Vehicle grille lamp control method, device and equipment

Abstract

The application relates to the technical field of automobile lighting, in particular to a vehicle grille lamp control method, device and equipment. A vehicle grille lamp control method comprises the following steps: collecting driving scene information of a current driving scene through a multi-modal sensor, and determining a structured feature vector according to the driving scene information; determining a corresponding three-dimensional weight under the current driving scene according to the structured feature vector, wherein the three-dimensional weight comprises a safety weight, an energy efficiency weight and an interaction weight; determining an utility score of each preset grille lamp mode under the current driving scene according to the three-dimensional weight and the structured feature vector; determining a grille lamp lighting strategy according to a preset priority rule and the utility score of each grille lamp mode under the current driving scene, and driving the grille lamp to execute the corresponding lighting strategy.

Description

Technical Field

[0001] This application relates to the field of automotive lighting technology, and in particular to a method, device and equipment for controlling vehicle grille lights. Background Technology

[0002] Grille lights are a type of lighting technology used in the front grille of automobiles. Traditional automotive grille lights have a single function, typically serving only as simple decorative lighting, with simple control logic, such as turning on in conjunction with the headlights. Some related technologies attempt to enhance the functionality of grille lights, such as adjusting brightness based on vehicle speed or allowing for customized patterns via a mobile application. However, these solutions suffer from the following inherent drawbacks:

[0003] Lack of environmental perception: It cannot adaptively adjust to complex traffic environments, such as rainy or snowy weather, oncoming vehicles, and pedestrian distances, which may cause glare or insufficient lighting problems.

[0004] Isolated control logic: Control strategies are usually based on a single or a few parameters, lacking multi-dimensional collaborative consideration of the overall vehicle status, such as energy consumption, battery status and external environment.

[0005] Weak interactive capabilities: It cannot achieve efficient and safe external information interaction, such as conveying clear vehicle intentions to pedestrians.

[0006] Lack of intelligent decision-making core: Existing solutions are mostly simple matching rules based on preset conditions, which do not have the ability to handle uncertainty and conflicting objectives. Summary of the Invention

[0007] This invention provides a vehicle grille light control method, apparatus, and device to address the problems of lack of environmental perception capability, isolated control logic, and weak interaction capability of vehicle grille lights in the prior art.

[0008] In a first aspect, embodiments of the present invention provide a vehicle grille light control method, the method comprising:

[0009] The driving scene information of the current driving scenario is collected by a multimodal sensor, and a structured feature vector is determined based on the driving scene information;

[0010] Based on the structured feature vector, the three-dimensional weights corresponding to the current driving scenario are determined respectively. The three-dimensional weights include safety weight, energy efficiency weight and interaction weight.

[0011] Based on the three-dimensional weights and the structured feature vectors, the utility scores of each preset grille light mode in the current driving scenario are determined respectively.

[0012] Based on preset priority rules and the utility scores of each grille light mode in the current driving scenario, a grille light illumination strategy is determined, and the grille lights are driven to execute the corresponding illumination strategy. The grille light illumination strategy includes target illumination mode, target brightness, target pattern parameters, and target color temperature.

[0013] Optionally, the multimodal sensor includes a vision unit, a radar unit, an environmental sensing unit, a vehicle sensor unit, and a positioning and communication unit;

[0014] The structured feature vectors determined based on the driving scenario information include environment vectors, vehicle state vectors, and user intent vectors.

[0015] The environment vector includes at least one of the following: ambient illuminance, correlated color temperature, visibility index, weather code, and a dynamic obstacle list. The dynamic obstacle list includes obstacle information for multiple obstacles, and the obstacle information includes obstacle type, distance between the vehicle and the obstacle, relative speed between the vehicle and the obstacle, and relative angle between the vehicle and the obstacle.

[0016] The vehicle state vector includes at least one of the following: vehicle speed, battery state of charge, instantaneous power consumption, driving mode, and autonomous driving level.

[0017] The user intent vector includes at least one of the following: turn signal status, brake pedal opening, accelerator pedal opening, and user personalized preference encoding.

[0018] Optionally, determining the three-dimensional weights corresponding to the current driving scenario based on the structured feature vector includes:

[0019] Determine the safety weights under the current driving scenario information, including:

[0020] Based on the dynamic obstacle list, the obstacle information of each obstacle is used to determine the real-time collision risk between the vehicle and each obstacle.

[0021] The obstacle with the highest real-time collision risk is identified as the target collision risk value.

[0022] The target collision risk value is normalized by a preset collision risk threshold to obtain the safety weight.

[0023] Determine the energy efficiency weights under the current driving scenario information, including:

[0024] The battery state of charge is normalized by a preset power threshold to obtain a power sensitivity value.

[0025] The energy efficiency weight is determined based on the safety weight and the power sensitivity value;

[0026] The safety weight is negatively correlated with the energy efficiency weight, while the power sensitivity value is positively correlated with the energy efficiency weight.

[0027] Determine the interaction weights under the current driving scenario information, including:

[0028] The interaction weight is determined based on the security weight and the energy efficiency weight;

[0029] The interaction weight is negatively correlated with the sum of the security weight and the energy efficiency weight.

[0030] Optionally, determining the utility score of each preset grille light pattern in the current driving scenario based on the three-dimensional weights and the structured feature vectors includes:

[0031] Based on the structured feature vectors, the sub-utility scores of each grille light mode in the current driving scenario are determined in the safety, energy efficiency, and interaction dimensions, respectively.

[0032] The comprehensive utility score is obtained by weighting and summing the sub-utility scores of the three dimensions using the aforementioned three-dimensional weights.

[0033] Specifically, determining the sub-utility score of the grille light in the safety dimension under the current driving scenario based on the structured feature vector includes:

[0034] The effective lighting area is determined based on the total number of grille light sectors, the effective lighting area of ​​each sector, and the number of sectors required to be lit for the corresponding grille light mode.

[0035] The visual gain is determined based on the ambient illuminance in the environment vector.

[0036] The glare risk probability is determined based on the relative distance between the vehicle and the obstacle in the environmental vector, the obstacle type, and the preset distance threshold between the vehicle and the obstacle.

[0037] The sub-utility score of the safety dimension is determined based on the effective lighting area, the visual gain, the preset glare area, and the glare risk probability.

[0038] The sub-utility score of the grille light in the energy efficiency dimension under the current driving scenario is determined based on the structured feature vector, including:

[0039] The energy efficiency value is determined based on the estimated power consumption of the corresponding grille light mode and the maximum theoretical power consumption.

[0040] The first modulation factor is determined based on the battery state of charge in the vehicle state vector.

[0041] The energy efficiency value is modulated by the first modulation factor to obtain the sub-utility score of the energy efficiency dimension;

[0042] The sub-utility score of the grille light in the interaction dimension under the current driving scenario is determined based on the structured feature vector, including:

[0043] The identification value is determined based on the contrast between the grille light pattern and the background of the corresponding grille light pattern, and the ambient illuminance in the environment vector.

[0044] The second modulation factor is determined using the current visibility and the maximum visibility in the environment vector;

[0045] The target identification value is obtained by modulating the identification value using the second modulation factor;

[0046] The degree of matching between the information conveyed by the grille light pattern and the user's intent is determined by the user intent vector in the corresponding grille light mode.

[0047] The sub-utility score of the interaction dimension is determined based on the target identification value and the matching degree value.

[0048] Optionally, determining the grille light illumination strategy based on preset priority rules and the utility scores of each grille light mode in the current driving scenario includes:

[0049] The grid light pattern corresponding to the highest overall utility score among all the overall utility scores is determined as the target lighting pattern in the grid light lighting strategy.

[0050] When multiple lighting modes have the same highest overall utility score, the grid light mode with the highest score is selected as the target lighting mode in the grid light lighting strategy based on the sub-utility scores of the safety dimension, energy efficiency dimension, and interaction dimension in turn.

[0051] If multiple grid light patterns still exist, the currently used grid light pattern will be used as the target lighting pattern in the grid light lighting strategy.

[0052] Optionally, the method for determining the grille light illumination strategy further includes:

[0053] Determine the base illumination brightness, base color temperature, and base length of the light spot in the target illumination mode;

[0054] The target brightness is obtained by correcting the basic lighting brightness based on the ambient illuminance in the environment vector;

[0055] The target pattern parameters are obtained by correcting the base length of the light spot based on the distance between the vehicle and the obstacle in the environment vector.

[0056] The base color temperature is corrected based on the correlated color temperature in the environment vector and the preset reference color temperature to obtain the target color temperature.

[0057] Optionally, after determining the grille light illumination strategy and driving the grille lights to execute it, the method further includes:

[0058] After a preset time period, feedback data is collected through multimodal sensors;

[0059] Based on the feedback data, the performance of the grille lights is evaluated in three dimensions: safety, energy efficiency, and interaction, to obtain the evaluation results.

[0060] When the preset learning conditions are met, the internal parameters are updated and trained based on the evaluation results, and the updated internal parameters are applied to subsequent grille light control decisions.

[0061] In a second aspect, embodiments of the present invention provide a vehicle grille light control device, the device comprising:

[0062] The acquisition module acquires driving scene information of the current driving scenario through a multimodal sensor, and determines a structured feature vector based on the driving scene information;

[0063] The first determining module determines the three-dimensional weights corresponding to the current driving scenario based on the structured feature vector. The three-dimensional weights include safety weights, energy efficiency weights, and interaction weights.

[0064] The second determining module determines the utility score of each preset grille light pattern in the current driving scenario based on the three-dimensional weights and the structured feature vectors.

[0065] The third determining module determines the grille light lighting strategy based on preset priority rules and the utility score of each grille light mode in the current driving scenario, and drives the grille light to execute the corresponding lighting strategy. The grille light lighting strategy includes target lighting mode, target brightness, target pattern parameters and target color temperature.

[0066] Thirdly, embodiments of the present invention provide an electronic device, including:

[0067] At least one processor; and

[0068] At least one memory communicatively connected to the processor, wherein:

[0069] The memory stores program instructions that can be executed by the processor, which can invoke the program instructions to perform the method as described in any of the first aspects.

[0070] Fourthly, embodiments of the present invention provide a storage medium including a stored program, wherein, when the program is executed, it controls the device where the storage medium is located to perform the method described in any of the first aspects.

[0071] This invention collects scene information of the current driving scenario through multimodal sensors, and makes decisions on the grille light lighting strategy based on this information. This enables the grille light to move from a single function of lighting to adaptive lighting, adjustment and interaction with the outside of the vehicle in complex environments, which significantly improves driving safety, energy efficiency and user experience.

[0072] Specifically, by using quantitative collision risk models and anti-glare models, the grille lights are integrated into the active safety system, achieving a leap in driving safety. By incorporating the power consumption of the grille lights into the vehicle's energy flow management, the driving range is maximized while ensuring safety, achieving a leap in energy efficiency. By generating interactive information highly correlated with vehicle status and environment, vehicle language is standardized and made intelligent, achieving natural and efficient external interaction. Through reinforcement learning, the system can continuously optimize its decisions, adapt to user personalization and new driving scenarios, and possesses flexibility and robustness unmatched by traditional rule-based systems. Attached Figure Description

[0073] To more clearly illustrate the technical solutions of the embodiments of this application, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0074] Figure 1 The diagram shown is a flowchart of a vehicle grille light control method provided in an embodiment of this application;

[0075] Figure 2 The diagram shown is a structural schematic of a vehicle grille light control device provided in an embodiment of this application;

[0076] Figure 3 The diagram shown is a structural schematic of an electronic device provided in an embodiment of this application. Detailed Implementation

[0077] To better understand the technical solution of this application, the embodiments of this application will be described in detail below with reference to the accompanying drawings.

[0078] It should be understood that the described embodiments are merely some, not all, of the embodiments in this application. All other embodiments obtained by those skilled in the art based on the embodiments in this application without inventive effort are within the scope of protection of this application.

[0079] like Figure 1 The diagram shown is a flowchart of a vehicle grille light control method provided in an embodiment of the present invention. See also... Figure 1 The specific steps of this method include:

[0080] S101 collects driving scene information of the current driving scenario through a multimodal sensor and determines a structured feature vector based on the driving scene information.

[0081] Specifically, multimodal sensors include vision units, radar units, environmental sensing units, vehicle sensor units, and positioning and communication units mounted on the vehicle.

[0082] The vision unit specifically includes a front-facing high-definition camera and a surround-view camera, used to identify environmental information such as pedestrians, vehicles, and traffic signs, as well as ambient light color temperature and perform illumination intensity detection.

[0083] The radar unit specifically includes a front-mounted millimeter-wave radar and a lidar, used to achieve accurate ranging and speed measurement, in order to compensate for the shortcomings of the vision unit in identifying environmental information in adverse weather conditions.

[0084] The environmental sensing unit specifically includes an ambient light sensor, a rain / humidity sensor, and an ambient temperature sensor, which are used to collect environmental weather information around the vehicle.

[0085] The vehicle sensor unit specifically includes a high-speed CAN / FD bus interface and sensors installed inside the vehicle, used to acquire vehicle status information in real time such as vehicle speed, steering angle, yaw rate, accelerator / brake pedal opening, drive mode, battery SOC, and on-board network load.

[0086] The positioning and communication unit specifically includes a GPS / BeiDou module and a V2X communication module (such as V2V, V2I, V2P) to acquire high-precision map information, traffic signal status, and the intentions of surrounding vehicles.

[0087] Based on the environmental information collected by the vision unit, and integrating information collected by the radar unit and the environmental sensing unit, the system performs target detection and classification of obstacles around the vehicle, generating a dynamic obstacle list for each obstacle around the vehicle. The dynamic obstacle list includes obstacle information for multiple identified and calculated obstacles, including at least one of the following: obstacle type, distance between the vehicle and the obstacle, relative speed between the vehicle and the obstacle, and relative angle between the vehicle and the obstacle.

[0088] The visibility index (Vis) and weather code (Wcode) are calculated using image textures collected by the vision unit, radar echoes collected by the radar unit, and rainfall data collected by environmental sensing data.

[0089] In this embodiment of the invention, the visibility index is estimated by fusing the image contrast acquired by the visual unit and the intensity of the radar echo acquired by the radar unit, and can be calculated, for example, by the following formula:

[0090]

[0091] in, This represents the contrast value of the image captured by the visual unit. The intensity of the radar echo. The weights are used to balance image contrast and radar echo intensity.

[0092] Weather coding is a multi-dimensional vector output by inputting environmental weather information collected by environmental sensing units into a pre-trained fuzzy classifier. Its vector formula can include confidence scores for rainfall, fog, etc.

[0093] Based on the collected driving scenario information, as well as the processed and calculated dynamic obstacle list, visibility index, weather code, etc., a structured feature vector is determined. The structured vector includes an environment vector, a vehicle state vector, and a user intent vector.

[0094] The environment vector includes at least one of the following: ambient illuminance, correlated color temperature, visibility index, weather code, and a dynamic obstacle list.

[0095] The vehicle state vector includes at least one of the following: vehicle speed, battery state of charge, instantaneous power consumption, driving mode, and autonomous driving level.

[0096] The user intent vector includes at least one of the following: turn signal status, brake pedal opening, accelerator pedal opening, and user personalized preference encoding.

[0097] S102, based on the structured feature vector, determine the corresponding three-dimensional weights in the current driving scenario.

[0098] Specifically, the three-dimensional weights include safety weight, energy efficiency weight, and interaction weight. These three-dimensional weights are used to characterize the importance of each dimension in the current driving scenario.

[0099] When determining the safety weights in the current driving scenario, it is necessary to determine the real-time collision risk between the vehicle and each obstacle based on the obstacle information of each obstacle in the dynamic obstacle list of the environment vector.

[0100] In this embodiment of the invention, the parameters are determined based on the relative speed between the vehicle and the obstacle, the relative angle between the vehicle and the obstacle, the relative distance between the vehicle and the obstacle, the type of obstacle, and the weather code.

[0101] The following is a formula for calculating real-time collision risk provided by an embodiment of the present invention:

[0102]

[0103] in, For the real-time collision risk between an obstacle and a vehicle, The relative speed between the vehicle and the obstacle. The relative angle between the vehicle and the obstacle. The relative distance between the vehicle and the obstacle. To address different obstacle types The set weights, To code different types of weather The weather deterioration factor set.

[0104] In the above formula, This reflects the inverse square relationship between relative speed and relative distance; the higher the relative speed and the closer the relative distance, the higher the collision risk. The relative angle describes the obstacle's direction of motion. Different weight values ​​are assigned based on different obstacle types, generally pedestrians > non-motorized vehicles > motorized vehicles. For example, pedestrians could be set to 1.0, non-motorized vehicles to 0.8, and motorized vehicles to 0.5. Different weather deterioration factors are set based on different weather types, generally dense fog > heavy rain > light rain > clear skies. For example, dense fog could be set to 2.0, heavy rain to 1.5, light rain to 1.2, and clear skies to 1.0.

[0105] After calculating the real-time collision risk of each obstacle, the maximum value among all real-time collision risks is selected as the target collision risk value at the current moment.

[0106] The target collision risk value is input into a preset sigmoid function. The target collision risk value is normalized by the sigmoid function and a preset collision risk threshold to obtain a safety weight.

[0107] The following is a calculation formula for normalizing the target collision risk value provided by an embodiment of the present invention:

[0108] =σ(x) =

[0109] in, For safety weights, σ(x) is a predefined sigmoid function. This is the slope coefficient of the sigmoid function, used to control the steepness of the weight changes; it can generally be set to 6. The calculated and selected target collision risk value. This is a preset collision risk threshold, typically set to 0.5. When the target collision risk value exceeds this threshold... At that time, the safety weight will increase sharply.

[0110] Optionally, in this embodiment of the invention, in addition to the collision risk threshold, a risk avoidance threshold is also set. Generally, the risk avoidance threshold is higher than the collision risk threshold. When the target collision risk value exceeds the risk avoidance threshold, the system can directly trigger an interruption mechanism, directly determine the lighting strategy of the grille light as the "emergency avoidance" lighting mode, skip the subsequent calculation process, and achieve the fastest response.

[0111] When determining the energy efficiency weight under the current driving scenario, the battery state of charge is input into another preset sigmoid function. The battery state of charge is normalized by the sigmoid function and the preset energy threshold to obtain the energy sensitivity value.

[0112] The following is a calculation formula for normalizing the state of charge of a battery, provided by an embodiment of the present invention:

[0113]

[0114] in, The battery is in its state of charge. The default sigmoid function. The slope coefficient of the sigmoid function. This is a preset power threshold, which is the critical power level that triggers the energy-saving mode. The preferred value is 0.15-0.20. When the battery power... When the power consumption is below this threshold, the power sensitivity value increases sharply, which will increase the energy efficiency weight, and the system will be more inclined to adjust the lighting mode of the grille lights to energy-saving mode.

[0115] The above formula maps continuous battery charge values ​​to a sensitivity coefficient within the interval [0,1] to characterize the sensitivity of energy efficiency to changes in battery charge.

[0116] After determining the power sensitivity value, the energy efficiency weight is determined based on the safety weight and the power sensitivity value. The safety weight and the energy efficiency weight are negatively correlated, while the power sensitivity value and the energy efficiency weight are positively correlated.

[0117] The following is a calculation formula for calculating energy efficiency weight provided by an embodiment of the present invention:

[0118]

[0119] in, As an energy efficiency weight, For safety weights, This is the power sensitivity value. The higher the safety weight, the lower the energy efficiency weight; the worse the battery's state of charge, i.e., the lower the remaining power, the higher the power sensitivity value and the higher the energy efficiency weight.

[0120] When determining the interaction weights in the current driving scenario, the interaction weights can be determined directly based on the values ​​of the safety weight and the energy efficiency weight.

[0121] Generally, the difference between 1 and the sum of the safety weight and the energy efficiency weight is used as the interaction weight. The interaction weight is negatively correlated with the sum of the safety weight and the energy efficiency weight.

[0122] The following is a calculation formula for calculating interaction weights provided by an embodiment of the present invention:

[0123]

[0124] in, For interaction weights, For safety weights, Energy efficiency weighting.

[0125] The autonomous driving system will allocate its remaining computing resources to interactive functions only when there is a clear interaction request or when safety and energy efficiency requirements are not urgent. The urgency of the interaction request, i.e., the interaction weight, will increase.

[0126] S103, based on the three-dimensional weights and structured feature vectors, determine the utility score of each preset grille light mode in the current driving scenario.

[0127] Specifically, based on the structured feature vectors, the sub-utility scores of each grille light mode in the current driving scenario are determined in the dimensions of safety, energy efficiency, and interaction.

[0128] The sub-utility scores of the three dimensions are weighted and summed using the three-dimensional weights calculated by S102 to obtain the comprehensive utility score for each grille light pattern. The comprehensive utility score is used to characterize the degree of matching between the features of different grille light patterns and the current driving scenario.

[0129] In this embodiment of the invention, the grille light mode may include, but is not limited to, strong lighting mode, energy-saving mode, interactive warning mode, personalized display mode, and emergency avoidance mode.

[0130] The strong lighting mode uses high-brightness white light for full-area illumination, suitable for scenarios such as nighttime rural areas, inclement weather, and tunnel driving; the energy-saving mode uses low-brightness blue light for contour indicator lighting, suitable for scenarios such as low battery, charging status, and economy driving mode; the interactive warning mode uses a combination of dynamic arrows and text to project light spots onto pedestrian walkways, suitable for scenarios such as autonomous driving yielding, pedestrian interaction at intersections, and construction area warnings; the personalized display mode uses custom patterns, suitable for scenarios such as vehicle stationary status, user-personalized expression, and vehicle display mode; and the emergency avoidance mode is suitable for scenarios where obstacles around the vehicle pose a danger to vehicle safety.

[0131] In determining the sub-utility score of the grille light in the safety dimension under the current driving scenario, it is necessary to determine the effective lighting area, visual gain, glare area, and glare risk probability respectively.

[0132] The effective lighting area is determined based on the total number of grille light sectors, the effective lighting area of ​​each sector, and the number of sectors required to be illuminated for the corresponding grille light mode. The effective lighting area is the area of ​​the road surface that is actually illuminated and useful to the driver under the corresponding grille light mode.

[0133] The following is a calculation formula for calculating the effective lighting area provided by an embodiment of the present invention:

[0134]

[0135] in, For effective lighting area, This represents the total number of sectors for the grille lights. The effective lighting area of ​​a sector To determine the number of sectors that need to be illuminated for a given mode of grille light, generally, a sector is illuminated when the light intensity received by the sector reaches the minimum effective light threshold, or when the actual brightness of the sector reaches the minimum brightness threshold visible to the human eye. This is the weighting function for the corresponding mode of the grille light, used to adjust the effective illumination area of ​​the corresponding mode of the grille light.

[0136] The visual gain is determined based on the ambient illuminance in the environment vector. The visual gain is the improvement in driver vision per unit illuminated area.

[0137] The following is a calculation formula for visual gain provided by an embodiment of the present invention:

[0138]

[0139] in, For visual gain, For adjustment coefficients, This refers to ambient light intensity. Visual gain is negatively correlated with ambient light intensity; the lower the ambient light intensity, such as at night, the higher the visual gain.

[0140] Glare area refers to the area that may cause glare to other road users, such as oncoming drivers and pedestrians, for the corresponding grille light pattern.

[0141] The glare risk probability is determined based on the distance between the vehicle and obstacles in the environmental vector, the obstacle type, and a preset distance threshold for the distance between the vehicle and obstacles. The glare risk probability characterizes the probability that the glare area will interfere with other road users.

[0142] The following is a calculation formula for calculating the probability of glare risk provided by an embodiment of the present invention:

[0143]

[0144] in, This represents the probability of glare. These are the preset adjustment parameters. The relative distance between the vehicle and the obstacle. A preset distance threshold for the distance between the vehicle and the obstacle. To address different obstacle types The weights set are generally: pedestrians > non-motorized vehicles > motorized vehicles.

[0145] After calculating the effective lighting area, visual gain, glare area, and glare risk probability, the sub-utility score of the safety dimension is determined accordingly.

[0146] The following is a calculation formula for calculating the sub-utility score of the security dimension provided by an embodiment of the present invention:

[0147]

[0148] in, For the sub-utility score of the safety dimension, For effective lighting area, For visual gain, Glare area, This represents the probability of glare risk.

[0149] In the above formula, the difference between the gain generated by illumination and the negative effect of illumination on glare to surrounding obstacles, i.e. the final actual gain, is used as the sub-utility score of the safety dimension.

[0150] When determining the sub-utility score of the grille light in the current driving scenario, the energy efficiency value needs to be determined based on the estimated power consumption of the corresponding grille light mode and the maximum theoretical power consumption. A first modulation factor is then determined based on the battery state of charge in the vehicle state vector. The energy efficiency value is modulated using this first modulation factor to obtain the sub-utility score of the energy efficiency dimension.

[0151] The following is a calculation formula for calculating the sub-utility score of the energy efficiency dimension provided by an embodiment of the present invention:

[0152]

[0153] in, For the sub-utility score of the energy efficiency dimension, The estimated power consumption for the corresponding mode is determined based on factors such as LED brightness, the number of illuminated sectors, and duty cycle. This is the maximum theoretical power consumption, typically the power consumption of all LEDs operating at maximum brightness and 100% duty cycle. This is the power sensitivity coefficient. It incurs a greater penalty to power consumption when the battery is low. When the battery level is low, this coefficient increases the penalty to power consumption, causing the system to more readily adopt power-saving mode. Its value typically ranges from 2 to 5. This indicates the battery's state of charge.

[0154] When determining the sub-utility score of the grille light in the current driving scenario, it is necessary to determine the target identification value and the degree of matching between the information conveyed by the grille light pattern and the user's intent in the corresponding grille light mode.

[0155] First, the recognition value is determined based on the contrast between the grille light pattern and the background of the corresponding grille light pattern, as well as the ambient illuminance in the environment vector. Then, a second modulation factor is determined using the current visibility and maximum visibility in the environment vector. The recognition value is then modulated using the second modulation factor to obtain the target recognition value. The target recognition value characterizes the degree to which the pattern or text displayed in the corresponding pattern is clearly recognizable by other road users under given environmental conditions.

[0156] The following is a calculation formula for calculating target identification values ​​provided by an embodiment of the present invention:

[0157]

[0158] in, The target identification value for the corresponding pattern. These are preset parameters. To adjust the contrast between the grille light pattern and the background. For ambient light intensity, A constant used to prevent division by zero. For current visibility, This represents the maximum visibility.

[0159] As can be seen from the above formula, the higher the contrast between the grid light pattern and the background, the lower the ambient light brightness, and the higher the ratio of current visibility to maximum visibility, the higher the target recognition value.

[0160] The matching degree between the information conveyed by the grille light pattern in the corresponding grille light mode and the user's intent is determined by the user intent vector. Specifically, this can be calculated based on a predefined mapping table or semantic similarity. In a specific embodiment, when the vehicle is preparing to turn, the matching degree value of the arrow pattern displayed in the interactive warning mode will be higher than that of other modes.

[0161] The sub-utility score of the interaction dimension is determined based on the target identification value and the matching degree value. Generally, the product of the target identification value and the matching degree value is used as the sub-utility score of the interaction dimension. The higher the target identification value and the higher the matching degree, the higher the sub-utility score of the interaction dimension.

[0162] After calculating the sub-utility scores for the three dimensions separately, the sub-utility scores for the three dimensions are weighted and summed using three-dimensional weights to obtain the comprehensive utility score for the corresponding pattern.

[0163] S104: Based on the preset priority rules and the utility scores of each grille light mode in the current driving scenario, determine the grille light lighting strategy and drive the grille lights to execute the corresponding lighting strategy.

[0164] Specifically, the grid lighting strategy includes target lighting mode, target brightness, target pattern parameters, and target color temperature.

[0165] When determining the target lighting mode, the grid light mode with the highest overall utility score among all modes in the current scene is determined as the target lighting mode in the grid light lighting strategy.

[0166] If multiple lighting modes have the same highest overall utility score, then the grid light mode with the highest score is selected as the target lighting mode in the grid light lighting strategy, based on the sub-utility scores of the safety dimension, energy efficiency dimension, and interaction dimension in turn.

[0167] That is, if there are multiple modes with the highest overall utility scores, the mode with the highest sub-utility score in the safety dimension is further prioritized as the target lighting mode; if there are multiple modes with the highest sub-utility scores in the safety dimension, the mode with the highest sub-utility score in the energy efficiency dimension is further prioritized as the target lighting mode; if there are multiple modes with the highest sub-utility scores in the energy efficiency dimension, the mode with the highest sub-utility score in the interaction dimension is further prioritized as the target lighting mode.

[0168] Furthermore, if multiple grid light patterns with the same sub-utility score exist across multiple interaction dimensions, the currently used grid light pattern will be selected as the target lighting pattern in the grid light illumination strategy. If no currently used grid light pattern exists, a grid light pattern will be randomly selected as the target lighting pattern.

[0169] After determining the target lighting pattern, it is necessary to fine-tune the brightness, pattern parameters, color, etc., based on the target lighting pattern and the current environmental parameters.

[0170] Determine the base illumination brightness, base color temperature, and base spot length of the target illumination pattern.

[0171] The target brightness is obtained by correcting the basic lighting brightness using the ambient illuminance in the environment vector.

[0172] The following is a calculation formula for calculating target brightness provided by an embodiment of the present invention:

[0173]

[0174] in, For target brightness, The base brightness for the target lighting mode, This is the brightness adjustment coefficient, used to control the sensitivity of brightness to changes in ambient light, and is typically between [0,1]. This refers to ambient light intensity.

[0175] As can be seen from the above formula, the darker the ambient light, the higher the base brightness will be.

[0176] The target pattern parameters are obtained by correcting the base length of the light spot using the distance between the vehicle and the obstacle in the environment vector.

[0177] The following is a calculation formula for calculating the spot length provided by an embodiment of the present invention:

[0178] Length =

[0179] Where Length is the length of the light spot. Where is the base length of the light spot in the target lighting mode, and D is the relative distance between the vehicle and the obstacle. For distance functions, the usual f(D) = min(1, ),in This is a reference distance, for example, it could be 10 meters.

[0180] As can be seen from the above formula, when the relative distance between the vehicle and the obstacle is small, the length of the light spot will be shortened proportionally to avoid the light spot being too long and exceeding the sidewalk.

[0181] In this embodiment of the invention, different pattern parameters will have different adjustment functions. The above-mentioned adjustment of the spot length is only an example of adjusting the pattern parameters.

[0182] The target color temperature is obtained by correcting the base color temperature using the correlated color temperature in the environment vector and the preset reference color temperature.

[0183] The following is a calculation formula for calculating the target color temperature provided by an embodiment of the present invention:

[0184]

[0185] in, For the target color temperature, The base color temperature in the target lighting mode. This is the amount for color temperature adjustment.

[0186] The amount of color temperature adjustment is determined by the correlated color temperature and the preset reference color temperature. Generally, =μ ( ),in, It is the reference ambient color temperature (e.g., 5500K), and μ is the adjustment coefficient (e.g., 0.1), used to make the color temperature of the grille light closer to the ambient color temperature and reduce the abruptness.

[0187] After determining the lighting strategy for the grille lights, the entire strategy is packaged into a control command package and sent to the driver unit within the grille lights. The driver unit then controls the LEDs in the high-resolution grille light array to illuminate according to the control commands, based on the mode ID, brightness value, color value, and dynamic pattern data in the command package. Simultaneously, the projection correction unit runs projection distortion correction and adaptive white balance algorithms to ensure accurate output.

[0188] Optionally, in this embodiment of the invention, after determining the grille light illumination strategy and driving the grille light to illuminate, it is necessary to collect feedback data and train the model based on the feedback data.

[0189] Specifically, feedback data is collected via multimodal sensors after a preset time period, such as 100ms. Based on the collected feedback data, the performance of the grille lights is evaluated in three dimensions: safety, energy efficiency, and interaction.

[0190] When conducting safety assessments, radar is used to monitor for abnormal behavior of vehicles ahead (such as emergency braking or sudden steering) that may be caused by glare; cameras are used to monitor the clarity and contrast of road patterns to ensure they do not interfere with the driver; and V2X messages are used to receive safety-related messages (such as complaints about glare) from other vehicles or roadside units. Based on this comprehensive information, a safety score is generated.

[0191] When evaluating energy efficiency, the energy consumption value is estimated by using the LED power model based on the actual brightness and dynamic effects, or by calculating the energy consumed in the current cycle based on the actual current sensor readings.

[0192] When evaluating the interaction dimension, V2X is used to receive thank-you signals (positive feedback) or complaint signals (negative feedback) from other vehicles. The interaction effectiveness is assessed by analyzing changes in the behavior of surrounding vehicles (such as whether they maintain a safe distance or drive smoothly). Finally, a comprehensive score is output, such as [-1, 1], to represent the transition from negative to positive feedback.

[0193] When the preset learning conditions are met, the model is updated and trained based on the evaluation results, and the updated model is applied to subsequent grille light control decisions.

[0194] After collecting the above feedback data and evaluation results, the collected feedback data and evaluation results are used as experience samples to perform machine learning training on the model, so as to enhance the model's learning ability and optimize the grid light control decision strategy.

[0195] The experience sample typically includes the state at the beginning of the current grille light control cycle, i.e., the environment vector, vehicle state vector, and user intent vector obtained in this cycle; the final decision, i.e., the final output grille light illumination strategy; and the three-dimensional evaluation of the corresponding grille light illumination strategy.

[0196] When training the model using empirical samples, a batch of samples is randomly selected from the pre-stored empirical samples. The gradient descent algorithm is then used to adjust the internal parameters of the weight calculation function and utility function used in this scheme, with the goal of maximizing long-term returns. Finally, the updated model parameters are saved and loaded at the beginning of the next control cycle, thus influencing all future decisions. At this point, the system has completed a full iterative optimization.

[0197] This invention collects driving scenario information from a multimodal sensor, and makes decisions on the grille light lighting strategy based on this information. This enables the grille light to move beyond a single function of illumination to adaptive illumination, adjustment, and interaction with the outside of the vehicle in complex environments, significantly improving driving safety, energy efficiency, and user experience.

[0198] Specifically, by using quantitative collision risk models and anti-glare models, the grille lights are integrated into the active safety system, achieving a leap in driving safety. By incorporating the power consumption of the grille lights into the vehicle's energy flow management, the driving range is maximized while ensuring safety, achieving a leap in energy efficiency. By generating interactive information highly correlated with vehicle status and environment, vehicle language is standardized and made intelligent, achieving natural and efficient external interaction. Through reinforcement learning, the system can continuously optimize its decisions, adapt to user personalization and new driving scenarios, and possesses flexibility and robustness unmatched by traditional rule-based systems.

[0199] Corresponding to the above-described vehicle grille light control method, this application also provides a vehicle grille light control device. See [link to relevant documentation]. Figure 2 This is a schematic diagram of the structure of a vehicle grille light control device provided in an embodiment of this application. The vehicle grille light control device may include: a data acquisition module 201, a first determination module 202, a second determination module 203, and a third determination module 204.

[0200] The acquisition module 201 acquires driving scene information of the current driving scene through a multimodal sensor, and determines a structured feature vector based on the driving scene information.

[0201] The first determining module 202 determines the three-dimensional weights corresponding to the current driving scenario based on the structured feature vector. The three-dimensional weights include safety weights, energy efficiency weights, and interaction weights.

[0202] The second determining module 203 determines the utility score of each preset grille light mode in the current driving scenario based on the three-dimensional weights and the structured feature vectors.

[0203] The third determining module 204 determines the grille light lighting strategy based on the preset priority rules and the utility score of each grille light mode in the current driving scenario, and drives the grille light to execute the corresponding lighting strategy. The grille light lighting strategy includes the target lighting mode, target brightness, target pattern parameters and target color temperature.

[0204] Figure 3 This is a schematic diagram illustrating the structure of one embodiment of the electronic device described in this specification. Specifically, the electronic device can be implemented as a device for executing the vehicle grille light control method provided in this embodiment. For example... Figure 3 As shown, the above-mentioned electronic device may include at least one processor; and at least one memory communicatively connected to the processing unit, wherein: the memory stores program instructions executable by the processing unit, and the processor can execute the vehicle grille light control method provided in this embodiment by calling the program instructions.

[0205] The aforementioned electronic device can be a device capable of intelligent dialogue with the user, such as a cloud server. This specification does not limit the specific form of the electronic device in the embodiments. It is understood that the electronic device here refers to the machine mentioned in the method embodiments.

[0206] Figure 3 A block diagram of an exemplary electronic device suitable for implementing embodiments of this specification is shown. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of use of the embodiments described in this specification.

[0207] like Figure 3 As shown, the electronic device is represented in the form of a general-purpose computing device. The components of the electronic device may include, but are not limited to: one or more processors 310, communication interface 320, memory 330, and communication bus 340 connecting different system components (including memory 330, communication interface 320 and processor 310).

[0208] Communication bus 340 represents one or more of several bus architectures, including a memory bus or memory controller, a peripheral bus, a graphics acceleration port, a processor, or a local bus using any of the various bus architectures. For example, these architectures include, but are not limited to, Industry Standard Architecture (ISA) buses, Micro Channel Architecture (MAC) buses, Enhanced ISA buses, Video Electronics Standards Association (VESA) local buses, and Peripheral Component Interconnect (PCI) buses.

[0209] Electronic devices typically include a variety of computer-readable media. These media can be any available media that can be accessed by the electronic device, including volatile and non-volatile media, and removable and non-removable media.

[0210] Memory 330 may include computer system readable media in the form of volatile memory, such as random access memory (RAM) and / or cache memory. The electronic device may further include other removable / non-removable, volatile / non-volatile computer system storage media. Memory 330 may include at least one program product having a set (e.g., at least one) of program modules configured to perform the functions of the embodiments described herein.

[0211] A program / utility having a set (at least one) of program modules may be stored in memory 330. Such program modules include, but are not limited to, an operating system, one or more application programs, other program modules, and program data. Each or some combination of these examples may include an implementation of a network environment. The program modules typically perform the functions and / or methods described in the embodiments of this specification.

[0212] The processor 310 executes various functional applications and data processing by running programs stored in the memory 330, such as implementing the vehicle grille light control method provided in the embodiments shown in this specification.

[0213] This specification provides a non-transitory computer-readable storage medium storing computer instructions that cause the computer to execute the vehicle grille light control method provided in the embodiments shown in this specification.

[0214] The aforementioned non-transitory computer-readable storage medium may be any combination of one or more computer-readable media. A computer-readable medium may be a computer-readable signal medium or a computer-readable storage medium. A computer-readable storage medium may be, for example, but not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of computer-readable storage media (a non-exhaustive list) include: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM), or flash memory, optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In this document, a computer-readable storage medium may be any tangible medium containing or storing a program that may be used by or in connection with an instruction execution system, apparatus, or device.

[0215] Computer-readable signal media may include data signals propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. Computer-readable signal media may also be any computer-readable medium other than computer-readable storage media, capable of sending, propagating, or transmitting programs for use by or in connection with an instruction execution system, apparatus, or device.

[0216] Program code contained on a computer-readable medium may be transmitted using any suitable medium, including but not limited to wireless, wire, optical fiber, RF, etc., or any suitable combination thereof.

[0217] Computer program code for performing the operations described herein can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, as well as conventional procedural programming languages ​​such as "C" or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or it can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0218] The foregoing has described specific embodiments of this specification. Other embodiments are within the scope of the appended claims. In some cases, the actions or steps recited in the claims may be performed in a different order than that shown in the embodiments and may still achieve the desired result. Furthermore, the processes depicted in the drawings do not necessarily require the specific or sequential order shown to achieve the desired result. In some embodiments, multitasking and parallel processing are possible or may be advantageous.

[0219] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this specification, "a plurality of" means at least two, such as two, three, etc., unless otherwise explicitly specified.

[0220] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing custom logic functions or processes, and the scope of the preferred embodiments of this specification includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which the embodiments of this specification pertain.

[0221] Depending on the context, the word "if" as used here can be interpreted as "when," "when," "in response to determination," or "in response to detection." Similarly, depending on the context, the phrase "if determination" or "if detection (of the stated condition or event)" can be interpreted as "when determination," "in response to determination," "when detection (of the stated condition or event)," or "in response to detection (of the stated condition or event)."

[0222] It should be noted that the terminals involved in the embodiments of this specification may include, but are not limited to, personal computers (PCs), personal digital assistants (PDAs), wireless handheld devices, tablet computers, mobile phones, MP3 players, MP4 players, etc.

[0223] In the embodiments provided in this specification, it should be understood that the disclosed systems, apparatuses, and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative; for instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed. Furthermore, the coupling or direct coupling or communication connection shown or discussed may be through some interfaces; the indirect coupling or communication connection between apparatuses or units may be electrical, mechanical, or other forms.

[0224] Furthermore, the functional units in the various embodiments of this specification can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or in a combination of hardware and software functional units.

[0225] The integrated units implemented as software functional units described above can be stored in a computer-readable storage medium. These software functional units, stored in a storage medium, include several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) or processor to execute some steps of the methods described in the various embodiments of this specification.

[0226] The above description is merely a preferred embodiment of this specification and is not intended to limit this specification. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this specification should be included within the scope of protection of this specification.

Claims

1. A method for controlling vehicle grille lights, characterized in that, The method includes: The driving scene information of the current driving scenario is collected by a multimodal sensor, and a structured feature vector is determined based on the driving scene information; Based on the structured feature vector, the three-dimensional weights corresponding to the current driving scenario are determined respectively. The three-dimensional weights include safety weight, energy efficiency weight and interaction weight. Based on the three-dimensional weights and the structured feature vectors, the utility scores of each preset grille light mode in the current driving scenario are determined respectively. Based on preset priority rules and the utility scores of each grille light mode in the current driving scenario, a grille light illumination strategy is determined, and the grille lights are driven to execute the corresponding illumination strategy. The grille light illumination strategy includes target illumination mode, target brightness, target pattern parameters, and target color temperature.

2. The method according to claim 1, characterized in that, The multimodal sensor includes a vision unit, a radar unit, an environmental sensing unit, a vehicle sensor unit, and a positioning and communication unit. The structured feature vectors determined based on the driving scenario information include environment vectors, vehicle state vectors, and user intent vectors. The environment vector includes at least one of the following: ambient illuminance, correlated color temperature, visibility index, weather code, and a dynamic obstacle list. The dynamic obstacle list includes obstacle information for multiple obstacles, and the obstacle information includes obstacle type, distance between the vehicle and the obstacle, relative speed between the vehicle and the obstacle, and relative angle between the vehicle and the obstacle. The vehicle state vector includes at least one of the following: vehicle speed, battery state of charge, instantaneous power consumption, driving mode, and autonomous driving level. The user intent vector includes at least one of the following: turn signal status, brake pedal opening, accelerator pedal opening, and user personalized preference encoding.

3. The method according to claim 2, characterized in that, The step of determining the three-dimensional weights corresponding to the current driving scenario based on the structured feature vector includes: Determine the safety weights under the current driving scenario information, including: Based on the dynamic obstacle list, the obstacle information of each obstacle is used to determine the real-time collision risk between the vehicle and each obstacle. The obstacle with the highest real-time collision risk is identified as the target collision risk value. The target collision risk value is normalized by a preset collision risk threshold to obtain the safety weight. Determine the energy efficiency weights under the current driving scenario information, including: The battery state of charge is normalized by a preset power threshold to obtain a power sensitivity value. The energy efficiency weight is determined based on the safety weight and the power sensitivity value; The safety weight is negatively correlated with the energy efficiency weight, while the power sensitivity value is positively correlated with the energy efficiency weight. Determine the interaction weights under the current driving scenario information, including: The interaction weight is determined based on the security weight and the energy efficiency weight; The interaction weight is negatively correlated with the sum of the security weight and the energy efficiency weight.

4. The method according to claim 2, characterized in that, The step of determining the utility score of each preset grille light pattern in the current driving scenario based on the three-dimensional weights and the structured feature vectors includes: Based on the structured feature vectors, the sub-utility scores of each grille light mode in the current driving scenario are determined in the safety, energy efficiency, and interaction dimensions, respectively. The comprehensive utility score is obtained by weighting and summing the sub-utility scores of the three dimensions using the aforementioned three-dimensional weights. Specifically, determining the sub-utility score of the grille light in the safety dimension under the current driving scenario based on the structured feature vector includes: The effective lighting area is determined based on the total number of grille light sectors, the effective lighting area of ​​each sector, and the number of sectors required to be lit for the corresponding grille light mode. The visual gain is determined based on the ambient illuminance in the environment vector. The glare risk probability is determined based on the relative distance between the vehicle and the obstacle in the environmental vector, the obstacle type, and the preset distance threshold between the vehicle and the obstacle. The sub-utility score of the safety dimension is determined based on the effective lighting area, the visual gain, the preset glare area, and the glare risk probability. The sub-utility score of the grille light in the energy efficiency dimension under the current driving scenario is determined based on the structured feature vector, including: The energy efficiency value is determined based on the estimated power consumption of the corresponding grille light mode and the maximum theoretical power consumption. The first modulation factor is determined based on the battery state of charge in the vehicle state vector. The energy efficiency value is modulated by the first modulation factor to obtain the sub-utility score of the energy efficiency dimension; The sub-utility score of the grille light in the interaction dimension under the current driving scenario is determined based on the structured feature vector, including: The identification value is determined based on the contrast between the grille light pattern and the background of the corresponding grille light pattern, and the ambient illuminance in the environment vector. The second modulation factor is determined using the current visibility and the maximum visibility in the environment vector; The target identification value is obtained by modulating the identification value using the second modulation factor; The degree of matching between the information conveyed by the grille light pattern and the user's intent is determined by the user intent vector in the corresponding grille light mode. The sub-utility score of the interaction dimension is determined based on the target identification value and the matching degree value.

5. The method according to claim 4, characterized in that, The process of determining the grille light illumination strategy based on preset priority rules and the utility scores of each grille light mode in the current driving scenario includes: The grid light pattern corresponding to the highest overall utility score among all the overall utility scores is determined as the target lighting pattern in the grid light lighting strategy. When multiple lighting modes have the same highest overall utility score, the grid light mode with the highest score is selected as the target lighting mode in the grid light lighting strategy based on the sub-utility scores of the safety dimension, energy efficiency dimension, and interaction dimension in turn. If multiple grid light patterns still exist, the currently used grid light pattern will be used as the target lighting pattern in the grid light lighting strategy.

6. The method according to claim 2, characterized in that, The method for determining the grille light illumination strategy further includes: Determine the base illumination brightness, base color temperature, and base length of the light spot in the target illumination mode; The target brightness is obtained by correcting the basic lighting brightness based on the ambient illuminance in the environment vector; The target pattern parameters are obtained by correcting the base length of the light spot based on the distance between the vehicle and the obstacle in the environment vector. The base color temperature is corrected based on the correlated color temperature in the environment vector and the preset reference color temperature to obtain the target color temperature.

7. The method according to claim 1, characterized in that, After determining the grille light illumination strategy and driving the grille light to execute it, the method further includes: After a preset time period, feedback data is collected through multimodal sensors; Based on the feedback data, the performance of the grille lights is evaluated in three dimensions: safety, energy efficiency, and interaction, to obtain the evaluation results. When the preset learning conditions are met, the internal parameters are updated and trained based on the evaluation results, and the updated internal parameters are applied to subsequent grille light control decisions.

8. A vehicle grille light control device, characterized in that, The device includes: The acquisition module acquires driving scene information of the current driving scenario through a multimodal sensor, and determines a structured feature vector based on the driving scene information; The first determining module determines the three-dimensional weights corresponding to the current driving scenario based on the structured feature vector. The three-dimensional weights include safety weights, energy efficiency weights, and interaction weights. The second determining module determines the utility score of each preset grille light pattern in the current driving scenario based on the three-dimensional weights and the structured feature vectors. The third determining module determines the grille light lighting strategy based on preset priority rules and the utility score of each grille light mode in the current driving scenario, and drives the grille light to execute the corresponding lighting strategy. The grille light lighting strategy includes target lighting mode, target brightness, target pattern parameters and target color temperature.

9. An electronic device, characterized in that, include: At least one processor; as well as At least one memory communicatively connected to the processor, wherein: The memory stores program instructions that can be executed by the processor, and the processor can invoke the program instructions to perform the method as described in any one of claims 1 to 7.

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

Images