Adaptive cruise control system, vehicle and method for controlling a vehicle
The adaptive cruise control system addresses the issue of non-customizable speed control by using a machine learning model to integrate driver and situational data, resulting in a more personalized and safe driving experience.
Patent Information
- Authority / Receiving Office
- WO · WO
- Patent Type
- Applications
- Current Assignee / Owner
- ZF CV SYST GLOBAL GMBH
- Filing Date
- 2025-12-09
- Publication Date
- 2026-06-25
Smart Images
Figure EP2025086052_25062026_PF_FP_ABST
Abstract
Description
[0001] Hannover, 19.12.2024 IP, Schaferjohann, Putzke / Kw 305698-IN-NP ID 305698
[0002] Adaptive Cruise Control System, Vehicle and Method for Controlling a Vehicle
[0003] The invention relates to an adaptive cruise control system, a vehicle and a method for controlling a vehicle.
[0004] Many modern vehicles are equipped with adaptive cruise control systems. Adaptive cruise control systems observe the distance between a vehicle and a second vehicle driving ahead of the vehicle and control the speed of the vehicle to maintain a desired speed while maintaining a safe distance between the vehicle and the second vehicle driving ahead.
[0005] The optimized speed of the vehicle results in improved and more efficient usage of fuel. Improved fuel efficiency means reduced carbon dioxide emissions and other pollutants associated with burning fossil fuels. This helps vehicle manufacturers to meet environmental regulations and individuals to reduce their carbon footprint and costs. By always keeping the proper safety distance to the vehicle ahead, the adaptive cruise control system also improves road safety. This is particularly important since two of the main causes of accidents with personal injury are insufficient distance to the vehicle ahead and unadjusted speed. An adaptive cruise control system, furthermore, improves convenience for the driver by maintaining the desired speed instead of the driver having to manually control the speed.
[0006] Adaptive cruise control systems typically control the speed of the vehicle on the basis of a plurality of parameters. These parameters are typically not customizable by the driver. Therefore, the driving style in which the adaptive cruise control system controls the speed of the vehicle might be perceived as too aggressive or too hesitant by the driver.
[0007] Existing solutions, such as described in US2017057517A1 , DE10343178A1 , and US1 1834042B2, try to learn the driving pattern of the driver through artificial intelligence or machine learning models which create acceleration and deceleration profiles suiting the driver. It is an object of the invention to improve the adaptation of the adaptive cruise control system to the driving behavior of the driver.
[0008] The object is achieved by an adaptive cruise control system in accordance with claim 1.
[0009] The adaptive cruise control system includes an adaptive cruise control unit. The adaptive cruise control unit is configured to automatically control a speed of a vehicle based on an input parameter set for the adaptive cruise control unit. The adaptive cruise control system comprises a driver behavior determination unit configured to determine a driver behavior information. The adaptive cruise control system comprises a situation determination unit configured to determine a situation information. The adaptive cruise control system comprises a parameterization unit. The parameterization unit is configured to output an input parameter set for the adaptive cruise control unit based on a machine learning model having the driver behavior information and the situation information as inputs.
[0010] Thus, the input parameter set is generated on the basis of not only the driver behavior information but also the situation information. This provides for a more finegrained adaptation of the adaptive cruise control system to the driver behavior. Instead of learning “one” driver behavior and applying it to all driving situations, the adaptive cruise control system uses additionally the situation information to replicate the individual and situation-dependent driving behavior.
[0011] The term “situation” refers to conditions that typically influence the driver behavior. Such conditions can, for example, pertain to an interior of the vehicle and / or to surroundings of the vehicle. More examples of situation information will be described below.
[0012] The driver behavior generally includes acceleration requests and braking requests issued by the driver. Thus, in a typical scenario the driver behavior determination unit may record actuation patterns of an accelerator padel, a brake padel, a retarder, if present, and / or a gear shift, if present. The machine learning model may preferably be configured to output an input parameter set for the adaptive cruise control unit. In an example, this input parameter set may be fed directly into the adaptive cruise control unit and / or the parameterization unit may be configured to directly output the output of the machine learning model as the input parameter set for adaptive cruise control unit. In an alternative example, it is possible to check, for example by a check unit, and, if necessary, change, for example by a change unit, the output of the machine learning model before outputting an input parameter set for the adaptive cruise control unit by the parameterization unit. The check unit may, for example, be configured to operate deterministically.
[0013] According to an embodiment, the parameterization unit is configured to train the machine learning model based on the driver behavior information and the situation information. This further improves adaptation of the behavior of the adaptive cruise control unit to the behavior of the driver. The driver behavior information and / or the situation information may, thus, form training data. Training data typically consists of examples or observations that the model uses to learn patterns and make predictions. By feeding this data into the machine learning model, it can adjust its parameters and improve its performance on the task of producing an appropriate output. Later when the adaptive cruise control unit controls the speed of the vehicle, the parameterization unit can produce an input parameter set based on live situation data and the machine learning model trained with the driver behavior information. Thus, the parameterization unit provides for a situation dependent speed control behavior that is highly similar to how the driver would behave if he or she controlled the speed of the vehicle.
[0014] Additionally, it is possible to pre-train the machine learning model before first use of the vehicle and to further train the machine learning model while using the vehicle equipped with the adaptive cruise control system by training the machine learning model during use of the vehicle.
[0015] According to an embodiment, the parameterization unit is configured to train the machine learning model based on the driver behavior information and the situation information, when the adaptive cruise control unit is inactive. When the adaptive cruise control unit is inactive, the driver has full control over the vehicle and operates the accelerator pedal and / or the brake pedal.
[0016] According to an embodiment, the parameterization unit is configured to train the machine learning model based on the driver behavior information and the situation information when the driver overrides the adaptive cruise control unit. If the driver actuates the accelerator pedal and / or the brake pedal while the adaptive cruise control unit is active, the adaptive cruise control system and / or the parameterization unit may detect a driver override. An override often means that the driver is not satisfied with the behavior of the adaptive cruise control unit in a given situation. Thus, considering overrides in the training process further improves adaptation of the adaptive cruise control unit to the driver’s behavior.
[0017] The situation determination unit may be configured to determine the situation information by means of one or more sensors. Such sensors may be or include a laser sensor, a radar sensor, a camera, and / or a GNSS sensor. Furthermore, the situation determination unit may be configured to determine the situation information based on a signal received from a communication unit, such as through V2X-communication and / or over the internet.
[0018] The situation information may comprise a trailer information about whether or not the vehicle comprises a trailer. Driving with a trailer typically makes the driver change his or her driving behavior. The additional weight of a trailer typically leads to an additional load on the brakes and extends the braking distance.
[0019] The situation information may comprise a slope information about a slope of a road on which the vehicle is driving. A slope information can comprise an information about whether the vehicle is driving uphill, downhill, on a flat road or a slope value of the road. The slope of the road can influence how the driver controls the driving speed as well as the degree of acceleration and deceleration.
[0020] The situation information may comprise a passenger situation information, preferably comprising a number of passengers in the vehicle and / or a number of seated passengers in the vehicle and / or a number of standing passengers in the vehicle. The driver may drive more cautiously, if passengers are present in the vehicle and, in particular, if standing passengers are present in the vehicle. The presence of passengers also may change the weight of the vehicle and the weight distribution within the vehicle, which also may affect the driver’s behavior.
[0021] The situation information may comprise a load condition of the vehicle. The load condition may represent an extent to which the vehicle is loaded, for example as a percentage of a maximum load of the vehicle or as a classification into classes such as “empty”, “half full”, “full”, or the like.
[0022] The situation information may comprise a load type of the vehicle. The load type may comprise information regarding the type of the load, such as “liquid”, “solid”, “passengers”, “goods”, “concrete”, “refrigerated”, “heated”, “dangerous goods” or the like.
[0023] The situation information may comprise a road condition of a road on which the vehicle is driving. A road condition can be, for example, “good”, “bad”, “smooth”, “bumpy”, “dangerous” or the like.
[0024] The situation information may comprise a road type information. A road type information can be, for example, “highway”, “country road”, “city road”, “trail” or the like.
[0025] The situation information may comprise a weather condition. A weather condition can be, for example, “windy”, “stormy”, “sunny”, “foggy”, “snowy” or “rainy”. Especially in heavy rain fall, fog or snow, the driver’s driving behavior may change drastically and might result in a much more careful way of driving.
[0026] The situation information may comprise a vision condition. A vision condition can be, for example, “bright”, “dark”, “clear” or “blurry”, or the like. If the vision is affected by lighting conditions or obstructed by fog, for example, the driver may be more likely to drive more cautiously.
[0027] The situation information may comprise a traffic density information. A traffic density information can be, for example, “low density”, “medium density”, “high density”, “congested”, or the like. The driver’s behavior typically depends on the given situation. With changing conditions, the driver’s behavior may change slightly or drastically. By considering the situation information and the driving behavior as input for the machine learning model, the machine learning model can learn the driver’s behavior in a particularly sophisticated manner and create a detailed and realistic reproduction of the driver’s driving behavior. The adaptive cruise control system can, thus, be perceived as highly pleasant and intelligent by the driver.
[0028] According to an embodiment, the machine learning model comprises a generative model. A generative model is particularly advantageous for producing an input parameter set for the adaptive cruise control unit. Generative models typically have the flexibility to generate data, learn from unlabeled data, and capture the full data distribution.
[0029] According to an embodiment, the generative model is a generative adversarial network. Generative adversarial networks generally produce high quality data and excel at modelling complex data distributions, allowing them to reproduce the driver’s behavior appropriate to the situation. Generative adversarial networks can operate in an unsupervised manner, they can learn without having to rely on labelled data. As an alternative example, the generative model may be a variational autoencoder.
[0030] According to an embodiment, the driver behavior information comprises a classification of the driver behavior. This can improve the performance of the system as it simplifies the driver behavior information. Classification of the driver’s behavior can preferably include classification into one group of a predetermined set of groups. Classes or groups can be, for example, “conservative”, “moderate”, “aggressive” or similar. This allows for a faster prediction of the driver’s behavior.
[0031] According to an embodiment, the driver behavior determination unit is configured to determine the driver behavior information and / or classify the driver behavior based on a second machine learning model. This is a simple and efficient solution. According to an embodiment, the parameterization unit may be configured to output the input parameter set for the adaptive cruise control unit based on a driver identity information. This further improves adaptation of the adaptive cruise control system to the driver’s behavior. The adaptive cruise control system may include a driver identification unit configured to identify a driver and to output the driver identity information. The parameterization unit may preferably be configured to use individual machine learning models for each driver identity.
[0032] The object is also achieved by a vehicle according to the claim directed thereto and / or a vehicle comprising an adaptive cruise control system as described above.
[0033] The vehicle can be, for example, a commercial vehicle, a truck, a passenger car or a bus. The vehicle can be, for example, a road vehicle or a rail vehicle. The vehicle can be, for example, a single vehicle, a tractor for a tractor-trailer combination or a tractor-trailer combination.
[0034] The object is also achieved by a method for controlling a vehicle in accordance with the claim directed thereto and / or a method for controlling a vehicle, for example a vehicle as described above and / or by means of an adaptive cruise control system as described above, the method comprising automatically controlling a speed of the vehicle according to an adaptive cruise control scheme based on an input parameter set; determining a driver behavior information; determining a situation information; and outputting an input parameter set for the adaptive cruise control scheme based on a machine learning model having the driver behavior information and the situation information as inputs.
[0035] If devices and methods are described herein, the methods described can advantageously be developed further by the embodiments and individual features of the devices, and vice versa.
[0036] The invention is described in more detail below with reference to examples, which are shown in schematic drawings.
[0037] Fig. 1 shows a vehicle with an adaptive cruise control system. Fig. 2 shows a vehicle with an adaptive cruise control system.
[0038] Fig. 3 shows a vehicle with an adaptive cruise control system.
[0039] Fig. 4 shows a method for controlling a vehicle.
[0040] Fig. 1 shows a vehicle 10 comprising an adaptive cruise control system 11. The adaptive cruise control system 11 comprises an adaptive cruise control unit 12. The adaptive cruise control unit 12 is configured to automatically control 13 a speed 14 of the vehicle 10 based on an input parameter set 16 for the adaptive cruise control unit 12.
[0041] The adaptive cruise control system 11 of the vehicle 10 of Fig. 1 comprises a driver behavior determination unit 18 configured to determine a driver behavior information 20, e.g. acceleration requests and braking requests issued by a driver 15 of the vehicle 10 actuating an accelerator padel or a brake padel (not shown in Fig. 1), respectively.
[0042] The adaptive cruise control system 11 of the vehicle 10 of Fig. 1 comprises a situation determination unit 24 configured to determine a situation information 26, i.e. an information about a fact that tends to influence the driving behavior of the driver. The situation information 20 may include, for example, information regarding road conditions and / or load conditions.
[0043] The adaptive cruise control system 11 of the vehicle 10 of Fig. 1 comprises a parameterization unit 28. The parameterization unit 28 is configured to output an input parameter set 16 for the adaptive cruise control unit 13 based on a machine learning model 30 having the driver behavior information 20 and the situation information 26 as inputs 32.
[0044] The parameterization unit 28 shown in Fig. 1 is configured to train the machine learning model 30 based on the driver behavior information 20 and situation information 26. Thus, the input parameter set 16 shown in Fig. 1 is generated on the basis of not only the driver behavior information 20 but also the situation information 24. This provides for a fine grained adaptation of how the adaptive cruise control unit 12 controls 13 the speed 14 to how the driver 15 would.
[0045] The machine learning model 30 of the parameterization unit 28 shown in Fig. 1 comprises a generative model 34. The generative model 34 may be or comprise a generative adversarial network 36.
[0046] The driver behavior information 20 of the adaptive cruise control system 12 of Fig. 1 comprises a driver behavior classification 38, such as “conservative”, “moderate”, “aggressive” or similar. The driver behavior determination unit 18 is configured to determine the driver behavior classification 38 based on a second machine learning model 40.
[0047] The adaptive cruise control system 11 of Fig. 1 comprises a driver identification unit 41 configured to identify a driver 15 and to output a driver identity information 42. The parameterization unit 28 is configured to output the input parameter set 16 for the adaptive cruise control system 12 based on the driver identity information 42. The parameterization unit 28 may preferably be configured to use individual machine learning models 30 for each driver 15.
[0048] Fig. 2 shows another vehicle 10 comprising an adaptive cruise control system 11. The adaptive cruise control system 11 comprises an adaptive cruise control unit 12. The adaptive cruise control unit 12 is configured to automatically control 13 a speed 14 of the vehicle 10 based on an input parameter set 16 for the adaptive cruise control unit 12. The adaptive cruise control system 11 comprises a driver behavior determination unit 18 configured to determine a driver behavior information 20.
[0049] The adaptive cruise control system 11 comprises a situation determination unit 24 configured to determine a situation information 26. The adaptive cruise control system 11 comprises a parameterization unit 28. The parameterization unit 28 is configured to output an input parameter set 16 for the adaptive cruise control unit 12 based on a machine learning model 30 having the driver behavior information 20 and the situation information 26 as inputs 32.
[0050] The situation information 26 may comprise a trailer information 44 about whether or not the vehicle 10 comprises a trailer 46, as shown in Fig. 2.
[0051] The situation information 26 may comprise a slope information 48 about a slope 50 of a road 52 on which the vehicle 10 is driving, as shown in Fig. 2. A slope information 48 can comprise an information about whether the vehicle 10 is driving uphill, downhill, on a flat road or a slope value of the road 52. The slope 50 of the road 52 can influence how the driver 15 controls the driving speed 14 as well as the degree of acceleration and deceleration.
[0052] The situation information 26 may comprise a passenger situation information 54, as shown in Fig. 2. One passenger 55 is present in the vehicle 10 in Fig. 2. The passenger situation information 54 may comprise a number of passengers 56 in the vehicle 10 and / or a number of seated passengers 58 in the vehicle 10 and / or a number of standing passengers 60 in the vehicle 10. The driver 15 may drive more cautiously, if passengers 55 are present in the vehicle 10 and, in particular, if standing passengers 55 are present in the vehicle 10.
[0053] The situation information 26 may comprise a load condition 66 of the vehicle 10, as shown in Fig. 2. The load condition 66 may represent an extent to which the vehicle 10 is loaded, for example as a percentage of a maximum load of the vehicle 10 or as a classification into classes such as “empty”, “half full”, “full” or the like.
[0054] The situation information 26 may comprise a load type 70 of the vehicle 10, as shown in Fig. 2. The load type 70 may comprise information regarding the type of a load 71 of the vehicle 10, such as “liquid”, “solid”, “passengers”, “goods”, “concrete”, “refrigerated”, “heated”, “dangerous goods” or the like.
[0055] The situation information 26 may comprise a road condition 72 of a road 52 on which the vehicle 10 is driving, as shown in Fig. 2. A road condition 72 can be, for example, one of “good”, “bad”, “smooth”, “bumpy”, “dangerous”, or the like. The situation information 26 may comprise a road type information 74, as shown in Fig. 2. A road type information 74 can be, for example, be “highway”, “country road”, “city road”, “trail”, or the like.
[0056] The situation information 26 may comprise a weather condition 76, as shown in Fig. 2. A weather condition 76 can be, for example, one of “windy”, “stormy”, “sunny”, “foggy”, “snowy”, “rainy” or the like. Especially in heavy rain fall, fog or snow, the driver’s driving behavior may change drastically and might result in a much more careful way of driving.
[0057] The situation information 26 may comprise a vision condition 78, as shown in Fig. 2. A vision condition 78 can be, for example, one of “bright”, “dark”, “clear”, “blurry”, or the like. If the vision of the driver 15 is affected by lighting conditions or obstructed by fog, the driver 15 may be more likely to drive more cautiously.
[0058] The situation information 26 may comprise a traffic density information 80, as shown in Fig. 2. A traffic density information 80 can be, for example, one of “low density”, “medium density”, “high density”, “congested”, or the like.
[0059] The situation determination unit 24 shown in Fig. 2 may be configured to determine the situation information 26 by means of one or more sensors 82. The one or more sensors 82 may be or include a laser sensor, a radar sensor, a camera, and / or a GNSS sensor.
[0060] Fig. 3 shows another vehicle 10 comprising an adaptive cruise control system 11. The adaptive cruise control system 11 comprises an adaptive cruise control unit 12. The adaptive cruise control unit 12 is configured to automatically control 13 a speed 14 of the vehicle 10 based on an input parameter set 16 for the adaptive cruise control unit 12. The adaptive cruise control system 11 comprises a driver behavior determination unit 18 configured to determine a driver behavior information 20. The adaptive cruise control system 11 comprises a situation determination unit 24 configured to determine a situation information 26. The adaptive cruise control system 11 comprises a parameterization unit 28. The parameterization unit 28 is configured to output an input parameter set 16 for the adaptive cruise control unit 12 based on a machine learning model 30 having the driver behavior information 20 and the situation information 26 as inputs 32.
[0061] Fig. 3 further shows a driver 15 controlling 33 the speed 14 the vehicle 10 by means of an accelerator padel 84 and a brake padel 86. The accelerator padel 84 and brake padel 86 issue acceleration requests 88 and braking requests 90, respectively, upon actuation by the driver 15. The acceleration requests 88 and braking requests 90 are forwarded to the driver behavior determination unit 18 and the driver behavior determination unit 18 is configured to determine the driver behavior information 20 based on the acceleration requests 88 and braking requests 90.
[0062] The parameterization unit 28 of the vehicle 10 of Fig. 3 is configured to train the machine learning model 30 based on the driver behavior information 20 and the situation information 26.
[0063] The parameterization unit 28 may be configured to train the machine learning model 30 based on the driver behavior information 20 and the situation information 26 when the adaptive cruise control unit 12 is inactive. When the adaptive cruise control unit 12 is inactive, the driver 15 has full control over the vehicle 10 and operates the accelerator pedal 84 and / or the brake pedal 86 as needed.
[0064] The parameterization unit 28 of the vehicle 10 of Fig. 3 may, additionally or alternatively, be configured to train the machine learning model 30 based on the driver behavior information 20 and the situation information 26, when the driver 15 overrides 87 the adaptive cruise control unit 12. If the driver 15 actuates the accelerator pedal 84 and / or the brake pedal 86 while the adaptive cruise control unit 12 is active, i.e. while the adaptive cruise control unit 12 is controlling 13 the speed 14 of the vehicle 10, the adaptive cruise control unit 12 and / or the parameterization unit 28 may detect a driver override 87.
[0065] Fig. 4 shows a method 92 for controlling 94 a vehicle. The method 92 comprises automatically controlling 13 a speed 14 of the vehicle according to an adaptive cruise control scheme 96 based on an input parameter set 16. The adaptive cruise control scheme 94 may be implemented in an adaptive cruise control unit 12. The method 92 further includes determining 98 a driver behavior information 20, determining 100 a situation information 26 and outputting 102 an input parameter set 16 for the adaptive cruise control scheme 96 based on a machine learning model 30 having the driver behavior information 20 and the situation information 26 as inputs 32.
[0066] Where similar or identical elements are shown in different figures, reference numerals are assigned accordingly. Multiple descriptions of similar or identical elements have been avoided for the sake of clarity. Nevertheless, the embodiments of the fig- ures can be combined with each other and developed further in accordance with the other embodiments and / or their individual features.
[0067] List of references (part of the description)
[0068] 10 vehicle
[0069] 11 adaptive cruise control system
[0070] 12 adaptive cruise control unit
[0071] 13 control
[0072] 14 speed
[0073] 15 driver
[0074] 16 input parameter set
[0075] 18 driver behavior determination unit
[0076] 20 driver behavior information
[0077] 24 situation determination unit
[0078] 26 situation information
[0079] 28 parameterization unit
[0080] 30 machine learning model
[0081] 32 inputs
[0082] 34 generative model
[0083] 36 generative adversarial network
[0084] 38 classification
[0085] 40 second machine learning model
[0086] 41 driver identification unit
[0087] 42 driver identity information
[0088] 44 trailer information
[0089] 46 trailer
[0090] 48 slope information
[0091] 50 slope
[0092] 52 road
[0093] 54 passenger situation information
[0094] 56 number of passengers
[0095] 58 number of seated passengers
[0096] 60 number of standing passengers
[0097] 62 passengers
[0098] 66 load condition 68 load
[0099] 70 load type
[0100] 71 load
[0101] 72 road condition
[0102] 74 road type information
[0103] 76 weather condition
[0104] 78 vision condition
[0105] 80 traffic density information
[0106] 82 sensors
[0107] 84 accelerator pedal
[0108] 86 brake pedal
[0109] 87 override
[0110] 88 acceleration request
[0111] 90 braking request
[0112] 92 method
[0113] 94 controlling
[0114] 96 adaptive cruise control scheme
[0115] 98 determining
[0116] 100 determining
[0117] 102 outputting
Claims
Claims1 . Adaptive cruise control system (11) for a vehicle (10), wherein the adaptive cruise control system (11) comprises an adaptive cruise control unit (12), wherein the adaptive cruise control unit (12) is configured to automatically control (13) a speed (14) of the vehicle (10) based on an input parameter set (16) for the adaptive cruise control unit (12), wherein the adaptive cruise control system (11) comprises a driver behavior determination unit (18) configured to determine a driver behavior information (20), wherein the adaptive cruise control system (11) comprises a situation determination unit (24) configured to determine a situation information (26), wherein the adaptive cruise control system (11) comprises a parameterization unit (28), wherein the parameterization unit (28) is configured to output an input parameter set (16) for the adaptive cruise control unit (13) based on a machine learning model (30) having the driver behavior information (20) and the situation information (26) as inputs (32).
2. Adaptive cruise control system (11) according to claim 1 , wherein the parameterization unit (28) is configured to train the machine learning model (30) based on the driver behavior information (20) and the situation information (26).
3. Adaptive cruise control system (11) according to claim 2, wherein the parameterization unit (28) is configured to train the machine learning model (30) based on the driver behavior information (20) and the situation information (26), when the adaptive cruise control unit (12) is inactive.
4. Adaptive cruise control system (11) according to claim 2 or 3, wherein the parameterization unit (28) is configured to train the machine learning model (30) based on the driver behavior information (20) and the situation information (26), when the driver (15) overrides (87) the adaptive cruise control unit (12).
5. Adaptive cruise control system (11) according to one of the preceding claims, wherein the situation information (26) comprises one or more of the following:- a trailer information (44) about whether or not the vehicle (10) comprises a trailer (46),- a slope information (48) about a slope (50) of a road (52) on which the vehicle (10) is driving,- a passenger situation information (54), preferably comprising a number of passengers (56) in the vehicle (10) and / or a number of seated passengers (58) in the vehicle (10) and / or a number of standing passengers (60) in the vehicle (10),- a load condition (66) of the vehicle (10),- a road condition (72) of a road (52) on which the vehicle (10) is driving,- a road type information (74),- a weather condition (76),- a vision condition (78),- a traffic density information (80).
6. Adaptive cruise control system (11) according to one of the preceding claims, wherein the machine learning model (30) comprises a generative model (34).
7. Adaptive cruise control system (11) according to claim 6, wherein the generative model (34) is a generative adversarial network (36).
8. Adaptive cruise control system (11) according to one of the preceding claims, wherein the driver behavior information (20) comprises a driver behavior classification (38).
9. Adaptive cruise control system (11) according to one of the preceding claims, wherein the driver behavior determination unit (18) is configured to determine the driver behavior classification (38) based on a second machine learning model (40).
10. Adaptive cruise control system (11) according to one of the preceding claims, wherein the parameterization unit (28) is configured to output the input parameterset (16) for the adaptive cruise control unit (12) based on a driver identity information (42).
11. Vehicle (10) comprising an adaptive cruise control system (11) in accordance with one of the preceding claims.
12. Method (92) for controlling (94) a vehicle (10), the method (92) comprising: automatically controlling (13) a speed (14) of the vehicle (10) according to an adap- tive cruise control scheme (96) based on an input parameter set (16); determining (98) a driver behavior information (20); determining (100) a situation information (26); outputting (102) an input parameter set (16) for the adaptive cruise control scheme (96) based on a machine learning model (30) having the driver behavior in- formation (20) and the situation information (26) as inputs (32).