DE102020204833A1 - Method and device for fusing a plurality of signals from an ultrasonic sensor system of a means of locomotion - Google Patents

Abstract

The present invention relates to a method and a device for fusing a plurality of signals from an ultrasonic sensor system (30, 35) of a means of locomotion (80). The method comprises the steps of: receiving a first weighting factor for a first signal, which represents a first hydroplaning risk value, receiving a second weighting factor for a second signal, which represents a second hydroplaning risk value, determining a third weighting factor representing the reciprocal sum of the reciprocal first weighting factor and the reciprocal second weighting factor, determining a weighted mean value for the first hydroplaning risk value and the second hydroplaning risk value from the sum of respective quotients from the first hydroplaning risk value and the first weighting factor, and the second hydroplaning risk value and the second weighting factor, the sum of these quotients being weighted by means of the third weighting factor Determination of a weighted variance for the first aquaplaning risk value and the second aquaplaning risk value on the basis of respective deviations of the respective aquaplan Ingrisikowerte from the weighted mean value taking into account the first weighting factor, the second weighting factor and the third weighting factor, and using the weighted mean value and the weighted variance to determine a resulting hydroplaning risk state, wherein the weighted mean value represents the merging of the first signal and the second signal.

Description

State of the art

The present invention relates to a method and a device for fusing a plurality of signals from an ultrasonic sensor system of a means of locomotion.

Means of locomotion are known from the prior art which are provided with one or more ultrasonic sensors for detecting the surroundings of the means of locomotion. For this purpose, the ultrasonic sensors are used to first emit ultrasonic waves into a predefined area of an area surrounding the means of transport. A transit time of the ultrasonic waves is then determined on the basis of the ultrasonic waves reflected or scattered by the environment to the ultrasonic sensors. Finally, on the basis of the transit time of the ultrasonic waves, a distance between the means of transport and objects in the vicinity of the means of transport can be determined. Distance information determined in this way to objects in the vicinity of the means of locomotion is received and used in the prior art, for example by maneuvering assistance systems of the means of locomotion (e.g. lane change assistant, parking assistant, etc.).

Furthermore, semi-autonomous and / or autonomously driving means of transport are known from the prior art, which take into account additional information about the conditions of roads currently being driven on (e.g. wetness, uneven road surfaces) when controlling the means of transport to ensure safe driving.

DE102016221809A1 describes a method for analyzing an unevenness of a roadway for a vehicle, with a signal relating to a position and / or movement of the vehicle being detected and with the unevenness of the roadway being characterized by evaluating the signal. The signal can be recorded using an ultrasonic sensor, among other things. By filtering the signal using a bandpass filter or using several bandpass filters in one or more frequency ranges, a computationally intensive frequency analysis can be dispensed with and the unevenness of the road can be described using a small amount of data. In addition, a mean value and a variance of the signal can be determined and used advantageously for analyzing the unevenness of the road.

DE 10 2018 118 679 A1 describes the communication and avoidance of hazards in a monitored driving environment on the basis of an application of a system for improved target object recognition in a vehicle, which is equipped with a laser recognition and range LIDAR system, through simultaneous transmission of several lasers at different wavelengths. The lasers are detected and separated according to wavelengths in order to shorten the recording time and / or increase the point density. The recorded sensor data are weighted in a suitable manner. The probability of a new object position typically results from the mean value, variance and standard deviation of newly scanned data.

In the unpublished at the time of registration DE 10 2019 210 767 describes how several ultrasound-based moisture values can be merged with one another with the help of a weighted variance. This type of fusion is suitable for recognizing disturbances that have a similarly strong effect on sensors arranged next to one another and to greatly reduce the respective weighting of the sensor values of the sensors concerned during the fusion in order to calculate a particularly disturbance-free moisture value.

An actual risk of aquaplaning is to be distinguished from a pure wetness value of the road in that the road is wet regardless of the means of transport being used, while the risk of aquaplaning only arises from the fact that a vehicle is moving on the wet road. The risk of aquaplaning depends on various factors. The speed of the means of locomotion has an influence on the risk of aquaplaning, for example, and can easily be adjusted by a driver in order to reduce the risk. In addition, the drainage properties of the road surface and tires have a significant influence on the risk of aquaplaning. In addition, the static and dynamic tire contact forces related to the tire contact area can have an influence on the risk of aquaplaning.

It is an object of the present invention to provide a method and a device for fusing a plurality of signals from an ultrasonic sensor system of a means of locomotion, on the basis of which, in particular, aquaplaning risk information recorded by the ultrasonic sensor system can be merged. This enables inter alia to achieve a higher reliability for a respective aquaplaning risk detection method in the means of transport.

Disclosure of the invention

To determine the situation-related aquaplaning risk that is individual for the means of transport, signals that are caused, for example, by tire noise, are to be measured with ultrasonic sensors of the means of transport will. These ultrasonic sensors can be part of a park pilot system, for example. Each individual ultrasonic sensor can calculate a hydroplaning risk value based on a noise level measurement of one or more frequency ranges in such a way that a characteristic value for road wetness is determined based on the noise level measurement and, by means of this characteristic value, determines an aquaplaning risk value for the roadway wetness, taking into account the current speed of the means of transport and possibly other conditions will. These ultrasound-based aquaplaning risk values are to be merged with one another in such a way that the influence of disturbances is reduced to a minimum.

According to a first aspect of the present invention, a method for fusing a plurality of signals from an ultrasonic sensor system of a means of locomotion is proposed. The means of locomotion can be, for example, a road vehicle (e.g. motorcycle, car, van, truck) or a rail vehicle or an aircraft / airplane and / or a watercraft.

In a first step of the method according to the invention, an evaluation unit according to the invention is used to receive a first weighting factor for a first signal which represents a first hydroplaning risk value. In a second step of the method according to the invention, the evaluation unit according to the invention is used to receive a second weighting factor for a second signal which represents a second aquaplaning risk value. The first signal and the second signal can be generated by different ultrasonic sensors or by one and the same ultrasonic sensor of the means of locomotion. In addition, the first signal and the second signal can each represent different or identical frequency ranges and / or measurement times of the ultrasonic sensor system. Preferably, 2 to 8 different frequency ranges and particularly preferably 3 to 4 different frequency ranges can be detected per ultrasonic sensor. Furthermore, the first signal and the second signal can be determined on the basis of the received noise level of an ultrasonic sensor and / or diffuse floor echoes of a previously transmitted ultrasonic excitation signal. The first and second aquaplaning risk values represented by the first signal and the second signal can be determined on the basis of a suitable algorithm which is set up to determine respective values for a current aquaplaning risk on the basis of the measured values generated by the ultrasonic sensors. For this purpose, characteristic noise levels present in the measured values can be classified into degrees of wetness that correspond to them. The degrees of wetness together with other parameters, such as the current speed of the means of transport, result in hydroplaning risk values which represent the current hydroplaning risk. The first weighting factor and / or the second weighting factor can be determined in different ways depending on the application context of the method according to the invention described here. The application context should be understood to mean different levels of fusion for the respective aquaplaning risk values. In other words, the method according to the invention can be used, for example, in a first merging stage, the first signal, the second signal and, if necessary, further signals that are generated, for example, by one and the same ultrasonic sensor (e.g. in different frequency ranges of the ultrasonic sensor and / or measured at successive measurement times) sensor-related, while in a second fusion stage a plurality of sensor-related fused aquaplaning risk values of a plurality of ultrasonic sensors of the means of transport can be merged, so that a result of the second merging stage can represent a hydroplaning risk value for the entire means of transport. Alternatively, it is also conceivable to merge aquaplaning risk values from a plurality of predefined groups of ultrasonic sensors of the means of transport in the second merging stage, which can then be merged in an additional third merging stage to form an aquaplaning risk value for the entire means of transport. In addition, it is also conceivable to use further suitable fusion stages in connection with the method according to the invention. It should be noted that in the first merging stage not necessarily only aquaplaning risk values originating from one ultrasonic sensor can be merged; instead, in the first merging stage, for example, all aquaplaning risk values of a respective frequency range of a plurality of ultrasonic sensors (i.e. from all ultrasonic sensors of the means of transport or from all ultrasonic sensors) can also be merged a respective group) are merged. In addition, any further combinations of frequency ranges and / or ultrasonic sensors and / or measurement times are basically conceivable.

The result, which represents a current risk of aquaplaning, can be displayed to a driver, for example, so that if the risk of aquaplaning is too high, he can reduce the speed and thus also the risk of aquaplaning. In addition, for example, the maximum speed of the speed controller of a driver assistance function of the means of locomotion or of an autonomously driving Means of locomotion can be reduced if the risk of aquaplaning is too high and the distance to the vehicle in front increased. In addition, vehicle stabilization systems such as ESP can be prepared for the increased probability of an imminent stabilization intervention and safety systems such as belt tensioners can be activated to reduce the consequences of an accident due to aquaplaning.

bestimmt werden.The method according to the invention enables the different merging stages to be implemented, preferably on the basis of multiple implementation of the method according to the invention, with the respective results of a respective preceding merging stage being able to flow into a subsequent iteration run (i.e. the subsequent merging stage) of the method according to the invention as input variables in the form of respective aquaplaning risk values and respective weighting factors . In this context, it should be noted that the respective weighting factors flowing into the first merging stage can preferably be predefined weighting factors for respective ultrasonic sensors and / or frequency ranges of the ultrasonic sensors, which can be determined or learned, for example, during a development phase of the means of transport. The weighting factors determined in this way can, for example, be stored in a memory unit connected to the evaluation unit for information technology and can in principle represent a value for an expected quality or reliability of the respective signals. In this way, signals that are rather unsuitable for determining the aquaplaning risk (e.g. due to a fault-prone arrangement position of the respective ultrasonic sensors corresponding to the unsuitable signals) can continuously flow into the merger with a correspondingly lower weighting factor, while more suitable signals by means of a correspondingly higher weighting factor a correspondingly larger one Can influence an overall result of the aquaplaning risk assessment. In each merging step following the first merging step, no predefined weighting factors flow into the method according to the invention, since the result of the first merging step always depends on current measured values or boundary conditions and is therefore dynamic. For the sake of a simplified description, a merger of sensor-related aquaplaning risk values will be assumed in the first merging stage (ie a merging of aquaplaning risk values of a single ultrasonic sensor) without restricting the method according to the invention to this variant. Correspondingly, the sensor-related indices or designations of variables used below also do not represent a restriction to this variant, but can also be transferred and applied in a suitable manner to the above-mentioned further combination options for merging aquaplaning risk values. The respective predefined weighting factors flowing into the first merger stage can preferably be based on signal-related variances (hereinafter also referred to as "individual variances") ${\sigma }_{\mathrm{<figure-callout\; id="2"\; label="i"\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">i</figure-callout>}}^2$

to be determined.

wobei ${\sigma }_{s,r}^2$

eine sensorbezogene Varianz und ${\sigma }_i^2$

die signalbezogene Varianz repräsentieren. Der Index r kennzeichnet die Berechnung auf Basis der jeweiligen reziproken Werte.In a third step of the method according to the invention, a third weighting factor representing the reciprocal sum of the reciprocal first weighting factor and the reciprocal second weighting factor is determined by means of the evaluation unit. This can preferably be done on the basis of the following formula: $${\sigma }_{s,\mathrm{<figure-callout\; id="2"\; label="r"\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">r</figure-callout>}}^2=\frac{1}{{\sum }_i\frac{1}{{\mathrm{<figure-callout\; id="2"\; label="\sigma \; i"\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">\sigma </figure-callout>}}_i^{}}}$$

whereby ${\sigma }_{s,\mathrm{<figure-callout\; id="2"\; label="r"\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">r</figure-callout>}}^2$

a sensor-related variance and ${\mathrm{<figure-callout\; id="2"\; label="\sigma \; i"\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">\sigma </figure-callout>}}_i^{}$

represent the signal-related variance. The index r indicates the calculation based on the respective reciprocal values.

wobei µ_s den gewichteten Mittelwert und µ_i die jeweiligen Aquaplaningrisikowerte repräsentieren.In a fourth step of the method according to the invention, a weighted mean value for the first aquaplaning risk value and the second aquaplaning risk value from the sum of the respective quotients from the first aquaplaning risk value and the first weighting factor, and the second aquaplaning risk value and the second weighting factor is determined by means of the evaluation unit, the sum of these being determined Quotient is weighted by means of the third weighting factor. This can preferably be done on the basis of the following formula: $${\mu }_s={\sigma }_{s,\mathrm{<figure-callout\; id="2"\; label="r"\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">r</figure-callout>}}^2{\displaystyle \sum _i\frac{{\mu }_i}{{\sigma }_i^2}}$$

where µ _{s represents} the weighted mean and µ _i the respective aquaplaning risk values.

wobei ${\sigma }_{s,\pi }^2$

die gewichtete Varianz bezogen auf den jeweiligen Ultraschallsensor repräsentiert.In a fifth step of the method according to the invention, a weighted variance for the first aquaplaning risk value and the second aquaplaning risk value is determined by means of the evaluation unit on the basis of respective deviations of the respective aquaplaning risk values from the weighted mean, taking into account the first weighting factor, the second weighting factor and the third weighting factor. This can preferably be done on the basis of the following formula: $${\sigma }_{s,\mathrm{<figure-callout\; id="2"\; label="w"\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">w</figure-callout>}}^2={\sigma }_{s,\mathrm{<figure-callout\; id="2"\; label="r"\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">r</figure-callout>}}^2{\displaystyle \sum _i\frac{{\left({\mu }_i-{\mu }_s\right)}^2}{{\mathrm{<figure-callout\; id="2"\; label="\sigma \; i"\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">\sigma </figure-callout>}}_i^{}}}$$

whereby ${\sigma }_{s,\mathrm{<figure-callout\; id="2"\; label="\pi "\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">\pi </figure-callout>}}^2$

represents the weighted variance related to the respective ultrasonic sensor.

In a sixth step of the method according to the invention, the weighted mean value and the weighted variance are used to determine a resulting hydroplaning risk condition, the weighted mean value representing the merging of the first signal and the second signal. As described above, using the results of the sixth method step can also mean carrying out the described method again using the results of the method as input variables for the subsequent iteration of the method. Alternatively, an improved aquaplaning risk determination can already be achieved on the basis of the results of the first merger stage.

It should be pointed out that the aforementioned preferred calculation rules are not intended to represent any restriction of the method according to the invention to their specific design. Instead, it is also conceivable to use calculation rules that differ from the calculation rules mentioned above, as long as the basic principle of the method according to the invention is retained on their basis. In addition, instead of the variances used above (or the standard deviations derived therefrom), it is also conceivable to use scatter parameters which deviate therefrom in conjunction with the method according to the invention.

The subclaims show preferred developments of the invention.

wobei σ² die Gesamtvarianz repräsentiert. Die Gesamtvarianz σ² wird dann groß, wenn alle individuellen Varianzen ${\sigma }_i^2$

σ² aufgrund eines Messrauschens groß sind. Sie wird aber auch dann groß, wenn die individuellen Varianzen ${\sigma }_i^2$

σ² aufgrund des Messrauschens klein sind, während sich die Aquaplaningrisikowerte µ_i stark unterscheiden. Allerdings kann die Gesamtvarianz σ² auch klein werden, wenn einer der Messwerte stark von anderen Messwerten abweichen sollte, während für diesen Messwert eine hohe Varianz aufgrund des Messrauschens bestimmt wurde.In a preferred embodiment of the present invention, the evaluation unit is used in an additional method step to determine a total variance for the first aquaplaning risk value and the second aquaplaning risk value on the basis of the third weighting factor and the weighted variance. In a further additional method step, this total variance is subsequently used as a weighted variance in the means of transport. These process steps are preferably carried out between the respective merging stages in order, among other things, to avoid potential divisions by zero. The calculation of the total variance can preferably be based on the following formula: $${\mathrm{<figure-callout\; id="2"\; label="\sigma "\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">\sigma </figure-callout>}}^2={\sigma }_{s,\mathrm{<figure-callout\; id="2"\; label="r"\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">r</figure-callout>}}^2+{\sigma }_{s,\mathrm{<figure-callout\; id="2"\; label="w"\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">w</figure-callout>}}^2$$

where σ ^{2 represents} the total variance. The total variance σ ² becomes large if all individual variances ${\sigma }_i^{}$${}_2^{}$

σ ² are large due to measurement noise. But it will also be great if the individual variances ${\sigma }_i^{}$${}_2^{}$

σ ² are small due to the measurement noise, while the aquaplaning risk _{values µ i} differ greatly. However, the total variance σ ^{2 can} also become small if one of the measured values should deviate significantly from other measured values, while a high variance was determined for this measured value due to the measurement noise.

wobei α den vierten Gewichtungsfaktor repräsentiert. Wird der Faktor a groß gewählt, dann bewirken potentiell einwirkende Störungen eine starke Erhöhung der Gesamtvarianz σ². Dadurch kann bei einer anschließenden Fusion mehrerer Aquaplaningrisikowerte der Einfluss der jeweils gestörten Ultraschallsensoren besonders gut reduziert werden. Wird der Faktor a hingegen klein gewählt, dann bewirken zufällige Unterschiede zwischen Signalen aufgrund des Messrauschens nur eine geringe Erhöhung der Gesamtvarianz σ². Dadurch kann bei einer anschließenden Fusion mehrerer Aquaplaningrisikowerte entsprechend der Einfluss des Messrauschens reduziert werden. Um einen geeigneten Kompromiss zwischen einer guten Erkennung von Störungen und einer guten Reduzierung des Messrauschens zu erreichen, kann der Faktor a bevorzugt in einem Bereich zwischen 0,2 und 10 und insbesondere bevorzugt in einem Bereich zwischen 1 und 2 gewählt werden. Die Wahl des Faktors α kann darüber hinaus bevorzugt für jeden Ultraschallsensor individuell festgelegt werden.In a further advantageous embodiment of the present invention, the weighted variance is weighted with a fourth weighting factor before determining the total variance by means of the evaluation unit. This can preferably be done on the basis of the following formula: $${\mathrm{<figure-callout\; id="2"\; label="\sigma "\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">\sigma </figure-callout>}}^2={\sigma }_{s,\mathrm{<figure-callout\; id="2"\; label="r"\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">r</figure-callout>}}^2+\alpha {\sigma }_{s,\mathrm{<figure-callout\; id="2"\; label="\pi "\; filenames="DE102020204833A1\_0014.png"\; state="\{\{state\}\}">\pi </figure-callout>}}^2$$

where α represents the fourth weighting factor. If the factor a is chosen to be large, then potentially influencing disturbances cause a strong increase in the total variance σ ² . As a result, when several aquaplaning risk values are subsequently merged, the influence of the ultrasonic sensors that are disturbed in each case can be reduced particularly well. If, on the other hand, the factor a is chosen to be small, then random differences between signals due to the measurement noise only cause a slight increase in the total variance σ ² . As a result, the influence of the measurement noise can be reduced accordingly in the event of a subsequent merger of several aquaplaning risk values. In order to achieve a suitable compromise between a good detection of disturbances and a good reduction of the measurement noise, the factor a can preferably be selected in a range between 0.2 and 10 and particularly preferably in a range between 1 and 2. In addition, the choice of the factor α can preferably be set individually for each ultrasonic sensor.

In a further advantageous embodiment of the present invention, the first signal and the second signal are generated by a predefined group of ultrasonic sensors and a weighted mean value assigned to the group and a weighted variance assigned to the group are determined over all ultrasonic sensors of the predefined group. As described above, the ultrasonic sensors of the means of locomotion can be assigned to a plurality of groups, wherein each ultrasonic sensor of the means of locomotion can be assigned to one, several or none of the groups. For the assignment of respective ultrasonic sensors to respective groups, ultrasonic sensors that are locally adjacent to one another can preferably be selected for a respective group assignment. This can be of particular advantage since disturbances acting on the ultrasonic sensors often have a locally limited effect on adjacent ultrasonic sensors and / or on individual adjacent and / or overlapping groups. All aquaplaning risk values of groups influenced by disturbances should preferably be given a low weight when calculating the merged aquaplaning risk value across several groups. Measurements for which it is known that a high variance is to be expected due to measurement noise or interference should accordingly be weighted less when calculating the variance of the merged aquaplaning risk value. Disturbances which, on the other hand, do not lead to an increased value for the variance, but can have different effects on different ultrasonic sensors, should, however, advantageously lead to a calculation of a large value for the variance. For this reason, the calculations for the group-related aquaplaning risk values and the group-related variances can be carried out analogously to the calculation steps described above for the individual ultrasonic sensors. Preferably between 2 and 4 adjacent ultrasonic sensors can be assigned to a common group without thereby limiting the number in the sense of the method according to the invention. As already mentioned, it can also be advantageous to assign one or more ultrasonic sensors to more than one group. As a result, groups can overlap, which can increase the probability that one or more groups of the means of locomotion will remain undisturbed during disruptions that occur and thereby generate a hydroplaning risk value with a low variance. Alternatively, for example, 5 or 6 ultrasonic sensors (or more) can be assigned to a respective group, in which case each ultrasonic sensor can preferably be assigned to three groups so that the effects of disturbances can continue to be clearly delimited.

In a further advantageous embodiment of the present invention, the evaluation unit determines a weighted mean value assigned to the means of transport and a weighted variance assigned to the means of transport for the first hydroplaning risk value and for the second hydroplaning risk value, the first hydroplaning risk value and the second hydroplaning risk value having a weighted mean value and a weighted mean value Represent variance for a respective ultrasonic sensor or a weighted mean value and a weighted variance for a respective group of ultrasonic sensors of the means of locomotion. As described above, the values relating to the means of locomotion can be calculated in a manner analogous to the method according to the invention described above. The variance calculated for the means of transport can in particular be used as an indicator for the reliability of the aquaplaning risk determination for the entire means of transport.

In a further advantageous embodiment of the present invention, the weighted mean value assigned to the means of transportation and the weighted variance assigned to the means of transportation are additionally determined on the basis of a hydroplaning risk value of a sensor system that deviates from the ultrasonic sensor system of the means of transportation. This deviating sensor system can be, for example, a camera and / or a lidar sensor and / or a satellite-supported location system (due to a changed multipath propagation at different degrees of wetness), the output signals of which are preferably preprocessed in such a way that they are weighted and weighted according to the method according to the invention Variances for the respective aquaplaning risk values of the ultrasonic sensors can be merged.

In a further advantageous embodiment of the present invention, the weighted mean value and the weighted variance are additionally determined on the basis of third aquaplaning risk values and / or on the basis of a weighted mean value of the third aquaplaning risk values and a weighted variance of the third aquaplaning risk values. The third aquaplaning risk values and / or the weighted mean values and weighted variances based thereon can be determined on the basis of a further ultrasonic sensor of the means of transport. Alternatively or additionally, these can also be determined on the basis of a moisture value provided by another means of transport (“crowd sourcing”) and / or a moisture value provided by an external server. In other words, aquaplaning risk values, which were determined with the help of moisture values determined outside the means of transport, can be merged with a respective result of a respective corresponding merging stage of the means of transport, whereby a further improvement in the reliability of the aquaplaning risk determination can be achieved if necessary.

In a further advantageous embodiment of the present invention, the first weighting factor and / or the second weighting factor represent speed-dependent and / or moisture-dependent and / or dependent on thermal measurement noise values, which in particular can be determined from a predefined map. By additionally taking into account the above-mentioned influencing variables on the aquaplaning risk determination, the reliability of the aquaplaning risk determination can be further increased accordingly. The above-mentioned influencing variables can preferably be determined during a development phase and / or a training phase for the means of locomotion and, for example, stored in the memory unit connected to the evaluation unit.

According to a further advantageous embodiment of the present invention, ultrasonic sensors that are arranged at the front of the means of transport, viewed in the direction of travel of the means of transport, are each considered separately from ultrasound sensors that are arranged at the rear of the means of transport, viewed in the direction of travel of the means of transport. A front fused aquaplaning risk value is thus determined by means of the front ultrasonic sensors and a rear fused aquaplaning risk value is determined by means of the rear ultrasonic sensors. The front fused aquaplaning risk value can in particular be assigned to a front axle of the means of locomotion, the rear fused aquaplaning risk value can in particular be allocated to a rear axle of the means of locomotion. This embodiment of the invention has the advantage that, in particular, a sudden change in the wetness of the roadway can be responded to in good time. In such a case, for example if there is a large puddle of water on the roadway, the front merged aquaplaning risk value can already indicate an increased risk of aquaplaning at a point in time when the wheels of the rear axle are still on a dry road, i.e. the rear aquaplaning risk value is determined to be significantly lower than the first aquaplaning risk value.

The tire noise of the front axle can dominate the noise measured by the ultrasonic sensors. Depending on the center distance and the speed of the means of locomotion, such a change in the risk of aquaplaning has a delayed effect on the tire noises that are generated by the rear axle and measured with the rear ultrasonic sensors. So that this delay in the merger of the first and second aquaplaning risk values of the front and rear ultrasonic sensors is not mistakenly considered to be a malfunction, the sensor values of the front ultrasonic sensors can be delayed by a dead time element, the dead time or the delay being selected in particular as a function of the current speed will. As a result of this delay, a sudden change in the risk of aquaplaning acts synchronously on the rear and front delayed ultrasonic sensor values, so that during such a change in the risk of aquaplaning, disturbances have less of an influence on the fusion.

In a particularly advantageous embodiment of the present invention, for a means of locomotion with at least four wheels, individual fused aquaplaning values are determined for each wheel. This information can be used, for example, by an ESP control of the means of transport, so that the driving dynamics control can better estimate the coefficient of friction of the individual wheels and thus the developing vehicle dynamics.

In order to appropriately output a hydroplaning risk for a driver or for the cruise control, the merged hydroplaning risk values can be filtered over a longer period of time. The reaction time of a driver to information about the risk of aquaplaning can be, for example, 0.2 to 1 second. In order to prevent incorrect information due to measurement noise, it can therefore be beneficial to filter the aquaplaning risk values, e.g. over 0.2 to 1 second. The fusion equations described above can be used to filter a plurality of aquaplaning risk values determined in succession in time. For example, a fusion filter with an infinite impulse response (IIR filter) can be implemented by weighting the output of the fusion filter or providing it with an offset and returning it as an input. Alternatively, a fusion filter with a finite impulse response (FIR filter) can be implemented by fusing a certain number of the most recently calculated input values with one another.

According to a second aspect of the present invention, a device for fusing a plurality of signals from an ultrasonic sensor system of a means of locomotion is proposed. The device comprises an evaluation unit with a data input and a data output, wherein the evaluation unit can be connected by means of the data input and by means of the data output in particular to an on-board network of the means of locomotion. Furthermore, the evaluation unit can be connected in terms of information technology to an internal and / or external storage unit in which data received and / or calculated by the evaluation unit can be stored for subsequent processing. The evaluation unit can be designed, for example, as an ASIC, FPGA, processor, digital signal processor, microcontroller, or the like, and can be set up in conjunction with a computer program to carry out method steps according to the invention described above.

The evaluation unit is set up in connection with the data input, a first one To receive a weighting factor for a first signal which represents a first hydroplaning risk value and to receive a second weighting factor for a second signal which represents a second hydroplaning risk value. Furthermore, the evaluation unit is set up to determine a third weighting factor representing the reciprocal sum of the reciprocal first weighting factor and the reciprocal second weighting factor, a weighted mean value for the first aquaplaning risk value and the second aquaplaning risk value from the sum of the respective quotients of the first aquaplaning risk value and the first weighting factor, and to determine the second hydroplaning risk value and the second weighting factor, the sum of these quotients being weighted by means of the third weighting factor and a weighted variance for the first hydroplaning risk value and the second hydroplaning risk value based on respective deviations of the respective hydroplaning risk values from the weighted mean value, taking into account the first weighting factor to determine the second weighting factor and the third weighting factor. In connection with the data output, the evaluation unit is set up to use the weighted mean value and the weighted variance to determine a resulting hydroplaning risk state, the weighted mean value representing the merging of the first signal and the second signal. In this context, it should be noted that the result of the method according to the invention, as described above, can be fed to one or more further merging stages in the form of a further iteration of the method according to the invention, whereby the result output at the data output via the data input of the evaluation unit within the evaluation unit itself can be further processed. In such a case, the data output and the data input can be understood as writing and reading memory access within the evaluation unit itself. Alternatively or additionally, a respective result of the method according to the invention can also be transmitted to one or more recipients of the result by means of the data output via the on-board network of the means of locomotion.

Figure list

• 1 ein Flussdiagramm veranschaulichend Schritte eines Ausführungsbeispiels eines erfindungsgemäßen Verfahrens;
• 2 eine schematische Übersicht einer erfindungsgemäßen Vorrichtung in Verbindung mit einem Fortbewegungsmittel;
• 3 eine grafische Darstellung eines Rechenbeispiels für die Fusion mehrerer Aquaplaningrisikowerte ohne äußere Störung, und
• 4 eine grafische Darstellung eines Rechenbeispiels für die Fusion mehrerer Aquaplaningrisikowerte mit äußerer Störung.

Exemplary embodiments of the invention are described in detail below with reference to the accompanying drawings. Show:

• 1 a flow chart illustrating steps of an exemplary embodiment of a method according to the invention;
• 2 a schematic overview of a device according to the invention in connection with a means of locomotion;
• 3 a graphic representation of a calculation example for the fusion of several hydroplaning risk values without external disturbance, and
• 4th a graphic representation of a calculation example for the fusion of several hydroplaning risk values with external disturbance.

Embodiments of the invention

1 FIG. 11 shows a flowchart illustrating steps of an exemplary embodiment of a method according to the invention for fusing a plurality of signals from an ultrasonic sensor system of a means of locomotion. In step 100 a first weighting factor for a first signal, which represents a first hydroplaning risk value, is received by means of an evaluation unit according to the invention, which is a microcontroller here. The first signal is generated on the basis of a measurement in a first frequency range of an ultrasonic sensor of the means of locomotion. In step 200 a second weighting factor for a second signal, which represents a second hydroplaning risk value, is received by means of the evaluation unit according to the invention. The second signal is generated on the basis of a measurement in a second frequency range of the ultrasonic sensor of the means of locomotion, which differs from the first frequency range. In step 300 a third weighting factor representing the reciprocal sum of the reciprocal first weighting factor and the reciprocal second weighting factor is determined. In step 400 a weighted average for the first hydroplaning risk value and the second hydroplaning risk value based on the ultrasonic sensor is determined from the sum of the respective quotients from the first hydroplaning risk value and the first weighting factor, and the second hydroplaning risk value and the second weighting factor, the sum of these quotients being weighted by means of the third weighting factor will. In step 500 a weighted variance related to the ultrasound sensor for the first hydroplaning risk value and the second hydroplaning risk value is determined on the basis of respective deviations of the respective hydroplaning risk values from the weighted mean, taking into account the first weighting factor, the second weighting factor and the third weighting factor. In step 600 the sensor-related weighted mean value and the sensor-related weighted variance are used to determine a resulting aquaplaning risk state. For this purpose, a weighted sensor-related total variance is determined on the basis of the sensor-related weighted variance, on the basis of the third weighting factor and on the basis of a fourth weighting factor. The sensor-related weighted mean and the sensor-related total variance are then used to calculate a plurality of weighted mean values and weighted total variances of a plurality of To fuse ultrasonic sensors analogously to the method described above to form respective values related to the means of locomotion.

2 shows a schematic overview of a device according to the invention in connection with a means of locomotion 80 . The means of transportation 80 comprises an evaluation unit according to the invention 10 which here is a microcontroller and which by means of a data input 12th and by means of a data output 14th with an on-board network of the means of locomotion 80 is information technology connected. The means of transportation 80 also includes a first ultrasonic sensor 30th , a second ultrasonic sensor 35 and a plurality of other ultrasonic sensors 50 , which in each case via the vehicle electrical system of the means of transport 80 information technology with the evaluation unit 10 are connected. A first subset of the ultrasonic sensors 30th , 35 and 50 is a first ultrasonic sensor group 40 and a second subset of the ultrasonic sensors 30th , 35 and 50 is a second ultrasonic sensor group 45 assigned. The evaluation unit 10 is also IT with an external storage unit 20th connected, in which by the evaluation unit 10 received and processed data are stored. It is also in the storage unit 20th a computer program is stored which is set up to carry out method steps according to the invention described above. The evaluation unit 10 is also information technology with a camera 70 of the means of transport 80 Information technology connected and set up, an output signal from the camera 70 with the respective measurement signals from the ultrasonic sensors 30th , 35 and 50 for determining the risk of aquaplaning in the vehicle 80 to merge in a suitable manner using the computer program. The evaluation unit is also 10 by means of the data output 14th with a control unit 60 of the means of transport 80 information technology connected, which is set up by the evaluation unit 10 to the control unit 60 transferred result of the method according to the invention in the means of transport 80 to use.

3 shows a diagram 700 multiple measurements 1-4 an aquaplaning risk value, as well as a resulting merged aquaplaning risk value 5 without external interference according to an embodiment of the invention. The aquaplaning risk value is plotted on the x-axis and the associated probability density on the y-axis. The aquaplaning risk values are given as percentages in this example.

In the calculation example shown, the merged aquaplaning risk value is 5 the standard deviations caused by the measurement noise for the first signal have been determined to be 25% 1 17.5%, the values for the second signal 2 and the third signal 3 at 25% each and the value of the fourth signal 4th elected to 37.5%. The mean value of the fused aquaplaning risk value without disturbance (corresponds to the maximum value of the probability density) corresponds to approx. 25%. The standard deviation α (distance between the point of inflection and the high point) corresponds to approx. 10%.

4th shows a diagram 800 multiple measurements 1'-4 ' an aquaplaning risk value, as well as a resulting merged aquaplaning risk value 5 ' with external disturbance according to a further embodiment of the invention. The aquaplaning risk value is plotted on the x-axis and the associated probability density on the y-axis. The aquaplaning risk values are also given as percentages in this example.

The disturbance is chosen so that it is for the first signal 1' an increase of 75 percentage points, the signals 2 ' and 3 ' Increase of 50% -points compared to the undisturbed measurement ( 3 ) and for the value 4 ' causes an increase of 25 percentage points. The mean of the fused aquaplaning risk value 5 ' without taking the disturbance into account (this corresponds to the maximum value of the probability density function) is approx. 60%. The deviation from the actually selected aquaplaning risk value of 25% is very large, but the calculated standard deviation of approx. 20% is also significantly larger than for the merged value 5 without interference. In the event of a subsequent fusion of the disturbed aquaplaning risk value with an aquaplaning risk value of an undisturbed ultrasonic sensor, the value of the undisturbed ultrasonic sensor is weighted significantly more strongly due to its lower standard deviation, whereby the influence of the disturbed ultrasonic sensor can be significantly reduced.

• Die Fusion mehrerer Werte eines Ultraschallsensors mit Hilfe der gewichteten Varianz hat den Vorteil, dass Störungen, die sich unterschiedlich stark auf die Aquaplaningrisikowerte auswirken zu einer hohen berechneten Varianz für den fusionierten Wert führen und dadurch der fusionierte Wert des Ultraschallsensors bei der weiteren Fusion mit Werten von weiteren Ultraschallsensoren, die nicht gestört sind, weniger stark gewichtet werden. Dadurch wird die Fusion unempfindlich gegenüber Störungen.

The invention achieves the following advantages in particular:

• The merger of several values of an ultrasonic sensor with the help of the weighted variance has the advantage that disturbances which have different effects on the aquaplaning risk values lead to a high calculated variance for the merged value and thus the merged value of the ultrasonic sensor when further merging with values of other ultrasonic sensors that are not disturbed are weighted less heavily. This makes the fusion insensitive to disturbances.

The fusion initially within an ultrasonic sensor and the subsequent fusion of several neighboring ultrasonic sensors within a group and the subsequent fusion several groups has the advantage that disturbances that do not have different effects on the aquaplaning risk values of individual ultrasonic sensors, but instead have different effects on neighboring ultrasonic sensors, lead to a high calculated variance for the merged value of the zone and thus the merged value the zone is weighted less heavily in the further fusion with values of zones that are not disturbed. As a result, the fusion becomes insensitive to disturbances which could not or not sufficiently be recognized by the affected ultrasonic sensors themselves.

The quality of the fused aquaplaning risk value can be read from the calculated variance.

The calibration of the function can be automated to a high degree. In this way, the parameters of the aquaplaning risk value and the variance maps can be calculated from measurements of a few driving tests. Very different and large disturbances have a small influence on the calculated hydroplaning risk value and the hydroplaning risk can nevertheless be calculated dynamically and precisely.

QUOTES INCLUDED IN THE DESCRIPTION

This list of the documents listed by the applicant was generated automatically and is included solely for the better information of the reader. The list is not part of the German patent or utility model application. The DPMA assumes no liability for any errors or omissions.

Patent literature cited

• DE 102016221809 A1 [0004]
• DE 102018118679 A1 [0005]
• DE 102019210767 [0006]

Claims (13)

A method for fusing a plurality of signals from an ultrasonic sensor system (30, 35) of a means of locomotion (80), comprising the steps: • receiving (100) a first weighting factor for a first signal which represents a first hydroplaning risk value, • receiving (200) a second weighting factor for a second signal, which represents a second hydroplaning risk value, • Determination (300) of a third weighting factor representing the reciprocal sum of the reciprocal first weighting factor and the reciprocal second weighting factor, • Determination (400) of a weighted mean value for the first hydroplaning risk value and the second hydroplaning risk value from the sum of the respective quotients from the first hydroplaning risk value and the first weighting factor, and the second hydroplaning risk value and the second weighting factor, the sum of these quotients being weighted by means of the third weighting factor , • determining (500) a weighted variance for the first hydroplaning risk value and the second hydroplaning risk value on the basis of respective deviations of the respective hydroplaning risk values from the weighted mean value, taking into account the first weighting factor, the second weighting factor and the third weighting factor, and Using (600) the weighted mean value and the weighted variance to determine a resulting hydroplaning risk condition, the weighted mean value representing the merging of the first signal and the second signal. Procedure according to Claim 1 , wherein the first signal and the second signal • by different ultrasonic sensors (30, 35), or • by one and the same ultrasonic sensor (30), the ultrasonic sensors (30, 35) of the means of locomotion (80) are generated. Method according to one of the preceding claims, wherein the first signal and the second signal are each different or identical • Frequency ranges, and / or • represent measurement times of the ultrasonic sensors (30, 35), and / or on the basis of • Noise levels, and / or • Diffuse ground echoes of a previously transmitted ultrasonic excitation signal are determined. A method according to any preceding claim further comprising • Determination of a total variance for the first aquaplaning risk value and the second aquaplaning risk value on the basis of the third weighting factor and the weighted variance, • then using the total variance as a weighted variance. Procedure according to Claim 4 , the weighted variance being weighted with a fourth weighting factor before the total variance is determined. The method according to any one of the preceding claims, wherein • the first signal and the second signal are generated by a predefined group (40) of ultrasonic sensors (30, 35), and • a weighted mean value assigned to the group (40) and a weighted variance assigned to the group (40) are determined over all ultrasonic sensors (30, 35) of the predefined group (40). Procedure according to Claim 6 wherein a front fused aquaplaning risk value is determined by means of ultrasonic sensors which are arranged at the front of the means of transport (80) as seen in the direction of travel of the means of transport (80), and by means of ultrasound sensors which, as seen in the direction of travel of the means of transport (80) are at the rear of the means of transport (80) a rear fused aquaplaning risk value is determined, the front fused aquaplaning risk value being assigned in particular to a front axle of the means of locomotion (80) and the rear fused aquaplaning risk value in particular being allocated to a rear axle of the means of locomotion (80). Procedure according to Claim 6 or 7th , wherein a front hydroplaning risk value is fused with a rear hydroplaning risk value, the signals that are generated by means of ultrasonic sensors that are arranged in the direction of travel of the means of locomotion (80) at the front of the means of locomotion before the fusion, in particular by a dead time element, with respect to signals that are generated by means of ultrasonic sensors which are arranged at the rear of the means of transport as seen in the direction of travel of the means of transport (80), the delay being selected in particular as a function of the current speed of the means of transport (80). Method according to one of the preceding claims, wherein • a weighted mean value assigned to the means of transport (80) and a weighted variance assigned to the means of transport (80) are determined for the first hydroplaning risk value and for the second hydroplaning risk value, and • the first hydroplaning risk value and the second hydroplaning risk value ◯ represent a weighted mean value and a weighted variance for a respective ultrasonic sensor (30, 35), or ◯ represent a weighted mean value and a weighted variance for a respective group (40) of ultrasonic sensors (30, 35) of the means of locomotion (80). Procedure according to Claim 7 , wherein the weighted mean value assigned to the means of transportation (80) and the weighted variance assigned to the means of transportation (80) are additionally determined on the basis of a hydroplaning risk value of a sensor system (70) that deviates from the ultrasonic sensor system (30, 35) of the means of transportation (80). Method according to one of the preceding claims, wherein the weighted mean and the weighted variance are additionally based on • third hydroplaning risk values, and / or • a weighted mean of the third aquaplaning risk values and a weighted variance of the third aquaplaning risk values are determined. Method according to one of the preceding claims, wherein the first weighting factor and / or the second weighting factor • speed-dependent, and / or • Depending on the degree of moisture, and / or • represent values dependent on a thermal measurement noise, and / or • be determined from a predefined map. Device for fusing a plurality of signals from an ultrasonic sensor system of a means of locomotion, comprising: • an evaluation unit (10), • a data input (12), and • a data output (14), the evaluation unit (10) being set up, • in connection with the data input (12) ◯ to receive a first weighting factor for a first signal which represents a first hydroplaning risk value, and ◯ to receive a second weighting factor for a second signal, which represents a second hydroplaning risk value, • to determine a third weighting factor representing the reciprocal sum of the reciprocal first weighting factor and the reciprocal second weighting factor, • to determine a weighted mean value for the first hydroplaning risk value and the second hydroplaning risk value from the sum of the respective quotients from the first hydroplaning risk value and the first weighting factor, and the second hydroplaning risk value and the second weighting factor, the sum of these quotients being weighted using the third weighting factor, • to determine a weighted variance for the first aquaplaning risk value and the second aquaplaning risk value on the basis of respective deviations of the respective aquaplaning risk values from the weighted mean, taking into account the first weighting factor, the second weighting factor and the third weighting factor, and • to use the weighted mean value and the weighted variance in connection with the data output (14) to determine a resulting hydroplaning risk condition, the weighted mean value representing the merging of the first signal and the second signal.

Images