preset threshold associated with that perception data, adjusting the parameters of the atomization capability presetly associated with that perception data, until the rest mode ends. The method of this application effectively solves the problem in related technologies where the vehicle rest mode adjustment function cannot simultaneously consider changes in user state and environmental changes, resulting in insufficient user experience.

Description

Technical Field

[0001] This application relates to the field of vehicle technology, and in particular to a method, device, equipment and medium for adaptive adjustment of vehicle rest mode. Background Technology

[0002] With the increasing prevalence of smart cars, users are spending more and more time resting in their vehicles, such as taking a lunch break, resting during long drives, or relaxing while waiting at a stop. The vehicle's rest mode adjustment function directly impacts the quality of rest and subsequent driving performance. A well-designed rest mode can automatically adjust seat position, air conditioning temperature, interior lighting, and acoustics to quickly create a comfortable resting environment, enhancing the user experience while avoiding the tediousness and interruptions of manual adjustments. Therefore, the level of intelligence in rest mode adjustment has become a crucial indicator of the vehicle's cabin experience.

[0003] In related technologies, the adjustment of vehicle rest modes generally adopts preset scripts or fixed schemes. After the mode is activated, it executes according to predetermined parameters, which cannot take into account changes in the user's state during the rest process (such as whether they are asleep or experiencing fluctuations in vital signs) or changes in the external environment (such as temperature fluctuations, noise interference, and changes in lighting). When the user's state or environmental conditions change, the system cannot make fine adjustments in real time, causing the rest environment to gradually deviate from the user's actual needs, which seriously reduces the user's overall experience. Summary of the Invention

[0004] This application provides a vehicle rest mode adaptive adjustment method, device, equipment, and medium to solve the problem that the vehicle rest mode adjustment function in related technologies cannot take into account changes in user status and environmental changes, resulting in insufficient user experience.

[0005] Firstly, this application provides a method for adaptive adjustment of a vehicle rest mode, comprising the following steps:

[0006] The system identifies the user in the vehicle when the rest mode is activated and acquires user status perception data and environmental perception data. The rest mode is the operating mode in which the vehicle provides rest services to the user.

[0007] Initial adjustment parameters are generated based on the pre-stored historical preference model corresponding to the user identity, the deviation between the current environmental perception data and the pre-stored historical environmental data, and the adjustment is executed according to the initial adjustment parameters. The adjustment is based on the atomic capability corresponding to the initial adjustment parameters, and the atomic capability is a vehicle control function unit that supports independent invocation.

[0008] During the rest mode operation, the changes in the preset types of sensor data are monitored, and when the change in any sensor data exceeds the preset threshold associated with that sensor data, the parameters of the atomization capability associated with that sensor data are adjusted until the rest mode ends.

[0009] In one embodiment of this disclosure, the adjustment is performed according to the initial adjustment parameters based on a reinforcement learning decision model. The reinforcement learning decision model takes user identity, user state perception data, and environment perception data as states, uses the invocation of atomic capabilities and parameter adjustment as actions, and aims to maximize the composite reward function. The composite reward function includes an immediate reward and a final reward. The immediate reward is determined based on whether the user manually adjusts the parameters or provides real-time feedback, and the final reward is determined based on whether the rest duration meets the recommended range and whether the user's deep rest duration meets the standard.

[0010] In one embodiment of this disclosure, when the change in any sensing data exceeds a preset threshold associated with the sensing data, the parameters of the atomization capability associated with the sensing data are adjusted. This includes: comparing the real-time changes of various types of sensing data with their corresponding trigger thresholds, wherein there is a preset mapping relationship between the sensing data and the corresponding atomization capability; when the change in a certain type of sensing data exceeds the corresponding trigger threshold, only the atomization capability that is presetly mapped to that type of sensing data is triggered to perform independent parameter adjustments, without changing the current state of other atomization capabilities.

[0011] In one embodiment of this disclosure, after the rest mode ends, the method further includes: recording the entire rest data, wherein the entire rest data includes user identification, initial adjustment parameters, dynamic adjustment records, user active intervention records, vital sign change curves, and user satisfaction feedback; and updating the user's historical preference model through incremental learning, wherein updating through incremental learning includes using the entire rest data as new training samples to progressively correct the parameters of the historical preference model, so that the model evolves smoothly with the increase in the number of uses.

[0012] In one embodiment of this disclosure, an initial adjustment parameter is generated based on a pre-stored historical preference model corresponding to the user's identity, the deviation between the current environmental perception data and the pre-stored historical environmental data, including: calling the historical preference model and obtaining the baseline adjustment parameter corresponding to the current user's identity; calculating the deviation between the current environmental perception data and the average environmental data stored in the user's historical rest scenarios; calculating the environmental correction coefficient based on the deviation; dynamically correcting the baseline adjustment parameter based on the environmental correction coefficient, and generating the initial adjustment parameter.

[0013] In one embodiment of this disclosure, acquiring user status perception data and environmental perception data includes: acquiring user identity, limb movement and eye movement data collected by an in-vehicle camera, user heart rate and body temperature data collected by a vital signs sensor, and user voice command data collected by a microphone, and integrating them to obtain user status perception data; acquiring in-vehicle temperature and humidity, light intensity and noise data collected by an in-vehicle environment sensor, and acquiring outside air quality, ambient temperature and weather data collected by an outside environment perception unit, and integrating them to obtain environmental perception data.

[0014] In one embodiment of this disclosure, after the rest mode ends, the method further includes: if an exit command is received or a preset exit condition is detected, calling the corresponding atomization capability to restore the seat posture, air conditioning settings, window light transmittance, and audio status to the state before the rest mode was entered; and collecting user satisfaction feedback on this rest based on the in-vehicle interactive interface, and using the satisfaction feedback as input data for the next incremental learning of the historical preference model.

[0015] Secondly, embodiments of this disclosure provide a vehicle rest mode adaptive adjustment device, comprising:

[0016] The acquisition module is used to determine the identity of the user in the vehicle when the rest mode is activated, and to acquire user status perception data and environmental perception data. The rest mode is the operating mode in which the vehicle provides rest services to the user.

[0017] The initial adjustment module is used to generate initial adjustment parameters based on the pre-stored historical preference model corresponding to the user identity, the deviation between the current environmental perception data and the pre-stored historical environmental data, and to perform adjustment according to the initial adjustment parameters. The adjustment is based on the atomic capability corresponding to the initial adjustment parameters, and the atomic capability is a vehicle control function unit that supports independent invocation.

[0018] The dynamic processing module is used to monitor the changes in preset types of sensor data during the rest mode operation, and when the change in any sensor data exceeds the preset threshold associated with that sensor data, it adjusts the parameters of the atomization capability associated with that sensor data until the rest mode ends.

[0019] Optionally, the initial adjustment module specifically includes: performing adjustment based on a reinforcement learning decision model according to initial adjustment parameters. The reinforcement learning decision model takes user identity, user state perception data, and environmental perception data as states, uses the invocation of atomic capabilities and parameter adjustment as actions, and aims to maximize the composite reward function. The composite reward function includes immediate reward and final reward. The immediate reward is determined based on whether the user manually adjusts the parameters or provides real-time feedback, and the final reward is determined based on whether the rest duration meets the recommended range and whether the user's deep rest duration meets the standard.

[0020] Optionally, the dynamic processing module is specifically used to compare the real-time changes of various types of sensing data with the corresponding trigger thresholds, wherein there is a preset mapping relationship between the sensing data and the corresponding atomization capabilities; when the change of a certain type of sensing data exceeds the corresponding trigger threshold, only the atomization capability that is preset-mapped to that type of sensing data is triggered to make independent parameter adjustments, without changing the current state of other atomization capabilities.

[0021] Optionally, the initial adjustment module is also used to record the entire rest data after the rest mode ends. The entire rest data includes user identification, initial adjustment parameters, dynamic adjustment records, user active intervention records, vital sign change curves, and user satisfaction feedback. The module also updates the user's historical preference model through incremental learning. Incremental learning includes using the entire rest data as new training samples to progressively correct the parameters of the historical preference model, so that the model evolves smoothly with the number of uses.

[0022] Optionally, the initial adjustment module is specifically used to: call the historical preference model to obtain the baseline adjustment parameters corresponding to the current user identity; calculate the deviation between the current environmental perception data and the average environmental data stored in the user's historical rest scenarios; calculate the environmental correction coefficient based on the deviation; dynamically adjust the baseline adjustment parameters based on the environmental correction coefficient, and generate the initial adjustment parameters.

[0023] Optionally, the acquisition module is specifically used to acquire user identity, body movements and eye movements data collected by the in-vehicle camera, user heart rate and body temperature data collected by the vital signs sensor, user voice command data collected by the microphone, and integrate them to obtain user status perception data; acquire in-vehicle temperature and humidity, light intensity and noise data collected by the in-vehicle environment sensor, and in-vehicle air quality, ambient temperature and weather data collected by the in-vehicle environment perception unit, and integrate them to obtain environmental perception data.

[0024] Optionally, the dynamic processing module is also used to, after the rest mode ends, receive an exit command or detect that the preset exit conditions are met, call the corresponding atomization capability to restore the seat posture, air conditioning settings, window light transmittance, and audio status to the state before entering the rest mode; collect user satisfaction feedback on this rest based on the in-vehicle interactive interface, and use this satisfaction feedback as input data for the next incremental learning of the historical preference model.

[0025] Thirdly, embodiments of this application provide a control device, including: a memory and a processor;

[0026] The memory stores instructions that the computer executes;

[0027] The processor executes computer execution instructions stored in memory, causing the processor to perform a vehicle rest mode adaptive adjustment method as described in the first aspect of this disclosure.

[0028] Fourthly, embodiments of this disclosure also provide a computer-readable storage medium storing computer-executable instructions, which, when executed by a processor, are used to implement the vehicle rest mode adaptive adjustment method as described in the first aspect of this disclosure.

[0029] Fifthly, embodiments of this disclosure also provide a computer program product comprising computer execution instructions, which, when executed by a processor, are used to implement the vehicle rest mode adaptive adjustment method as described in the first aspect of this disclosure.

[0030] The vehicle rest mode adaptive adjustment method, apparatus, device, and medium provided in this disclosure determine the user's identity and acquire user status and environmental perception data. Then, based on the user's historical preference model and the deviation between the current and historical environments, initial adjustment parameters are generated, and independently controllable atomization capabilities are invoked to execute the adjustment. During the rest period, the changes in various perception data are continuously monitored. When the change in any data exceeds its preset associated threshold, only the atomization capability associated with that data is independently adjusted until the rest ends. Thus, by combining historical preferences with real-time environmental deviations to generate initial parameters, the problem of traditional solutions being unable to personalize the user's current state is solved. Furthermore, by setting independent thresholds for different perception data and adjusting only the associated atomization capabilities, the problem of prior art disrupting rest continuity due to global parameter adjustments is solved. This achieves real-time, refined response to changes in user status and environment, significantly improving the accuracy and personalization level of adjustment while ensuring the continuity of the rest process. Attached Figure Description

[0031] The accompanying drawings, which are incorporated in and form part of this specification, illustrate embodiments consistent with this application and, together with the description, serve to explain the principles of this application.

[0032] Figure 1 This is an application scenario diagram of the vehicle rest mode adaptive adjustment method, device, equipment and medium provided in the embodiments of this disclosure;

[0033] Figure 2 A flowchart of a vehicle rest mode adaptive adjustment method provided in one embodiment of this disclosure;

[0034] Figure 3 A flowchart of a vehicle rest mode adaptive adjustment method provided in yet another embodiment of this disclosure;

[0035] Figure 4A schematic diagram of the structure of a vehicle rest mode adaptive adjustment device provided in yet another embodiment of this disclosure;

[0036] Figure 5 This is a schematic diagram of the structure of a control device provided in yet another embodiment of this disclosure.

[0037] The accompanying drawings illustrate specific embodiments of this application, which will be described in more detail below. These drawings and descriptions are not intended to limit the scope of the concept in any way, but rather to illustrate the concept of this application to those skilled in the art through reference to particular embodiments. Detailed Implementation

[0038] Exemplary embodiments will now be described in detail, examples of which are illustrated in the accompanying drawings. When the following description relates to the drawings, unless otherwise indicated, the same numbers in different drawings denote the same or similar elements. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with this application. Rather, they are merely examples of apparatuses and methods consistent with some aspects of this application as detailed in the appended claims.

[0039] The vehicle rest mode is designed to quickly create a suitable resting environment for users in parking scenarios. The core challenge of its adaptive adjustment lies in the fact that the user's state during rest is dynamically evolving, progressing from wakefulness to light sleep and then to deep sleep, with vital signs such as heart rate, body temperature, and limb movements constantly changing. Simultaneously, external environmental factors such as temperature, noise, and light intensity also exhibit unpredictable fluctuations. The adjustment system needs to capture these changes in real time and respond accurately without user intervention, and any adjustment action should not interrupt the user's rest continuity. This requires the system to not only identify the type and degree of change but also determine which adjustments are necessary and which should remain unchanged.

[0040] Existing technologies generally employ preset scripts or fixed schemes, executing according to predetermined parameters after the rest mode is activated. They fail to perceive real-time changes in the user's state or respond to dynamic fluctuations in the external environment. While some solutions allow users to manually adjust settings and remember preferences, their adjustment logic is essentially post-event recording rather than real-time feedback. Furthermore, adjustments are often global, such as simultaneously changing multiple parameters like seat, air conditioning, and lighting, easily disrupting the user's rest continuity. More critically, existing solutions lack a sophisticated mechanism to determine "what changes trigger what adjustments," failing to distinguish the importance and correlation of different sensory data, resulting in adjustments that are either excessively frequent or sluggish. These persistent technical obstacles make truly adaptive rest adjustment difficult to achieve.

[0041] The adaptive adjustment method, device, equipment, and medium for vehicle rest modes provided in this application decompose the adjustment process into two stages: "initialization based on historical preferences and environmental deviations" and "real-time independent adjustment based on differentiated thresholds." The former utilizes a user's historical preference model and incorporates current environmental deviation corrections to achieve personalized initial state settings; the latter sets independent thresholds for each type of sensing data. When a data point exceeds its limit, only its associated atomic capabilities are adjusted. This ensures continuous rest while responding to changes in the environment and the user in real time, avoiding cascading changes in irrelevant parameters.

[0042] Figure 1 This is a schematic diagram illustrating the application scenario of the vehicle rest mode adaptive adjustment method, device, equipment, and medium provided in this application, such as... Figure 1 As shown, during vehicle operation, after receiving the instruction to start the rest mode, the vehicle control system 100 will identify the user 110 and automatically monitor the user status perception data 120 and environmental perception data 130 to determine the corresponding in-vehicle environmental parameter adjustment command 140, so as to achieve adaptive adjustment of the vehicle rest mode until the rest mode ends.

[0043] It should be noted that, Figure 1 The scenario shown includes vehicle control system, user identity, user status perception data, and environmental perception data. Only one or a specific number of these are used as examples for illustration, but this disclosure is not limited to this. That is to say, the number of vehicle control system, user identity, user status perception data, and environmental perception data can be arbitrary.

[0044] The technical solution of this application and how the technical solution of this application solves the above-mentioned technical problems are described in detail below with specific embodiments. These specific embodiments can be combined with each other, and the same or similar concepts or processes may not be described again in some embodiments. The embodiments of this application will now be described with reference to the accompanying drawings.

[0045] Figure 2 This is a flowchart illustrating the adaptive adjustment method for vehicle rest mode provided in this application embodiment. The following is a summary of the process. Figure 2 The main process of the adaptive adjustment method for vehicle rest mode is explained below:

[0046] S201. Determine the identity of the user in the vehicle when the rest mode is activated, and obtain user status perception data and environmental perception data.

[0047] Among them, the rest mode is the operating mode in which the vehicle provides rest services for users.

[0048] Specifically, in this embodiment, the vehicle rest mode adaptive adjustment method is executed by an on-board processor or on-board control system. It is communicatively connected to various sensors, controllers, and execution units within the cabin, enabling it to acquire user data, vehicle status, and environmental information, and after analysis and decision-making, send control commands to the execution mechanism. For ease of description, these will be collectively referred to as the system below.

[0049] When a user triggers the rest mode via voice commands such as "I want to rest for twenty minutes", the rest mode icon on the touch screen, or a specific gesture, the system first needs to determine the identity of the user in the current vehicle.

[0050] The system can capture facial images of the driver or passengers through in-vehicle cameras, extract facial feature points using locally deployed lightweight facial recognition algorithms, and compare them with user registration information pre-stored in the vehicle's storage unit to identify the specific user's identity.

[0051] If the vehicle supports a multi-user account system, the system can also assist in confirming identity through pairing information from the mobile phone Bluetooth key or trigger signals from the seat memory button.

[0052] For unregistered temporary users, the system can assign a generic identity and invoke the factory-preset initial preference model.

[0053] Upon verifying the user's identity, the system initiates the process of collecting multimodal perception data.

[0054] User status perception data includes: limb movement frequency and eye movement status collected by in-vehicle infrared cameras; heart rate and body surface temperature collected by piezoelectric vital signs sensors embedded in the seat; and ambient noise level and user breathing characteristics collected by in-vehicle microphone arrays.

[0055] The environmental perception data is divided into two parts: in-vehicle and out-of-vehicle. The in-vehicle temperature and humidity sensor collects the real-time temperature and relative humidity in the cabin, the light sensor collects the light intensity in the cabin, and the air quality sensor detects the carbon dioxide concentration and volatile organic compound content. The out-of-vehicle environmental perception unit obtains local weather forecast data from the cloud through the vehicle communication module, including outdoor temperature, weather conditions, and air quality index, and performs real-time verification using the out-of-vehicle temperature sensor.

[0056] All sensing data is timestamped and stored in the system's circular buffer for subsequent initialization calculations and real-time monitoring.

[0057] In some embodiments, the system may also include a data preprocessing module for filtering out instantaneous noise from the sensor, such as using median filtering to remove motion artifacts from heart rate data and using moving average to smooth fluctuations in temperature data.

[0058] Thus, the system obtains complete input information for personalized adjustments, including user identity tags, the user's current physiological and behavioral state, and parameters of the vehicle's internal and external environment.

[0059] S202. Generate initial adjustment parameters based on the pre-stored historical preference model corresponding to the user identity, the deviation between the current environmental perception data and the pre-stored historical environmental data, and perform adjustment according to the initial adjustment parameters.

[0060] The adjustment is based on the atomization capability corresponding to the initial adjustment parameters, and the atomization capability is a vehicle control function unit that supports independent invocation.

[0061] Specifically, after obtaining a user's identity, the system retrieves the historical preference model bound to that identity from non-volatile memory. This model stores the user's historical adjustment records under different environmental conditions in the form of key-value pairs or lightweight decision trees. For example, when the ambient temperature is between 26°C and 28°C, the user usually sets the air conditioner to 22°C and turns on the second fan speed, adjusts the seat back angle to 120 degrees, raises the leg rest to 80%, closes the window light transmittance to 10%, and plays white noise.

[0062] If this is the user's first time using the rest mode, the system will call the group default model based on a large amount of anonymous user data as the initial baseline.

[0063] Next, the system needs to calculate the deviation between the current environmental perception data and the pre-stored historical environmental data.

[0064] The system can extract the average environmental data recorded when the user started the rest mode three times in the past three times from the historical preference model as a baseline environmental vector, including average temperature, average light intensity, average noise decibels, etc.

[0065] The deviation of each dimension is obtained by subtracting the currently collected ambient temperature, light intensity, and noise level from the corresponding baseline values.

[0066] For example, if the current outdoor temperature is 35℃, and the user's historical average temperature during rest periods is 28℃, then the temperature deviation is +7℃. In this case, the system corrects the baseline adjustment parameters based on the magnitude and direction of the deviation: for temperature deviation, a linear correction coefficient is used; for every 1℃ increase above the baseline value, the air conditioning target temperature is lowered by 0.5℃, and the fan speed is increased by one level; for lighting deviation, if the current lighting intensity is higher than the historical average, the window transmittance is automatically reduced to a darker level; for noise deviation, if the current external noise is higher than the historical average, the intensity of active noise cancellation is actively increased or the volume of white noise is increased.

[0067] After the correction is completed, the system generates an initial set of adjustment parameters, which contains target values ​​for the atomization capabilities of the seating system, environmental system, acoustic system, and optical system.

[0068] Subsequently, the system distributes various adjustment commands to the corresponding execution units through the central control interface of the atomic capability execution layer.

[0069] Specifically, after receiving parameters such as backrest angle, leg rest position, and massage mode, the seat controller can drive the motor to perform actions at a smooth speed to avoid causing user discomfort due to rapid movement; the air conditioning controller can start the compressor and fan according to the target temperature, air volume, and air outlet mode; the window controller changes the light transmittance by adjusting the voltage of the electrochromic glass; and the audio system starts playing preset sleep-aid tracks.

[0070] All adjustments can be pre-configured to avoid abrupt changes, such as limiting seat movement speed to less than 5 degrees per second and gradually changing the air conditioning temperature at a rate of one degree every 30 seconds.

[0071] In addition, the system can also record the initial adjustment parameters generated this time, which can be used as the basis for subsequent learning and updates.

[0072] S203. During the operation of the rest mode, monitor the change in the preset types of sensory data, and when the change in any sensory data exceeds the preset threshold associated with that sensory data, adjust the parameter of the atomization capability associated with that sensory data until the rest mode ends.

[0073] Specifically, during the rest mode operation, the system continuously collects various sensing data defined in step S201 at fixed time intervals.

[0074] In some embodiments, the sampling frequency varies depending on the data type. For example, vital signs such as heart rate and body temperature are collected once per second, ambient temperature and humidity are collected once every 10 seconds, noise levels are continuously monitored and the average decibel value is calculated every 2 seconds, and light intensity is collected once every 5 seconds. Other periodic intervals may also be used depending on the actual situation.

[0075] The system presets independent trigger thresholds for each type of sensing data. These thresholds can be either factory-calibrated values ​​or dynamically adapted based on the user's historical adjustment behavior.

[0076] For example, the threshold for heart rate change is a change of more than 5 beats per minute, the threshold for body temperature change is 0.3℃, the threshold for ambient temperature change is 1℃, and the threshold for noise level change is 5 decibels.

[0077] The system can compare each newly collected sensing data value with the previous adjusted baseline value and calculate the absolute value of the change.

[0078] When the change in a certain piece of perceived data exceeds its corresponding trigger threshold for the first time, the system immediately triggers an adjustment action for the atomization capability associated with that data.

[0079] The correlation mapping is predefined: heart rate changes are mainly related to air conditioning temperature, air volume, and audio system volume, because an increased heart rate often means that the user feels too hot or is disturbed by external factors; body temperature changes are also related to the air conditioning system; ambient temperature changes are related to the air conditioning target temperature and air volume; noise level changes are related to the intensity of active noise cancellation and the volume of white noise; light intensity changes are related to the light transmittance of the car windows and the brightness of the interior ambient lighting.

[0080] It is important to emphasize that the system only adjusts the atomization capabilities directly related to the over-limit perception data, without changing the status of other operating capabilities. For example, when the ambient noise increases from 40 decibels to 48 decibels, exceeding the 5-decibel threshold, the system only increases the active noise cancellation intensity by one level and raises the white noise volume from 20% to 30%, while other parameters such as seat angle, air conditioning temperature, and window light transmittance remain unchanged.

[0081] This differentiated and independent adjustment mechanism avoids the disruption to users' rest caused by global parameter reset.

[0082] After the adjustment is executed, the system uses the new parameter values ​​as the benchmark for subsequent comparisons and resets the cumulative change of the sensed data.

[0083] In some embodiments, if multiple sensing data exceed the limit simultaneously or sequentially, the system processes them in order of priority. The priority rule can be: vital signs data related to user safety take precedence over environmental comfort data, such as abnormal heart rate taking precedence over temperature changes.

[0084] In some embodiments, the system may also set a minimum adjustment interval, such as adjusting the same atomization capability at most once within 30 seconds, to prevent frequent jitter.

[0085] The above monitoring and adjustment process is executed cyclically until the system receives an exit signal, such as the user's voice command "exit rest mode", the preset rest time expires, or the vehicle gear changes.

[0086] Thus, the system achieves real-time and refined responses to changes in user status and environment, continuously maintaining the optimal rest environment while ensuring the continuity of rest.

[0087] The vehicle rest mode adaptive adjustment method provided in this application determines the user's identity and obtains user status and environmental perception data. Then, based on the user's historical preference model and the deviation between the current environment and historical environments, initial adjustment parameters are generated, and independently controllable atomization capabilities are invoked to execute the adjustment. During the rest period, the changes in various perception data are continuously monitored. When the change in any data exceeds its preset associated threshold, only the atomization capability associated with that data is independently adjusted until the rest ends. Thus, by combining historical preferences with real-time environmental deviations to generate initial parameters, the problem of traditional solutions being unable to personalize the user's current state is solved. Furthermore, by setting independent thresholds for different perception data and adjusting only the associated atomization capabilities, the problem of prior art disrupting rest continuity due to global parameter adjustments is solved. This achieves real-time, refined response to changes in user status and environment, significantly improving the accuracy and personalization level of adjustment while ensuring the continuity of the rest process.

[0088] Figure 3 Another embodiment of the vehicle rest mode adaptive adjustment method provided in this disclosure, in Figure 2 Based on the illustrated embodiment, the following is combined with Figure 3 The specific steps in implementing the adaptive adjustment method for vehicle rest mode are described in detail, including the following steps:

[0089] S301. Determine the identity of the user in the vehicle when the rest mode is activated.

[0090] Specifically, the system performs a startup condition check before entering rest mode.

[0091] Specifically, the system detects whether the vehicle speed is zero, whether the gear is in parking (P) gear, whether the brake pedal is not depressed, and whether the remaining battery power meets the minimum threshold for rest mode operation, such as being greater than 10%.

[0092] If any condition is not met, the system will prompt the user via voice or screen that they cannot enter rest mode and provide corresponding operation suggestions.

[0093] This ensures the vehicle is safely parked and prevents accidental activation of the rest mode while driving.

[0094] S302. Acquire user identity, body movements and eye movement data through in-vehicle cameras, user heart rate and body temperature data through vital sign sensors, and user voice command data through microphones, and integrate them to obtain user status perception data.

[0095] Specifically, the system can use an in-vehicle infrared camera to collect the user's facial expressions, head posture, limb joint positions, eye closure frequency, and pupil changes to determine the user's level of drowsiness and sleep depth.

[0096] Vital signs sensors are typically integrated into the seat cushion and backrest, using piezoelectric thin film or capacitive sensing technology to collect heart rate and body surface temperature without contact. The sampling frequency can be set to once per second.

[0097] The microphone array is used not only to receive voice commands, but also to continuously collect ambient noise inside the vehicle as well as the frequency and amplitude of the user's breathing sounds.

[0098] The system aligns the aforementioned multi-source data by timestamp to form a user status awareness data vector for subsequent decision-making.

[0099] In some embodiments, the system also performs anomaly filtering on sensor data during the initialization phase.

[0100] Specifically, the system performs a reasonableness check on the raw data collected by each sensing module. For example, if the heart rate data exceeds the normal range of 30-150 beats / minute, the ambient temperature exceeds the reasonable range of -20℃ to 60℃, or a certain sensor has no signal output for a long time, the system determines that the data is an abnormal value, replaces it with the valid value of the previous moment or the average value of adjacent sensors, and records the abnormal event in the log.

[0101] This avoids decision-making errors caused by sensor malfunctions or momentary interference.

[0102] S303: Acquire in-vehicle temperature, humidity, light intensity, and noise data through in-vehicle environmental sensors, and acquire outside air quality, ambient temperature, and weather data through the outside environment sensing unit, and integrate them to obtain environmental sensing data.

[0103] Specifically, in-vehicle environmental sensors include, but are not limited to: temperature and humidity sensors installed in the roof or near the air conditioning vents; light sensors typically located below the windshield in front of the dashboard; and noise sensors utilizing existing microphone arrays.

[0104] The vehicle's external environment perception unit measures the outside temperature in real time through the vehicle's external temperature sensor, and obtains weather forecast data from the cloud through the vehicle's T-BOX module, including weather conditions such as sunshine and rain, air quality index (AQI), and wind speed.

[0105] The system integrates in-vehicle and out-of-vehicle data into an environmental perception data vector for subsequent deviation calculation.

[0106] S304. Call the historical preference model and obtain the baseline adjustment parameters corresponding to the current user identity.

[0107] Specifically, the system can read historical preference model files associated with user identities from non-volatile storage.

[0108] In some embodiments, the model may be stored in a lightweight JSON format, containing combinations of user preference parameters across multiple environmental condition ranges.

[0109] If the current environmental conditions fall within a certain range, the parameters corresponding to that range will be directly output as the baseline adjustment parameters; if they fall outside the range, nearest neighbor interpolation or linear interpolation will be used for calculation.

[0110] For new users, the model can return the factory-preset universal baseline parameters.

[0111] S305. Calculate the deviation between the current environmental perception data and the average environmental data stored in the user's historical rest scenarios.

[0112] Specifically, the system extracts the weighted average of the environmental data recorded when the user recently started the rest mode from the user's historical preference model, which is called the historical average environmental data.

[0113] For example, calculate the arithmetic mean of the ambient temperature, light intensity, and noise level at the three most recent rest periods. Subtract the current environmental data from the corresponding historical averages to obtain the deviation for each dimension.

[0114] In some implementations, if data for a certain dimension is missing, the default deviation is zero.

[0115] S306. Calculate the environmental correction factor based on the deviation.

[0116] Specifically, the system calculates the corresponding correction coefficients based on the deviation of each dimension using a preset correction function.

[0117] The correction function can be a linear mapping, such as increasing the correction coefficient by 0.05 for every 1°C increase in temperature deviation; or it can be a nonlinear piecewise function, where the correction coefficient is small when the deviation is small, and increases rapidly after the deviation exceeds a certain threshold. The correction coefficient is a dimensionless scalar, and its value can range from 0.5 to 1.5.

[0118] The system can set different correction functions for different atomization capabilities. For example, the correction coefficient for air conditioning temperature is strongly correlated with temperature deviation, while the correction coefficient for seat angle is not sensitive to environmental deviation.

[0119] In some embodiments, to improve processing efficiency, the system may use a pre-calibrated mapping table for lookup calculations.

[0120] S307. Based on the environmental correction factor, dynamically correct the benchmark adjustment parameters and generate the initial adjustment parameters.

[0121] Specifically, the system multiplies or weights the baseline adjustment parameters obtained in step S304 with the correction coefficients for each dimension obtained in step S306 to obtain the corrected adjustment parameters.

[0122] For example, if the base air conditioner target temperature is 22℃ and the temperature deviation correction factor is 1.1, then the corrected target temperature is 24.2℃, and the system rounds it to 24℃.

[0123] For non-continuously adjustable parameters such as seat massage modes, the correction factor only affects whether the mode is enabled and does not change the mode type.

[0124] The system summarizes all the revised atomization capability target values ​​into an initial set of adjustment parameters.

[0125] S308. Perform adjustment according to the initial adjustment parameters.

[0126] Specifically, the system distributes the initial set of adjustment parameters to each atomized capability execution unit via the vehicle bus. The seat controller, air conditioning controller, window controller, audio system, etc., each execute actions according to the target values.

[0127] During execution, the system records the deviation between the actual execution result and the target value for subsequent learning.

[0128] In some embodiments, the adjustment according to the initial adjustment parameters is implemented based on a reinforcement learning decision model. The reinforcement learning decision model takes user identity, user state perception data, and environment perception data as states, uses the invocation of atomic capabilities and parameter adjustment as actions, and aims to maximize the composite reward function. The composite reward function includes immediate reward and final reward. The immediate reward is determined based on whether the user manually adjusts the parameters or provides real-time feedback, and the final reward is determined based on whether the rest duration meets the recommended range and whether the user's deep rest duration meets the standard.

[0129] Specifically, the decision-making model adopts the classic reinforcement learning framework, in which the agent is an AI agent specifically for the rest mode, and the environment is the vehicle cabin and the user's state.

[0130] At each decision-making moment, the agent observes the current state S. The state space includes user identity, user state perception data such as heart rate, body temperature, and limb movement frequency, as well as environmental perception data such as temperature, light intensity, and noise. The agent selects an action A according to the current policy π. The action space corresponds to the invocation of atomic capabilities and parameter adjustments, such as lowering the air conditioning temperature by 1°C, increasing the seat back angle by 5°, and increasing the intensity of active noise cancellation.

[0131] After the action is performed, the environment transitions to the next state S′, and the agent receives a reward value R.

[0132] The reward function R is designed as a composite function, consisting of an immediate reward and a final reward.

[0133] Immediate rewards are given immediately after each action: if the user does not manually intervene in the action and does not give negative feedback, a positive reward such as +1 is given; if the user manually adjusts the parameter or explicitly expresses dissatisfaction, a negative reward such as -1 is given.

[0134] The final reward is calculated after the rest mode has completely ended: if the user's actual rest time is within the recommended range, such as 15-30 minutes, and the deep rest time accounts for more than 50%, a larger positive reward, such as +10, will be given.

[0135] The goal of an intelligent agent is to learn an optimal strategy that maximizes cumulative discount rewards through continuous interaction with the environment.

[0136] During training, each time the system completes a rest mode, it stores the current interaction trajectory state, action, reward, and next state into the experience replay pool. Periodically, it randomly samples small batches of data from the pool and updates the parameters of the neural network model using Q-learning or policy gradient methods.

[0137] Through multiple iterations, the agent gradually learns to select the optimal sequence of atomic capability calls for different user identities and environmental conditions, thereby achieving personalized adaptive adjustment.

[0138] In some embodiments, the reinforcement learning model can run on an automotive embedded platform using an open-source framework such as TensorFlow Lite Micro, without relying on commercial software.

[0139] S309. During the operation of the rest mode, monitor the changes in the sensing data of preset types.

[0140] Specifically, the system can continuously collect all the sensing data defined in steps S302 and S303 at fixed intervals, and maintain a "previous adjusted baseline value" for each type of data.

[0141] Each newly collected data point is compared with the corresponding baseline value, and the absolute value of the change is calculated.

[0142] In some embodiments, the monitoring frequency is dynamically adjusted according to different data types; relevant details can be found in [reference needed]. Figure 2 The descriptions of the corresponding steps in the illustrated embodiments will not be repeated here.

[0143] S310. Compare the real-time changes of various types of sensing data with the corresponding trigger thresholds.

[0144] There is a pre-defined mapping relationship between the perceived data and the corresponding atomization capabilities.

[0145] Specifically, the system presets an independent trigger threshold for each type of sensing data. The threshold can be dynamically updated based on the standard deviation of the user's historical adjustment data, or the factory calibration value can be used.

[0146] Meanwhile, the system can maintain a mapping table that defines which atomic capabilities should be adjusted when each type of perceived data exceeds its limits. For example, exceeding the heart rate limit is mapped to air conditioning temperature and volume, exceeding the ambient temperature limit is mapped to air conditioning target temperature and airflow, and exceeding the light limit is mapped to window transmittance and ambient light brightness.

[0147] Depending on the actual situation, the mapping relationship can be one-to-many or many-to-one, but the adjustment action only applies to directly related capabilities.

[0148] S311. When the change in a certain type of sensing data exceeds the corresponding trigger threshold, only the atomization capability that is pre-mapped to that type of sensing data is triggered to make independent parameter adjustments, without changing the current state of other atomization capabilities.

[0149] Specifically, when the comparison result of step S310 is that a certain sensing data exceeds the limit, the system only issues adjustment instructions to the atomization capabilities associated with the sensing data in the mapping table, while other atomization capabilities remain unchanged.

[0150] For example, when the ambient temperature rises above the 1°C threshold, the system only adjusts the target temperature and airflow of the air conditioner, while the seat angle, window light transmittance, etc., remain unchanged.

[0151] If multiple sensing data exceed the limit simultaneously, the system processes them sequentially according to the preset priority order, with higher priority data being executed first. Furthermore, the same atomic capability will only respond once within a short period of time, such as 30 seconds, to avoid oscillation.

[0152] In some embodiments, the system also provides a rest depth assessment function.

[0153] Specifically, the system uses a lightweight classification model to assess the user's rest depth based on the user's heart rate variability, respiratory rate stability, limb movement frequency, and eye movement status collected during the rest process, classifying the rest state into three levels: awake, light sleep, and deep sleep.

[0154] The evaluation results are displayed in real time in the corner of the central control screen or broadcast to users for reference, and also serve as the quantitative basis for "deep rest duration" in the final reward function.

[0155] The system can also automatically reduce screen brightness to the minimum, turn off ambient lighting, and further reduce media volume after detecting that the user has entered a deep sleep state, in order to create a more suitable sleep environment.

[0156] S312. If an exit command is received or the preset exit conditions are detected, the corresponding atomization capability is invoked to restore the seat posture, air conditioning settings, window light transmittance, and audio status to the state before entering rest mode.

[0157] Specifically, the system continuously monitors exit conditions, and the priority of exit signals is as follows: emergency situations such as vehicle collision signals or gear changes are the highest, followed by user-initiated exit commands such as voice "exit rest mode" or touch operation, and finally the expiration of the preset rest time.

[0158] When any condition is met, the system reads snapshots of the system states saved before the start of the rest mode from the memory, generates a set of recovery instructions, and restores the seats, air conditioning, windows, audio system, etc. to their state before entering the rest mode through the atomic capability execution layer.

[0159] The recovery process also uses a smooth, gradual approach to avoid sudden changes that could cause user discomfort.

[0160] S313. Based on the in-vehicle interactive interface, collect user satisfaction feedback on this rest period, and use this satisfaction feedback as input data for the next incremental learning of the historical preference model.

[0161] Specifically, after exiting the rest mode, the system will proactively ask the user about their satisfaction with the rest on the central control screen or voice assistant, for example, "Please rate this rest experience: satisfied, average, dissatisfied", and can further inquire about specific aspects of dissatisfaction, such as "the air conditioning is a bit cold" or "the seat is not reclined enough".

[0162] User feedback was quantified into satisfaction scores and specific adjustment suggestions, which were stored as tag data in the entire data of this rest period for subsequent incremental learning by the system.

[0163] S314. Record all data for this rest period.

[0164] The data for the entire rest period includes user identification, initial adjustment parameters, dynamic adjustment records, user active intervention records, vital sign change curves, and user satisfaction feedback.

[0165] Specifically, after the rest mode is completely exited, the system will package and store all the data generated during this rest period.

[0166] The initial adjustment parameters are the parameter set generated in step S307; the dynamic adjustment record includes the time of each threshold trigger, the trigger reason, the adjusted parameters and the adjustment amount; the user active intervention record includes the operation log of the user manually adjusting any atomization ability through voice or touch during the rest period; the vital sign change curve is the sequence data of heart rate, body temperature and other changes over time; and the user satisfaction feedback comes from step S313.

[0167] All data is stored in a structured format in the user's private database.

[0168] S315. Update the user's historical preference model through incremental learning.

[0169] Incremental learning updates include using the entire data from this rest period as new training samples to progressively correct the parameters of the historical preference model, allowing the model to evolve smoothly as the number of uses increases.

[0170] Specifically, the system uses an incremental learning algorithm to update the historical preference model, rather than retraining the entire dataset each time.

[0171] Specifically, if the historical preference model is based on linear regression or a lightweight decision tree, the entire data of this rest period is used as the new sample. The gradient update of the model parameters is calculated and weighted with a small learning rate, such as 0.1, so that the model parameters slowly drift towards the new sample.

[0172] If the historical preference model is a statistical model, such as the average preference under each environmental interval, then the sample mean and sample variance of that interval are updated directly.

[0173] By employing incremental learning, the model is guaranteed not to change drastically due to a single instance of abnormal data, while also being able to track the long-term migration of user preferences.

[0174] In some embodiments, the system may be implemented using the incremental learning module in the open-source scikit-learn library.

[0175] The updated model takes effect immediately and is used for initial adjustments during the next rest mode.

[0176] In some embodiments, the system supports uninterrupted fine-tuning by the user during rest periods. When the user issues an adjustment command via voice such as "turn the air conditioner up" or gesture such as waving upwards, the system parses the target capability in the command (air conditioning) and the adjustment direction (increasing the temperature), and adjusts only that atomized capability independently without interrupting the current rest mode or triggering any changes to other capabilities.

[0177] The system can simultaneously record the type, time, and amount of user intervention, which can be used as negative reward samples for subsequent incremental learning to indicate that the initial parameters or dynamic adjustment strategy failed to fully meet user preferences.

[0178] In some embodiments, the system supports the reservation and preheating functions for rest mode.

[0179] For example, users can set a scheduled start time via a mobile app or in-car voice command, such as "Start rest mode for 20 minutes at 12:15 PM." The system begins preheating 5 minutes before the scheduled time: based on the current ambient temperature and the user's historical preferences, it pre-activates the air conditioning to adjust the cabin temperature to a comfortable range, pre-adjusts the seat to a resting position, and pre-loads sleep-inducing audio that the user frequently listens to. Therefore, when the scheduled time arrives, the system directly enters the prepared resting environment, eliminating the need for the user to wait for the adjustment process to complete, further enhancing the user experience.

[0180] In some embodiments, the system can also be configured to have group learning and model sharing capabilities. With user authorization and data anonymization, the system uploads the user's preferred model parameters to a cloud server.

[0181] Therefore, the server performs cluster analysis on the preference data of a large number of anonymous users to generate default model parameters for different group profiles, such as "people who prefer cool environments" and "people who prefer high seat backs".

[0182] When a new user uses the rest mode for the first time, the system can download the default model of the most similar group as the initial preference, which significantly reduces the number of personalized adaptations during the cold start phase.

[0183] It is important to emphasize that the learning process for this group employs federated learning or differential privacy technology to ensure that user privacy is not compromised.

[0184] The adaptive adjustment method for vehicle rest modes provided in this disclosure generates initial parameters through identity recognition, multimodal perception, historical preferences, and environmental deviation correction, and optimizes decision-making using a reinforcement learning composite reward function. During operation, it only adjusts the associated atomic capabilities based on differential thresholds. After exiting, it records all data and incrementally updates the model through incremental learning. This effectively solves the problems of existing technologies failing to respond in real time to individual state and environmental changes, and abruptly interrupting the rest continuity with abrupt adjustments. It also overcomes the deficiency of models failing to evolve with use, achieving a precise, coherent, and continuously optimized adaptive rest adjustment experience.

[0185] Figure 4 This is a schematic diagram of the structure of a vehicle rest mode adaptive adjustment device provided in one embodiment of this disclosure. Figure 4 As shown, the vehicle rest mode adaptive adjustment device 400 includes:

[0186] The acquisition module 410 is used to determine the identity of the user in the vehicle when the rest mode is started, and to acquire user status perception data and environmental perception data. The rest mode is the operating mode in which the vehicle provides rest services to the user.

[0187] The initial adjustment module 420 is used to generate initial adjustment parameters based on the pre-stored historical preference model corresponding to the user identity, the deviation between the current environmental perception data and the pre-stored historical environmental data, and to perform adjustment according to the initial adjustment parameters. The adjustment is based on the atomization capability corresponding to the initial adjustment parameters, and the atomization capability is a vehicle control function unit that supports independent invocation.

[0188] The dynamic processing module 430 is used to monitor the changes in preset types of sensor data during the operation of the rest mode, and when the changes in any sensor data exceed the preset threshold associated with the sensor data, adjust the parameters of the atomization capability associated with the sensor data until the rest mode ends.

[0189] Optionally, the initial adjustment module 420 specifically includes: performing adjustment based on a reinforcement learning decision model according to the initial adjustment parameters. The reinforcement learning decision model takes user identity, user state perception data, and environmental perception data as states, uses the invocation of atomic capabilities and parameter adjustment as actions, and aims to maximize the composite reward function. The composite reward function includes immediate reward and final reward. The immediate reward is determined based on whether the user manually adjusts the parameters or provides real-time feedback, and the final reward is determined based on whether the rest duration meets the recommended range and whether the user's deep rest duration meets the standard.

[0190] Optionally, the dynamic processing module 430 is specifically used to compare the real-time changes of various types of sensing data with the corresponding trigger thresholds, wherein there is a preset mapping relationship between the sensing data and the corresponding atomization capabilities; when the change of a certain type of sensing data exceeds the corresponding trigger threshold, only the atomization capability that is presetly mapped to that type of sensing data is triggered to make independent parameter adjustments, without changing the current state of other atomization capabilities.

[0191] Optionally, the initial adjustment module 420 is further configured to record the entire rest data after the rest mode ends, wherein the entire rest data includes user identification, initial adjustment parameters, dynamic adjustment records, user active intervention records, vital sign change curves, and user satisfaction feedback as the entire rest data; and update the user's historical preference model through incremental learning, wherein incremental learning includes using the entire rest data as new training samples to progressively correct the parameters of the historical preference model, so that the model evolves smoothly with the increase of usage.

[0192] Optionally, the initial adjustment module 420 is specifically used to: call the historical preference model to obtain the baseline adjustment parameters corresponding to the current user identity; calculate the deviation between the current environmental perception data and the average environmental data stored in the user's historical rest scenarios; calculate the environmental correction coefficient based on the deviation; dynamically correct the baseline adjustment parameters based on the environmental correction coefficient, and generate the initial adjustment parameters.

[0193] Optionally, the acquisition module 410 is specifically used to acquire user identity, body movements and eye movements data collected by the in-vehicle camera, user heart rate and body temperature data collected by the vital signs sensor, user voice command data collected by the microphone, and integrate them to obtain user status perception data; acquire in-vehicle temperature and humidity, light intensity and noise data collected by the in-vehicle environment sensor, and in-vehicle air quality, ambient temperature and weather data collected by the in-vehicle environment perception unit, and integrate them to obtain environmental perception data.

[0194] Optionally, the dynamic processing module 430 is also used to, after the rest mode ends, receive an exit command or detect that the preset exit conditions are met, call the corresponding atomization capability to restore the seat posture, air conditioning settings, window light transmittance, and audio status to the state before entering the rest mode; collect user satisfaction feedback on this rest based on the in-vehicle interactive interface, and use the satisfaction feedback as input data for the next incremental learning of the historical preference model.

[0195] In this embodiment, the vehicle rest mode adaptive adjustment device, through the combination of various modules, solves the problem in related technologies that the vehicle rest mode adjustment function cannot take into account changes in user status and environmental changes, resulting in insufficient user experience.

[0196] Figure 5 This is a schematic diagram of the structure of a control device provided in one embodiment of the present disclosure, as shown below. Figure 5 As shown, the control device 500 includes a memory 510 and a processor 520.

[0197] The memory 510 stores a computer program that can be executed by at least one processor 520. This computer program is executed by at least one processor 520 to enable the control device to implement the vehicle rest mode adaptive adjustment method provided in any of the above embodiments.

[0198] The memory 510 and the processor 520 can be connected via a bus 530.

[0199] The relevant explanations can be understood by referring to the corresponding descriptions and effects in the method embodiments, and will not be repeated here.

[0200] One embodiment of this disclosure provides a computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement the vehicle rest mode adaptive adjustment method provided in any of the above embodiments.

[0201] The computer-readable storage medium may be ROM, random access memory (RAM), CD-ROM, magnetic tape, floppy disk, and optical data storage device, etc.

[0202] One embodiment of this disclosure provides a computer program product comprising computer-executable instructions that, when executed by a processor, are used to implement the vehicle rest mode adaptive adjustment method provided in any of the above embodiments.

[0203] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by a computer program instructing related hardware. The program can be stored in a computer-readable storage medium, and when executed, it can include the processes of the embodiments of the methods described above. The storage medium can be a magnetic disk, optical disk, read-only memory (ROM), or random access memory (RAM), etc.

[0204] In the above embodiments, the descriptions of each embodiment have their own emphasis. For parts not described in detail in a certain embodiment, please refer to the relevant descriptions of other embodiments. The technical features of the above embodiments can be combined arbitrarily. For the sake of brevity, not all possible combinations of the technical features in the above embodiments are described. However, as long as the combination of these technical features does not contradict each other, it should be considered within the scope of this specification.

[0205] Other embodiments of this application will readily occur to those skilled in the art upon consideration of the specification and practice of the invention disclosed herein. This application is intended to cover any variations, uses, or adaptations of this application that follow the general principles of this application and include common knowledge or customary techniques in the art not disclosed herein. The specification and examples are to be considered exemplary only, and the true scope and spirit of this application are indicated by the following claims.

[0206] It should be understood that this application is not limited to the precise structure described above and shown in the accompanying drawings, and various modifications and changes can be made without departing from its scope. The scope of this application is limited only by the appended claims.

Claims

1. A method for adaptive adjustment of vehicle rest mode, characterized in that, The method includes the following steps: The system determines the identity of the user in the vehicle when the rest mode is activated, and acquires user status perception data and environmental perception data, wherein the rest mode is the operating mode in which the vehicle provides rest services to the user. Initial adjustment parameters are generated based on the pre-stored historical preference model corresponding to the user identity, the deviation between the current environmental perception data and the pre-stored historical environmental data, and the adjustment is performed according to the initial adjustment parameters. The adjustment is based on the atomization capability corresponding to the initial adjustment parameters, and the atomization capability is a vehicle control function unit that supports independent invocation. During the rest mode operation, the changes in the preset types of sensor data are monitored, and when the change in any sensor data exceeds the preset threshold associated with that sensor data, the parameters of the atomization capability associated with that sensor data are adjusted until the rest mode ends.

2. The method according to claim 1, characterized in that, The adjustment according to the initial adjustment parameters is implemented based on a reinforcement learning decision model. The reinforcement learning decision model takes user identity, user state perception data, and environment perception data as states, uses the invocation of atomic capabilities and parameter adjustment as actions, and aims to maximize the composite reward function. The composite reward function includes an immediate reward and a final reward. The immediate reward is determined based on whether the user manually adjusts the parameters or provides real-time feedback, while the final reward is determined based on whether the rest duration meets the recommended range and whether the user's deep rest duration meets the standard.

3. The method according to claim 1, characterized in that, When the change in any sensed data exceeds a preset threshold associated with that sensed data, the parameter of the atomization capability preset associated with that sensed data is adjusted, including: The real-time changes of various types of sensing data are compared with the corresponding trigger thresholds, wherein there is a preset mapping relationship between the sensing data and the corresponding atomization capability; When the change in a certain type of sensing data exceeds the corresponding trigger threshold, only the atomization capability that is pre-mapped to that type of sensing data is triggered to make independent parameter adjustments, without changing the current state of other atomization capabilities.

4. The method according to claim 1, characterized in that, After the rest mode ends, it also includes: Record all data during this rest period, including user identification, initial adjustment parameters, dynamic adjustment records, user active intervention records, vital sign change curves, and user satisfaction feedback. The user's historical preference model is updated through incremental learning, wherein the incremental learning update includes using the entire data of this rest as new training samples to progressively correct the parameters of the historical preference model, so that the model evolves smoothly as the number of uses increases.

5. The method according to any one of claims 1 to 4, characterized in that, The process of generating initial adjustment parameters based on the pre-stored historical preference model corresponding to the user's identity, the deviation between the current environmental perception data and the pre-stored historical environmental data, includes: Invoke the historical preference model and obtain the baseline adjustment parameters corresponding to the current user identity; Calculate the deviation between the current environmental perception data and the average environmental data stored in the user's historical rest scenarios; Calculate the environmental correction factor based on the deviation amount; Based on the environmental correction coefficient, the baseline adjustment parameters are dynamically corrected, and initial adjustment parameters are generated.

6. The method according to any one of claims 1 to 4, characterized in that, The acquisition of user state awareness data and environment awareness data includes: The system acquires user identity, body movements, and eye movement data through in-vehicle cameras, heart rate and body temperature data through vital sign sensors, and voice command data through microphones, and integrates these data to obtain user status perception data. The system acquires in-vehicle temperature, humidity, light intensity, and noise data through in-vehicle environmental sensors, and outside air quality, ambient temperature, and weather data through an external environmental sensing unit, and integrates these data to obtain environmental sensing data.

7. The method according to any one of claims 1 to 4, characterized in that, After the rest mode ends, it also includes: If an exit command is received or the preset exit conditions are met, the corresponding atomization capability is invoked to restore the seat posture, air conditioning settings, window light transmittance, and audio status to the state before entering rest mode. Based on the in-vehicle interactive interface, user satisfaction feedback on this rest period is collected, and this satisfaction feedback is used as input data for the next incremental learning of the historical preference model.

8. A vehicle rest mode adaptive adjustment device, characterized in that, include: The acquisition module is used to determine the identity of the user in the vehicle when the rest mode is activated, and to acquire user status perception data and environmental perception data, wherein the rest mode is the operating mode in which the vehicle provides rest services to the user; The initial adjustment module is used to generate initial adjustment parameters based on the pre-stored historical preference model corresponding to the user identity, the deviation between the current environmental perception data and the pre-stored historical environmental data, and to perform adjustment according to the initial adjustment parameters. The adjustment is based on the atomization capability corresponding to the initial adjustment parameters, and the atomization capability is a vehicle control function unit that supports independent invocation. The dynamic processing module is used to monitor the changes in preset types of sensor data during the rest mode operation, and when the change in any sensor data exceeds the preset threshold associated with that sensor data, it adjusts the parameters of the atomization capability associated with that sensor data until the rest mode ends.

9. A control device, characterized in that, include: Memory, processor; The memory stores computer-executed instructions; The processor executes computer execution instructions stored in the memory, causing the processor to perform the method as described in any one of claims 1 to 7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer-executable instructions, which, when executed by a processor, are used to implement the method as described in any one of claims 1 to 7.

Images