DE102019212877A1 - LiDAR system as well as vehicle and method for detecting weather conditions - Google Patents

Abstract

A LiDAR system (1) is shown, which is set up to determine weather conditions in an environment of the LiDAR system (1). The LiDAR system (1) is set up to determine information about the weather conditions based on a statistical evaluation of signal intensities. Furthermore, a vehicle (2) with such a LiDAR system (1) is shown, the LiDAR system (1 ) is operatively connected to the vehicle (2). Furthermore, a method for detecting weather conditions in an environment of a LiDAR system (1) is shown. The method comprises the step of determining information about the weather conditions by the LiDAR system (1) based on a statistical evaluation of signal intensities.

Description

The present invention relates to a LiDAR system which is set up to determine weather conditions in an environment of the LiDAR system, as well as a vehicle with such a LiDAR system and a method for detecting weather conditions in an environment of a LiDAR system.

State of the art

A LiDAR system (not exclusively, but also automotive LiDAR) can be based on various measurement methods. One possible approach would be a so-called “direct time-of-flight” or dToF approach, in which pulsed laser light is emitted. The depth information here is the time it takes a laser pulse to get to a target and back to the detector. A sufficiently large pulse power or a large signal-to-noise ratio (SNR) in the receive path is decisive for the success of this approach.

Atmospheric weather events, such as rain, snow and fog, are an important influencing variable for the pulse power or the SNR in the detection path, which significantly influences the performance of a LiDAR system. These can significantly reduce the range of vision through, for example, scattering and absorption of laser light and thus make it more difficult to recognize objects. For the design of a LiDAR system, it is therefore important to evaluate the influence of atmospheric weather events on the system performance in order to recognize system degradation in good time and to report this to a system interface as an error message, for example.

According to the current state of the art, the following is particularly known from the automotive sector: Qualitative measurement of rain events using LEDs and detectors in the windshield, indirect estimation of visibility in fog using driver assistance sensors (camera, LiDAR, ...), for example due to laser pulse broadening, radio / TMC-based fog and rain warning (weather report supplied by measuring stations, meteorological data sources), manual testing by the driver (human visibility estimation) or car-to-X (C2X) communication.

The DE 69 933 932 T2 describes a LiDAR system as a terrain sensor that records weather information in addition to terrain information in certain scans. In particular, aircraft weather radar used to obtain off-road radar reflectance data is also disclosed.

The DE 20 2016 103 719 U1 discloses a door assistant system for a vehicle. The vehicle can have a vision module with a LiDAR system. The door assistant system can use a global positioning system to identify at least one weather condition, particularly wind speeds, in which the door is likely to be unexpectedly repositioned due to a gust of wind.

The EP 2 820 632 A1 describes a permanently mounted LiDAR system, for example on a bridge, for monitoring multiple lanes. The system can detect the presence of fog and snow and thus provide weather information.

The EP 3 206 046 A1 teaches a LiDAR system with improved detection of rain, snow, fog and smoke and can provide information about current weather conditions by means of pattern recognition based on low frequency signals and peaks.

From the EP 3 284 646 A1 a vehicle with a LiDAR sensor is known. An area condition detection device is set up to receive weather information (such as rain, amount of rain, fog and degree of fog) and traffic information from the outside in order to determine areas which are unsuitable for autonomous driving.

Disclosure of the invention

According to the invention, a LiDAR system is made available which is set up to determine information about the weather conditions based on a statistical evaluation of signal intensities.

Advantages of the invention

The LiDAR system according to the invention has the advantage that it allows the information about the weather conditions to be determined simply based on a statistical evaluation of the signal intensities without having to rely on further systems or manual support. An existing LiDAR system in a vehicle can take on a further function and derive information about the weather conditions from measurement data.

To determine the weather conditions, the LiDAR system preferably uses signals from a specific area (with unchanged conditions, for example from a reference target).

A preferred LiDAR system comprises a light source, in particular a laser source, for scanning the environment with a laser beam and a detector for recording that from the environment reflected laser beam. The LiDAR system usually allows depth information about the environment to be obtained from the laser beam sent and received again. It is preferred that the LiDAR system has a processor unit with one or more processors and a memory that is connected to the processor unit in order to carry out the statistical evaluation. In embodiments it is provided that the LiDAR system is set up to determine the pulse energy of a received beam of the LiDAR system in order to carry out the statistical evaluation thereon. The pulse energy of a transmission beam of the LiDAR system is preferably constant, in particular over a statistical evaluation.

The LiDAR system is preferably arranged to be movable. This means in particular that the LiDAR system can be mounted on a vehicle. In some embodiments, however, the LiDAR system is arranged statically.

In particular, this means that the LiDAR system can be mounted on a road bridge, such as a motorway bridge, or a hard shoulder post.

Preferred information about the weather conditions is the type of weather condition, in particular snow, rain and fog, and the intensity of the weather condition, preferably in discrete steps. It is preferred that the LiDAR system is arranged in such a way that part of the field of view that is used to determine the weather conditions permanently looks at a static target at a fixed distance from the sensor. The target is preferably a reference target attached to the vehicle in the immediate field of vision or also the road. Precipitation detection / classification methods and corresponding systems can thus be provided, in particular based on statistical evaluation of the signal intensities of LiDAR echo pulses at a static distance to an object that is consistently reflective over the evaluation period.

In particular, the LiDAR system is set up for the evaluation of dToF-LiDAR signals with regard to atmospheric weather events, preferably in motion. For this purpose, a determination of statistical fluctuations in a pulse energy (or amplitude) of the LiDAR system is preferably provided. Some embodiments provide that the LiDAR system is set up to uniquely identify a rain event. It is preferred that the LiDAR system is set up to determine a rain rate, in particular in several classes, such as "very much" ... "very little", preferably in 5 stages. Furthermore, the disclosed LiDAR system is particularly preferably set up to distinguish rain from fog. These and other preferred features are explained in more detail below.

In embodiments it is provided that the LiDAR system is set up to determine a pulse energy of a reception beam of the LiDAR system in order to determine a type of weather condition as the information about the weather conditions. This means that the type of weather condition in embodiments of the invention represents special information about the prevailing weather conditions, which can be determined by the LiDAR system. It is preferred that the LiDAR system determines the pulse energy as a function of time. One possible type of weather condition that can be determined is rain. Other possible types of weather conditions that can be determined are snow or fog. Rain is a statistically inhomogeneous medium. Unlike fog, which consists of the finest, micrometer-sized droplets and thus acts as a homogeneous medium on the laser beam, rain imposes statistics on the signal due to its comparatively large droplet size and the relatively large distance between the droplets. A number of raindrops, which are located in the cross section of the laser beam of the LiDAR system, varies from pulse to pulse. Correspondingly, the measured pulse energy also varies as a function of time from pulse to pulse, since a slightly different attenuation of the signal occurs for each implementation. This attenuation, also called extinction α, is made up of the scattering at each drop and the absorption when passing through each individual drop. It is preferably provided that the LiDAR system is set up to measure a predetermined number of pulses as a function of time and to create a histogram of the pulse energy from the measurements of the pulses. The histogram preferably has the pulse energy on an X-axis and the assigned absolute frequency of pulses on a Y-axis.

It is preferred that the LiDAR system is set up to determine an expected value and its standard deviation of the pulse energy in order to determine the type of weather condition. A particularly meaningful statistical evaluation of the signal intensities can thus be carried out. Expected value and standard deviation can preferably be obtained from the histogram. It is preferred that the LiDAR system is set up to calculate the standard deviation from the expected value from a scatter of the pulse energy around the expected value.

In some embodiments, the LiDAR system is set up to recognize a first expected value as an indicator for a first weather condition intensity and a second expected value, which is smaller than the first expected value, as an indicator for a second Detect weather condition intensity. It is thus possible to determine different weather condition intensities on the basis of different expected values solely on the basis of the statistical evaluation. For example, the first expected value can indicate light rain and the second expected value can show comparatively heavy rain. In heavy rain, the mean or expected value of the pulse energies is lower than in comparatively light rain due to the higher extinction. The standard deviation, on the other hand, increases in comparison with heavy rain, since a higher variation in energies is to be expected with heavy rain than with comparatively light rain. However, the first expected value can also indicate light fog, while the second expected value, on the other hand, shows comparatively dense fog. In embodiments, the LiDAR system can have a look-up table in a memory of the LiDAR system, compare measured expected values with typical expected values in the look-up table and thus determine the prevailing weather condition intensities. It is preferred that the distance dependency can be compensated for when the sensor or objects are moving. This is preferably 1 / r ² if the observed area always remains larger than one pixel of the receiving path. In this way, an intensity fluctuation due to a change in the orientation of a reflection surface can still remain.

The LiDAR system is preferably set up to recognize a standard deviation from the expected value that is greater than a predefined threshold value as an indicator for a first type of weather condition and a standard deviation from the expected value that is smaller than the predefined threshold value as an indicator for a second type of weather condition to be recognized. A preferred first type of weather condition is rain, for example. A preferred second type of weather condition is fog, for example. Another preferred type of weather condition, which can be the first or second type of weather condition, is snow. Two types of weather conditions, for example fog and rain, can cause the same mean attenuation in the histogram, but cause a very different spread, i.e. standard deviation. In the case of fog, the standard deviation can be significantly smaller than in the case of rain. In fog, the standard deviation is only caused by intrinsic time-dependent pulse-to-pulse fluctuations in the light source of the LiDAR system. The look-up table of the LiDAR system may contain predetermined standard deviations associated with corresponding types of weather conditions. The LiDAR system can be set up to determine the standard deviation of the pulse energy by statistical evaluation, to compare the determined standard deviation with the predetermined standard deviations in the look-up table and thus to determine the type of weather condition.

The LiDAR system can be set up to carry out a first method for statistical evaluation of the signal intensities. The first method can have the step of an offset correction. The offset correction preferably includes subtracting a possible offset in the measured intensity signal of the received pulse. After the offset correction, a step of determining the pulse energy can be provided. This step can include determining the pulse energy of each pulse by integrating the signal over time. Thereafter, a step of determining the standard deviation of the resulting distribution of pulse energies can be provided. In some embodiments, the LiDAR system can then optionally provide a reference of the standard deviation to the mean pulse energy of the pulses in the time window considered. In this way, a relative signal fluctuation can be determined. Thereafter, a step can be provided to use the relative signal fluctuation as a measure for the information about the weather conditions, that is to say a measure for, in particular, the type of weather condition or the intensity of the weather condition.

The first method described can, however, have the disadvantage that the mean pulse energy, which decreases during a rain or fog event, results in a distortion of the signal statistics if the time interval used to determine the signal statistics is correspondingly long. If the mean value shifts over time (for example due to the onset of rain or a change in the reflective properties of the target in a realistic scene, for example in an inhomogeneous environment, for example during automated driving in an urban environment), the pulse energies are no longer with a simple statistical evaluation distributed around a constant, rather than a varying mean value. The resulting standard deviation is larger than the intensity fluctuations induced by the rain. Although this can be compensated for by shorter evaluation times, the statistical significance then also suffers, since fewer pulses flow into the statistics due to the shorter observation interval.

wobei

die Pulsenergie des i-ten Pulses ist. Zusätzlich kann die hinter der Summe stehende Differenz der Pulsenergien in diesem Schritt in manchen Ausführungsformen auch quadriert werden: $$S={\displaystyle \sum _{i=1}^N{\left({\xi }_{i+\perp }-{\xi }_i\right)}^2}$$

The LiDAR system can therefore be set up to carry out a second method for statistical evaluation of the signal intensities. The second method can be preferred over the first method in certain scenarios described above by way of example. The second method is insensitive to a change in the mean pulse energy. The second method may include the step of subtracting a possible offset of the measured pulse, i.e. an offset correction. The pulse energy of each pulse can then be determined by integrating the signal over the course of time. The amount of the difference in pulse energies from one pulse to the next can then be determined for a block of N pulses, N N 2, and a subsequent summation, in particular according to the formula $$\mathrm{S.}={\displaystyle \sum _{i=1}^N\left|{\xi }_{i+\perp }-{\xi }_i\right|},$$

in which

is the pulse energy of the i-th pulse. In addition, the difference in the pulse energies behind the sum can also be squared in this step in some embodiments: $$\mathrm{S.}={\displaystyle \sum _{i=1}^N{\left({\xi }_{i+\perp }-{\xi }_i\right)}^2}$$

wobei n(t) die schnellen Puls-zu-Puls-Energiefluktuationen sind und s₀(t) der zeitlich langsam absinkende Mittelwert. Die zeitliche Ableitung ist nun: $$\frac{d\xi }{dt}=\frac{dn\left(t\right)}{dt}+\frac{d{s}_0\left(t\right)}{dt}$$

The procedure can be analogous to the integration with continuous signals. Let the time-varying pulse energy be defined by $$\xi \left(t\right)=n\left(t\right)+s_0\left(t\right),$$

where n (t) is the rapid pulse-to-pulse energy fluctuations and s ₀ (t) is the mean value, which decreases slowly over time. The time derivative is now: $$\frac{d\xi }{dt}=\frac{dn\left(t\right)}{dt}+\frac{d{s}_0\left(t\right)}{dt}$$

Then the resulting integral would be: $${\displaystyle {\int }_{t1}^{\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="DE102019212877A1\_0012.png"\; state="\{\{state\}\}">t</figure-callout>}2}\left|\frac{dn\left(t\right)}{dt}+\frac{d{s}_0\left(t\right)}{dt}\right|}dt\sim {\displaystyle {\int }_{t1}^{\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="DE102019212877A1\_0012.png"\; state="\{\{state\}\}">t</figure-callout>}2}\left|\frac{dn\left(t\right)}{dt}\right|dt}$$

The contribution of the second term to the integral is negligibly small, since the mean pulse energy only changes very slowly during the integration time.

The length of the integration time is determined by the quantity N, which specifies how many pulses should be added. If the integration time or the N is sufficiently large, the integral or the above-mentioned sum is dominated by the fast-moving pulse-to-pulse energy fluctuations n (t). The slowly changing mean value is no longer noticeable. A later step can then, optionally, include a relation of the result from the previous step to the mean pulse energy of the pulses in the considered time window. The relative signal fluctuation (fluctuation) determined in this way can finally also be used by the LiDAR system as the measure for the information about the weather conditions, preferably the weather condition intensity or the weather condition type, in particular the precipitation rate.

The LiDAR system can be set up to determine changes in the received pulse energy of the LiDAR system in order to determine the information about the weather conditions therefrom. In particular, the LiDAR system can be set up to determine a change in the type of weather condition and / or a change in the intensity of the weather condition from the changes in the received pulse energy. For example, the LiDAR system can in particular be set up to determine a change in the intensity of the weather conditions, preferably the onset of rain, from a time profile of an average pulse energy. With the onset of precipitation, such as rain, fog or snow, the pulse energy of the received light beam decreases with a constant transmission power of the light source. When precipitation ceases, such as rain, fog or snow, the pulse energy of the received light beam increases with constant transmission power of the light source. This means that additional precipitation sensors can be superfluous. The LiDAR system is therefore also set up in particular to determine the changes in the pulse energy as a function of time and to determine whether an average pulse energy decreases or increases as a function of time in order to determine the change in the intensity of the weather conditions therefrom. For example, cessation of precipitation, such as rain, fog or snow, can be determined if the mean pulse energy increases as a function of time. For example, the onset of precipitation, such as rain, fog or snow, can be determined if the mean pulse energy decreases as a function of time. The increase or decrease in the mean pulse energy is caused by the decrease or increase in precipitation particles in the ambient air. The LiDAR system is therefore preferably set up to detect the increase or decrease in precipitation particles in the environment.

The LiDAR system is set up in particular to determine pulse energy fluctuations around the mean value of the received pulse energy as a function of time in order to determine the information about the weather conditions therefrom. In this way, the change in the intensity of the weather conditions can preferably be determined. While the mean pulse energy drops over time when it starts to rain, for example, the relative statistical fluctuation of the pulse energies caused by the rain (relative to the mean value in the respective time interval considered) or the standard deviation, based on a base level, increases with the rain rate. Basic level means that deviations from the average pulse energy value caused only by intrinsic pulse energy fluctuations, which, for example, are generated by the light source (or the laser driver) itself and are always present. The LiDAR system can be set up to carry out an offset adjustment (subtracting the base level) in order to obtain a signal which can be calibrated to the rain rate or to a rain classification. Information about the weather conditions can thus also be determined from the pulse energy fluctuations around the mean value, such as in particular the onset or cessation of rain or snow. This also makes additional precipitation sensors superfluous. The pulse energy fluctuations around the mean pulse energy can preferably be determined from the above-mentioned histogram.

It is preferred that the LiDAR system is set up to distinguish fog from rain, in particular from the strength of fluctuations around the mean value of the received pulse energy over time. In embodiments, however, the LiDAR system can also be set up to distinguish fog from snow or also snow from rain. While the fluctuations in the pulse energy around the mean pulse energy clearly increase and are well above the base level in the case of rain, a fog event shows no change in the relative pulse energy fluctuations, but only a decrease in the mean pulse energy due to damping. The course of the fluctuations around the mean pulse energy remains at the basic level over time. The LiDAR system can therefore be set up to distinguish fog from rain by determining fog as the weather condition type if the relative pulse energy fluctuations fall below a predetermined fluctuation threshold value as a function of time, and / or by determining rain as the weather condition type if the relative pulse energy fluctuations are time-dependently the specified Exceed fluctuation threshold. A preferred measure for the fluctuation is the size "S" from the above formula. The fluctuation threshold value can be established, for example, by appropriate calibration of the sensor by means of a measurement at different rain rates. In embodiments it is provided to define the threshold value with a calibration of the system, ie for a certain rain event with a defined rain rate (or several defined rain events) to calculate the fluctuations "S" from the measurement data (according to the above formula) and then to assign the threshold value so that the rain just measured is identified as rain of a certain strength.

According to the invention, a vehicle is also made available with which a LiDAR system of the type mentioned is operationally connected, which is set up to determine the information about the weather conditions based on a statistical evaluation of signal intensities.

The vehicle according to the invention has the advantage that it allows the information about the weather conditions to be determined simply on the basis of a statistical evaluation of the signal intensities without having to rely on further systems or manual support. The LiDAR system that is often found in vehicles nowadays can take on a further function and derive information about the weather conditions from measurement data.

To determine the weather conditions, the LiDAR system preferably uses signals from a specific area (with unchanged conditions, for example from a reference target).

A preferred vehicle is a road vehicle. Preferred road vehicles are motor vehicles, especially passenger cars, trucks, buses and motorcycles. Alternative vehicles are aircraft, particularly airplanes, spaceships, and helicopters. Other alternative vehicles are watercraft, in particular ships.

According to the invention, a method for recognizing weather conditions in an environment of a LiDAR system is also provided, which includes the step of determining the information about the weather conditions by the LiDAR system based on a statistical evaluation of signal intensities.

The method according to the invention has the advantage that it allows the information about the weather conditions to be determined simply on the basis of a statistical evaluation of the signal intensities without having to rely on further systems or manual support. The LiDAR system often found in vehicles nowadays can take on a further function and derive information about the weather conditions from the measurement data that is already available.

Preferred process steps follow from the above description of the LiDAR system mutatis mutandis. The advantages and features are not repeated at this point, but they also apply to the method according to the invention. In particular, but not exclusively, the method according to the invention can comprise the above-mentioned steps of the first method and / or the second method. The order of the steps may be reversed in different embodiments as long as their purpose is achieved.

Advantageous further developments of the invention are given in the subclaims and described in the description.

Figure list

• 1 ein Fahrzeug, mit dem ein LiDAR-System gemäß der Erfindung betrieblich verbunden ist;
• 2 ein erstes Verfahren gemäß der Erfindung, und
• 3 ein zweites Verfahren gemäß der Erfindung.

Embodiments of the invention are explained in more detail with reference to the drawings and the following description. Show it:

• 1 a vehicle to which a LiDAR system according to the invention is operatively connected;
• 2 a first method according to the invention, and
• 3rd a second method according to the invention.

Embodiments of the invention

In the 1 is LiDAR system 1 shown. The LiDAR system 1 is with a vehicle 2 , here a schematically shown passenger car, operationally connected and thus, by moving the vehicle 2 , movably arranged. The LiDAR system 1 is with a battery 3rd of the vehicle 2 electrically connected and is powered by the battery 3rd supplied with power.

The LiDAR system 1 is set up to be in motion, i.e. while the vehicle is in motion 2 drives, weather conditions in an environment of the LiDAR system 1 to determine. Weather conditions include weather condition types and weather condition intensities. In this exemplary embodiment, types of weather conditions are snow, rain and fog. The LiDAR system 1 is set up to determine information about the weather conditions based on a statistical evaluation of signal intensities, namely in this exemplary embodiment by a processor 4th and one with the processor 4th effectively connected storage 5 that are in the LiDAR system 1 are arranged. The LiDAR system 1 is thereby set up for the statistical evaluation of a reception beam of the LiDAR system 1 perform by a light source (not shown) of the LiDAR system 1 is released into the environment, primarily to obtain in-depth information about the environment.

The LiDAR system 1 is set up for a pulse energy of the receiving beam of the LiDAR system 1 to determine the type of weather condition as the information on the weather conditions. The processor 4th is programmed to determine the expected value and its standard deviation of the pulse energy in order to determine the type of weather condition.

The LiDAR system 1 is further set up to recognize a first expected value as an indicator for a first weather condition intensity and to recognize a second expected value, which is smaller than the first expected value, as an indicator for a second weather condition intensity. The first expected value of the pulse energy, given as the first setpoint in a look-up table in the memory 5 , indicates light rain. The second expected value of the pulse energy, specified as the second setpoint in the look-up table in the memory 5 , indicates comparatively heavier rain. In the case of comparatively heavy rain, the expected value is lower than in the case of comparatively lighter rain. The specific one, i.e. through the LiDAR system 1 measured value of the pulse energy is made by the processor 4th compared with the expected values given as target values in the look-up table, whereby the intensity of the weather conditions in the environment is caused by the LiDAR system 1 can be determined.

More precisely, the embodiment shown is the LiDAR system 1 set up to recognize five discrete weather condition intensities, in particular rain rates, purely by way of example. Each of the five discrete weather condition intensities is a predetermined target value for the expected value, with increasing intensity on a scale of 1 to 5 , associated with the one in the look-up table in memory 5 is stored. Every measured value of the pulse energy is processed by the LiDAR system, especially its processor 4th , assigned by comparison with the look-up table to the target value for the expected value that is closest to it, and then read out the associated weather condition intensity from the look-up table. With the LiDAR system 1 five different weather condition intensities determined and assigned to a person in the vehicle 2 can be provided, for example via a display in a passenger cabin 6th . Possible discrete rain rates, which are defined using the look-up table, are "light rain", "moderate rain", "rain", "heavy rain" and "extreme rain". Instead of five, however, in other, not shown embodiments, more or less discrete weather condition intensities can be recognized, possibly one or more weather condition intensities, preferably three or more, particularly preferably five or more or even ten or more. Ultimately, this only depends on how many different discrete setpoint values are stored as expected values in the look-up table. Distances between adjacent setpoints can be the same or different. In embodiments that are not shown, it is provided that the weather determination intensity is not assigned to a person in the vehicle 2 is provided, but to a system for a driver assistance function or for partially or fully automated driving.

The LiDAR system 1 is further set up to be a standard deviation of that To recognize expected value that is greater than a predetermined threshold value as an indication of a first type of weather condition and to recognize a standard deviation from the expected value that is smaller than the predetermined threshold value as an indication of a second type of weather condition. Possible types of weather conditions are, in particular, snow, rain and fog. The LiDAR system 1 is set up to recognize rain as the first type of weather condition and to recognize fog as the second type of weather condition. The LiDAR system 1 thus first determines an expected value for the pulse energy by means of a histogram over a predetermined time interval, the standard deviation from the histogram and compares the determined standard deviation with the predetermined threshold value. In order to distinguish more than two types of weather conditions, two or more threshold values can be specified in embodiments (not shown), in particular a first threshold value for differentiating between snow and rain and a second threshold value for differentiating between rain and fog.

The LiDAR system 1 is also set up for changes in the received pulse energy of the LiDAR system 1 to determine in order to determine the information about the weather conditions therefrom. The LiDAR system is more precise 1 by means of the processor 4th and memory 5 set up to recognize time-dependent decreasing pulse energy as onset of precipitation and to recognize time-dependent increasing pulse energy as cessation of precipitation. Which type of precipitation, i.e. which type of weather condition, is present in detail, can then result from other statistical values that the LiDAR system 1 certainly. For example, the onset of rain can be determined by the LiDAR system based on a decreasing pulse energy and a comparatively large standard deviation from the expected value of the signal statistics 1 be recognized. If the standard deviation falls below a predetermined threshold value, the LiDAR system 1 however, the onset of fog was recognized. If the standard deviation is comparatively small, fog can be recognized as the weather condition type, but if the standard deviation is comparatively large, rain can be recognized as the weather condition type.

The LiDAR system 1 is also set up to determine pulse energy fluctuations around an average value of the received pulse energy as a function of time in order to determine the information about the weather conditions therefrom. If the pulse energy fluctuations exceed a predetermined fluctuation threshold, the LiDAR system detects 1 Rain. If the pulse energy fluctuations fall below the fluctuation threshold, the LiDAR system detects 1 Fog.

The LiDAR system 1 is therefore set up to distinguish fog from rain. The LiDAR system is thus used to detect weather conditions in the environment of the LiDAR system 1 furnished. The LiDAR system is used for this 1 in detail various methods according to the invention for the statistical evaluation of the signal intensities. The methods thus include the determination of information about the weather conditions by the LiDAR system 1 in the environment of the LiDAR system 1 based on a statistical evaluation of signal intensities. The statistical evaluation can be done in different ways. Two of the procedures are detailed in the 2 and 3rd described.

A first method for the statistical evaluation of the signal intensities is in 2 illustrated in a first flow chart. The first method has one step S21 an offset correction, which subtracts a possible offset in the measured intensity signal of the received pulse by the LiDAR system 1 includes. A step follows after the offset correction S22 a determination of the pulse energy, the determination of the pulse energy of each pulse by integrating the signal over time by the LiDAR system 1 includes. After that is a step S23 a determination of the standard deviation of the resulting distribution of pulse energies by the LiDAR system 1 intended. It follows, again through the LiDAR system 1 , a step S24 a reference of the standard deviation to the mean pulse energy of the pulses in the considered time window. A relative signal fluctuation is thus determined. After that is a step S25 provided in which the relative signal fluctuation caused by the LiDAR system 1 is used as a measure of the type of weather condition, here for precipitation.

wobei die Pulsenergie des i-ten Pulses ist. Alternativ kann die hinter der Summe stehende Differenz der Pulsenergien in diesem Schritt in einer anderen, nicht gezeigten Ausführungsform auch quadriert werden: $$S={\displaystyle \sum _{i=1}^N{\left({\xi }_{i+1}-{\xi }_i\right)}^2}$$

A second method for the statistical evaluation of the signal intensities is in 3rd illustrated in a second flow chart. The second method is insensitive to a change in the mean pulse energy. The second method comprises one step S31 a subtraction of a possible offset of the measured pulse by the LiDAR system 1 , also an offset correction. This is followed by a further step S32 determining the pulse energy of each pulse by integrating the signal over time using the LiDAR system 1 . This is followed by the LiDAR system 1 in one step S33 a determination of the amount of the difference in pulse energies from one pulse to the next for a block of N pulses, N ≥ 2, and a subsequent summation according to the formula $$\mathrm{S.}={\displaystyle \sum _{i=\perp }^N\left|{\xi }_{i+\perp }-{\xi }_i\right|},$$

where is the pulse energy of the i-th pulse. Alternatively, the difference in the pulse energies behind the sum can also be squared in this step in another embodiment, not shown: $$\mathrm{S.}={\displaystyle \sum _{i=1}^N{\left({\xi }_{i+1}-{\xi }_i\right)}^2}$$

wobei n(t) die schnellen Puls-zu-Puls-Energiefluktuationen sind und s₀(t) der zeitlich langsam absinkende Mittelwert. Die zeitliche Ableitung ist nun: $$\frac{d\xi }{dt}=\frac{dn\left(t\right)}{dt}+\frac{d{s}_0\left(t\right)}{dt}$$

The procedure can be analogous to the integration with continuous signals. Let the time-varying pulse energy be defined by $$\xi \left(t\right)=n\left(t\right)+s_0\left(t\right),$$

where n (t) is the rapid pulse-to-pulse energy fluctuations and s ₀ (t) is the mean value, which decreases slowly over time. The time derivative is now: $$\frac{d\xi }{dt}=\frac{dn\left(t\right)}{dt}+\frac{d{s}_0\left(t\right)}{dt}$$

Then the resulting integral would be: $${\displaystyle {\int }_{t1}^{\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="DE102019212877A1\_0012.png"\; state="\{\{state\}\}">t</figure-callout>}2}\left|\frac{dn\left(t\right)}{dt}+\frac{d{s}_0\left(t\right)}{dt}\right|}dt\sim {\displaystyle {\int }_{t1}^{\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="DE102019212877A1\_0012.png"\; state="\{\{state\}\}">t</figure-callout>}2}\left|\frac{dn\left(t\right)}{dt}\right|dt}$$

The contribution of the second term to the integral is negligibly small, since the mean pulse energy only changes very slowly during the integration time.

The length of the integration time is determined by the quantity N, which specifies how many pulses should be added. If the integration time or the N is sufficiently large, the integral or the above-mentioned sum is dominated by the fast-moving pulse-to-pulse energy fluctuations n (t). The slowly changing mean value is no longer noticeable. One later step S34 then includes a reference to the result from the previous step S33 the mean pulse energy of the pulses in the time window under consideration by the LiDAR system 1 . The relative signal fluctuation (fluctuation) determined in this way can finally be used in a next step S35 through the LiDAR system 1 can also be used as the measure of the weather condition intensity, here the precipitation rate. In an alternative embodiment, the LiDAR system 1 instead of the relative signal fluctuation, consider the absolute signal fluctuation, ie the result of S33 not related to the mean pulse energy.

The invention thus allows a clear detection of rain events including classification of a rain rate (for example heavy rain, moderate rain, little rain or also 5 levels) and a differentiation between rain and fog with the aid of a statistical evaluation method. Precipitation detection / classification methods can be provided based on statistical evaluation of the signal intensities of LiDAR echo pulses at a static distance to an object that reflects consistently over the evaluation period. It is easy to distinguish between rain and fog. In the case of a dToF-based LiDAR system 1 With the output of the signal intensity in the receive path, the methods described here can be used without additional hardware or modification of the device. Just an adapted software to control the processor 4th may have to be provided.

The invention is not restricted to the exemplary embodiments shown. Possible areas of application are generally automotive LiDAR systems for autonomous driving of the steps 4/5 , Weather monitoring systems, for example at weather stations or at airports, new types of weather classification sensors in an aircraft, Car2X applications, on-board warning systems for danger spots (rain, black ice, generally slippery surfaces), coefficient of friction analysis through road wetness / road estimation, etc.

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 69933932 T2 [0005]
• DE 202016103719 U1 [0006]
• EP 2820632 A1 [0007]
• EP 3206046 A1 [0008]
• EP 3284646 A1 [0009]

Claims (10)

LiDAR system (1) which is set up to determine weather conditions in an environment of the LiDAR system (1), characterized in that the LiDAR system (1) is set up to provide information about the weather conditions based on a statistical evaluation to determine signal intensities. LiDAR system (1) Claim 1 , wherein the LiDAR system (1) is set up to determine a pulse energy of a reception beam of the LiDAR system (1) in order to determine a type of weather condition as the information about the weather conditions. LiDAR system (1) Claim 2 , wherein the LiDAR system (1) is set up to determine the expected value and its standard deviation of the pulse energy in order to determine the type of weather condition. LiDAR system (1) Claim 3 , wherein the LiDAR system (1) is set up to recognize a first expected value as an indicator for a first weather condition intensity and to recognize a second expected value, which is smaller than the first expected value, as an indicator for a second weather condition intensity. LiDAR system (1) Claim 3 or Claim 4 , wherein the LiDAR system (1) is set up to recognize a standard deviation from the expected value that is greater than a predefined threshold value as an indication of a first type of weather condition and a standard deviation from the expected value that is smaller than the predefined threshold value , to be recognized as an indication of a second type of weather condition. LiDAR system (1) according to one of the Claims 2 to 5 , wherein the LiDAR system (1) is set up to determine changes in the received pulse energy of the LiDAR system (1) in order to determine the information about the weather conditions therefrom. LiDAR system (1) according to one of the Claims 2 to 6th , wherein the LiDAR system (1) is set up to determine pulse energy fluctuations around an average value of the received pulse energy as a function of time in order to determine the information about the weather conditions therefrom. LiDAR system (1) according to one of the preceding claims, wherein the LiDAR system (1) is set up to distinguish fog from rain. Vehicle (2) with a LiDAR system (1) according to one of the Claims 1 to 8th , wherein the LiDAR system (1) is operationally connected to the vehicle (2). Method for recognizing weather conditions in an environment of a LiDAR system (1), characterized by the step: - Determination of information about the weather conditions by the LiDAR system (1) based on a statistical evaluation of signal intensities.

Images