METHOD FOR ADJUSTING EXTERNAL AIR INTAKE IN AN INTERIOR OF A VEHICLE

Abstract

A method for adjusting external air intake in an interior of a vehicle involves continuously identifying an interior pollution level during a driving operation of the vehicle using recorded signals of a pollution sensor arranged in the interior. A pollution level of external air on a section of road ahead of the vehicle is predicted and the external air intake is automatically regulated depending on the predicted pollution level of the external air.

Claims

1-8. (canceled)

9. A method for adjusting external air intake in an interior of a vehicle, the method comprising: continuously identifying, during a driving operation of the vehicle, an interior pollution level using recorded signals of a pollution sensor arranged in the interior of the vehicle; identifying, using recorded signals from at least one vehicle-mounted visual recording unit, visual information; extrapolating semantic information from the identified visual information; predicting a pollution level of external air on a section of road ahead of the vehicle based on the semantic information; and automatically regulating the external air intake depending on the predicted pollution level of the external air.

10. The method of claim 9, further comprising: extrapolating, using the visual information, a fingerprint for identifying pollution-producing objects in signals from the at least one visual recording unit that will be recorded in the future.

11. The method of claim 9, wherein pollution-producing moving objects and stationary objects in the visual information about an environment of the vehicle are recorded as the semantic information.

12. The method of claim 10, further comprising: continuously identifying a current position of the vehicle.

13. The method of claim 12, further comprising: training, based on the extrapolated semantic information, the extrapolated fingerprint, or the identified current position of the vehicle, a time-delayed model in the vehicle in such a way that the interior pollution level is predicted as a set target.

14. The method of claim 13, further comprising: determining a time delay between a high interior pollution level and a cause of the high pollution level extrapolated by the time-delayed model.

15. The method of claim 13, further comprising: supplying the trained model to a central computer unit.

16. The method of claim 13, further comprising: supplying the trained model to further vehicles of a vehicle fleet.

Description

BRIEF DESCRIPTION OF THE SOLE FIGURE

[0017] Here:

[0018] The sole FIGURE schematically shows a vehicle with a pollution sensor and a visual recording unit as well as different stationary and moving objects in an environment of the vehicle.

DETAILED DESCRIPTION

[0019] The sole FIGURE illustrates a vehicle 1 with a pollution sensor 2 and a visual recording unit 3 in the form of a camera, wherein in addition a lorry 4 and a bicycle 5 are shown as moving objects O1, an industrial plant 6 is shown as a stationary object O2, and a central computer unit 7 is shown.

[0020] The pollution sensor 2 is arranged in an interior of the vehicle 1 and continuously records signals during a driving operation of the vehicle 1, based on which signals the pollution level of the interior is identified. Alternatively, or additionally to the pollution sensor 2, a sensor can also be arranged in the interior of the vehicle 1 that records signals, based on which smells that are perceptible to humans can be recognized.

[0021] The pollution level of the interior that is identified based on recorded signals of the pollution sensor 2 represents a time-delayed and time-integrated function of a pollution level of an external environment of the vehicle 1, i.e., of external air.

[0022] If an interior pollution level is identified based on the recorded signals of the pollution sensor 2, a ventilation of an interior of the vehicle 1 is controlled, as is known from the prior art.

[0023] In order to perform an adjustment of an external air intake in the interior of the vehicle 1, in which the interior pollution level is kept as low as possible, so that a health of occupants can be protected in terms of a pollution level, a method described in the following is provided.

[0024] A pollution level of the external air on a section of road ahead of the vehicle 1 is thereby predicted and the external air intake is automatically regulated depending on the predicted pollution level.

[0025] In particular, the method provides that the external air intake into the interior of the vehicle 1 is turned off if a comparatively high pollution level of the external air is expected in the future. By means of the method, the external air intake is thus reduced before pollutants enter the interior of the vehicle 1 with the external air.

[0026] As described above, signals are continuously recorded during the driving operation of the vehicle 1 by means of the pollution sensor 2 in the vehicle, based on which signals an interior pollution level is determined.

[0027] The vehicle 1 has the visual recording unit 3 in the form of the camera, the recording area of which is directed in front of the vehicle 1 and by means of which signals are continuously recorded during the driving operation of the vehicle 1, based on which signals an environment of the vehicle 1 and objects O1, O2 in this environment are detected.

[0028] Furthermore, the vehicle 1 comprises a satellite-supported position-determining unit that is not shown in more detail and a digital map, so that a current position of the vehicle 1 can be determined.

[0029] Based on the recoded signals of the visual recording unit 3, visual information is identified, which is evaluated.

[0030] Based on the visual information, semantic information can thereby be extrapolated by means of methods of machine learning, wherein, in particular, object recognition is used in relation to known objects O1, O2. I.e., it can be recognized which objects O1, O2 are in the environment of the vehicle 1. The objects O1, O2 detected in the environment of the vehicle 1 can, in particular, be differentiated into moving objects O1 and stationary objects O2.

[0031] Additionally, based on the semantic information, it can also be identified which of the objects O1, O2 produce pollution. If, e.g., a bicycle 5 is detected as a moving object O1 in the environment of the vehicle 1, then based on certain features of the bicycle 5 it is recognized that it is a bicycle 5, which does not produce pollution.

[0032] Furthermore, a fingerprint, in particular a perception fingerprint, can be extrapolated based on the visual information, for example by means of convolutional neural networks, without it being necessary to assign predefined objects O1, O2. During a development of a model created by means of the method that is discussed further below, non-predefined and known objects O1, O2 can thereby by identified by means of the fingerprint or by means of several extrapolated fingerprints.

[0033] Based on the semantic information, the perception fingerprint and/or based on position information identified due to the current position of the vehicle 1, a time-delayed model in the vehicle 1 is trained with machine learning methods. The model is thereby trained in such a way that a measured interior pollution level is predicted as a set target. The model can, for example, be a regression model or a model using reinforcement learning.

[0034] In the model, a time delay and a temporal integration of the pollution level of the external air is taken into account, for example by using time series information.

[0035] Such a model implicitly enables determining the time delay between a comparatively high interior pollution level, in particular a pollution level that exceeds a predetermined threshold, and a cause for a comparatively high pollution level of the external air that is identified by the model. For example, such a cause of the high pollution level is a lorry 4 and/or a tractor with a trailer full of manure, so-called slurry, driving in front of the vehicle.

[0036] This model trained in such a way is used in the vehicle 1, in order to predict an expected interior pollution level of the vehicle 1 by means of input information, in particular by means of the semantic information, of the perception fingerprint and/or position information. If a future comparatively high interior pollution level is predicted, then an interior ventilation is regulated in such a way that the external air intake into the interior of the vehicle 1 is reduced or switched off according to the predicted interior pollution level.

[0037] In one embodiment, the model can be pre-trained during a development of the vehicle 1, wherein the model can also be user- or region-specifically trained and/or trained further in further vehicles of a vehicle fleet.

[0038] It is also conceivable that general, user- and/or region-specific models be merged or aggregated in a central computer unit 7, for example of a vehicle manufacturer, by distributed learning, so-called federated learning, based on experiences of the further vehicles of the vehicle fleet.

[0039] The model additionally enables a recognition of causes of a comparatively high pollution concentration, e.g., a lorry 4 and/or an industrial plant 6, which can be region-specific. Such information can, for example, be used to identify sources of a comparatively high pollution concentration in the environment of the vehicle 1 in a country- and/or region-specific way.

[0040] By means of the central computer unit 7, the respective models and/or the aggregated model can be made available to the vehicle 1 and to the further vehicles of the vehicle fleet.

[0041] The method thus provides that, by means of visual information from the visual recording unit 3, a geolocation, i.e., a current position of the vehicle 1, and, based on the pollution sensor 2, a model, is trained, which predicts the interior pollution level, i.e., an interior pollution concentration, in particular based on the visual information.

[0042] For example, known moving objects O1 and stationary objects O2 can correlate with a comparatively high pollution level of the external air and thus also correlate in a time-delayed manner with a comparatively high interior pollution level, wherein the presented model learns this correlation.

[0043] Although the invention has been illustrated and described in detail by way of preferred embodiments, the invention is not limited by the examples disclosed, and other variations can be derived from these by the person skilled in the art without leaving the scope of the invention. It is therefore clear that there is a plurality of possible variations. It is also clear that embodiments stated by way of example are only really examples that are not to be seen as limiting the scope, application possibilities or configuration of the invention in any way. In fact, the preceding description and the description of the figures enable the person skilled in the art to implement the exemplary embodiments in concrete manner, wherein, with the knowledge of the disclosed inventive concept, the person skilled in the art is able to undertake various changes, for example, with regard to the functioning or arrangement of individual elements stated in an exemplary embodiment without leaving the scope of the invention, which is defined by the claims and their legal equivalents, such as further explanations in the description.