DE102005059902A1 - Distance sensor condition detecting method for measuring motor vehicle distance, involves determining distance values from sensor signals and producing sensor condition signals that correlates with result of statistical evaluation of values - Google Patents

Abstract

The method involves producing sensor signals, which correlate with a distance of a motor vehicle (1) to a driveway limitation, by distance sensors (8a, 8b, 9a, 9b). A set of time and/or location dependent distance values are determined from the signals, and the values are stored. The stored distance values are statistically evaluated, and sensor condition signals, which correlate with the result of the evaluation, are produced. Independent claims are also included for the following: (1) a distance measuring device for a vehicle for measuring a distance of the vehicle from a driveway limitation (2) a parking assistance system of a vehicle for outputting parking indications.

Description

State of technology

The The present invention relates to a sensor state detection method a sensor mounted on a vehicle and a parking assistance system and a distance measuring device of a vehicle for measuring the Distance of the vehicle from a track boundary.

The increasing traffic density and increased development of open space narrow the traffic area, especially in conurbations continuously one. The available Standing parking space is getting tighter and the search for a suitable parking space is straining the driver in addition to the ever increasing traffic. Among other things, therefore, were semiautonomous Parking assistance systems (SPA) designed to help the driver park support should. The driver is thereby the decision whether an existing parking lot for a parking operation sufficient, relieved or accepted.

There are a number of different parking assistance systems known, including, for example, parking assistance systems with so-called "parking space measurement function", with the side of the vehicle mounted proximity sensors to measure the size of a parking space in which the vehicle drives past DE 102 25 894 A1 and from the DE 196 16 447 A1 known. If the system detects a parking space large enough for the vehicle, this can be signaled to the driver. During the subsequent parking process, the system gives the driver instructions or warning signals for parking.

The for parking space measurement provided near range sensors are usually as ultrasonic sensors designed with ranges of up to several meters. Are these Ultrasonic sensors, however, polluted or by deposits of ice or snow or the like impaired, then the sensitivity decreases the sensors off and in particular more distant objects can not More can be detected so that parking spaces are not reliably measured can be. In extreme cases, such defective sensors to collisions during the Parking procedure lead, because the driver yes when parking just on such a reliable warning leaves.

Similar Problems can occur when the sensitivity of the sensors changes due to aging or the detection thresholds of the sensors are set poorly.

Advantages of invention

• – Erzeugen eines mit dem Abstand des Fahrzeugs zur Fahrwegbegrenzung korrelierenden Sensorsignals durch den Sensor;
• – Ermitteln einer Vielzahl von zeit- und/oder ortsabhängigen Abstandswerten aus dem erzeugten Sensorsignal und Abspeichern der ermittelten Abstandswerte;
• – Statistisches Auswerten der gespeicherten Vielzahl von Abstandswerten; und
• – Erzeugen eines mit dem Ergebnis des statistischen Auswertens korrelierenden Sensorzustandssignals.

Accordingly, a method is provided for detecting the sensor state of a vehicle-mounted sensor for measuring the distance of the vehicle to a track boundary, in particular a lateral track boundary, comprising the steps of:

• - Generating a correlated with the distance of the vehicle to the track boundary sensor signal by the sensor;
• - Determining a plurality of time and / or location dependent distance values from the generated sensor signal and storing the determined distance values;
• - Statistical evaluation of the stored plurality of distance values; and
• Generating a correlated with the result of the statistical evaluation sensor state signal.

• – einen Sensor, der ein mit dem Abstand des Fahrzeugs zur Fahrwegbegrenzung korrelierendes Sensorsignal ausgibt;
• – eine Abstandswertermittlungsvorrichtung, die eine Vielzahl von zeit- oder ortsabhängigen Abstandswerten aus dem von dem Sensor erzeugten Sensorsignal ermittelt;
• – einen Speicher, in dem die von der Abstandswertermittlungsvorrichtung ermittelten Abstandswerte gespeichert werden;
• – eine Auswertevorrichtung, die die im Speicher gespeicherte Vielzahl von Abstandswerten statistisch auswertet und den Sensorzustand des Sensors anhand des Ergebnisses der statistischen Auswertung ermittelt, und ein mit dem Sensorzustand korrelierendes Sensorzustandssignal erzeugt.

A corresponding distance measuring device for a vehicle for measuring the distance of the vehicle from a track boundary comprises:

• A sensor which outputs a sensor signal correlated with the distance of the vehicle to limit the travel distance;
• A distance value determining device that determines a plurality of time or location dependent distance values from the sensor signal generated by the sensor;
• A memory in which the distance values determined by the distance value determining device are stored;
• An evaluation device which statistically evaluates the plurality of distance values stored in the memory and determines the sensor state of the sensor on the basis of the result of the statistical evaluation, and generates a sensor state signal correlating with the sensor state.

in this connection means "with sensor signal "correlating the distance of the vehicle to the path limitation from the closed to the distance of the vehicle to the track boundary can be, for example the received signal of an ultrasonic sensor. "With the Sensor state correlating sensor state signal "means a signal that one information about contains the sensor state. For example, it may be a signal which may assume different discrete values associated with different states (e.g., "normal," "dirty," "defective,") also a signal which can take continuous values which for example, the performance of the sensor are proportional.

The idea underlying the invention is to provide a statistical evaluation of the with the help of Ab sensors to determine distance values, so as to close on the state of the distance sensors, so for example, to detect a possible defect of the sensors. A significant advantage of the invention is that the information on the state of the distance sensors can be used to warn, for example, the driver in the case of a defective or only weakly powerful sensor, or the thresholds (characteristics to define the sensitivity) of the sensors to adapt to their performance. Another central idea of the invention is to take into account the geographical position of the vehicle in such a sensor test, or even in the decision whether the parking assistant should be turned on.

The statistical evaluation of the distance values can, for example, the Forming an average of the plurality of distance values, wherein this average value is compared with a predetermined comparison value is, and the sensor sensor signal generated depends on whether the mean greater or is less than the predetermined comparison value.

The Statistical evaluation of the distance values can also determine the relative frequencies of the plurality of distance values, these being relative frequencies are compared with predetermined comparison values, and the generated Sensor state signal depends on whether the relative frequencies bigger or are smaller than the predetermined comparison values. For example, the relative frequencies for certain Distance ranges are determined in the manner of a histogram.

The Statistical evaluation of the distance values can also determine a frequency distribution of curvature values one of the distance values and while determining the distance values traveled by the vehicle Path defined curve, and the generated sensor state signal depends on, whether the ratio the proportion of below-threshold curvature values in the frequency distribution for Proportion of the curvature values above the threshold value in the frequency distribution greater or is smaller than a predetermined comparison ratio.

Further the statistical evaluation of the distance values can be the classification the page boundary into different object types and determining the frequencies of the respective object types, these frequencies the respective object types compared with predetermined comparison values and the generated sensor condition signal depends on whether the relative frequencies bigger or are smaller than the predetermined comparison values. Examples of different Object types to which the page boundaries can be classified are: Car, truck, curb, tree, house, or the like

The Statistical evaluation of the distance values can also be a comparison a time sequence of distance values with specified standard gradients. The specified standard curves can be found in be stored in a memory in the vehicle.

The statistical evaluation of the distance values can continue to form an autocorrelation of the distance values.

If at least two sensors are provided on the vehicle, then can statistically evaluating the distance values forming a cross-correlation the temporal sequence of the distance values, which from those of the at least two sensors generated sensor signals were determined include.

In an advantageous embodiment The statistical evaluation of the distance values comprises a weighting the distance values depending on the vehicle speed. For example, it is possible to use distance values, which were determined at a vehicle speed of at least 30 km / h, to ignore. this makes possible a more reliable Determination of the sensor state, since distance values, in such a high speed are more unreliable than distance values, which are determined at a lower speed. alternative it is also possible to which is determined at a vehicle speed of 0 to 15 km / h have to be considered with a weighting factor of 100% (ie full), Distance values at a vehicle speed of 15 to 30 km / h, with a weighting factor of 50% (ie only limited) to take into account and distance values determined at a vehicle speed above 30 km / h were taken into account, with a weighting factor of 0% (ignore). Such a weighting allows an even more reliable determination the sensor state.

In In a further advantageous embodiment, the statistical comprises Evaluating the distance values a weighting of the distance values depending on of whether a parking assistant provided on the vehicle is switched on is.

In an advantageous embodiment of the invention, a warning signal is output in dependence on the sensor state signal, which indicates the state of the sensor. Thus, the driver can For example, be signaled when the sensor is defective or reduced powerful.

• – Vergleichen einer Amplitude des Sensorsignals mit einem Schwellwert;
• – Ermitteln eines Zeitpunktes des Überschreitens des Schwellwertes durch das Sensorsignal;
• – Ermitteln eines Abstandswertes aus dem ermittelten Zeitpunkt des Überschreitens des Schwellwertes durch das Sensorsignal;

wobei der Schwellwert in Abhängigkeit vom Sensorzustandssignal festgesetzt wird. Somit kann der Schwellwert des Sensors (entspricht der Empfindlichkeit des Sensors) beispielsweise auf die (vom Verschmutzungsgrad abhängige) Leistungsfähigkeit des Sensors abgestimmt werden, so dass auch mit verschmutzten Sensoren akkurate Abstandswerte gemessen werden können. Dabei ist es vorteilhaft, wenn für unterschiedliche Abstandswerte die Amplitude des Sensorsignals mit unterschiedlichen Schwellwerten verglichen wird. In einer vorteilhaften Weiterbildung wird ferner der in Abhängigkeit vom Sensorzustandssignal festgesetzte Schwellwert in Korrelation mit umgebungsabhängigen Parametern, insbesondere Temperatur, Luftfeuchtigkeit und/oder Ort, gespeichert. Unter Verwendung dieser gespeicherten Schwellwerte kann somit der Schwellwert an die momentan vorliegenden Umgebungsparameter angepasst werden.In an advantageous development of the invention, the determination of the distance values comprises the following substeps:

• Comparing an amplitude of the sensor signal with a threshold value;
• - Determining a time of exceeding the threshold by the sensor signal;
• - Determining a distance value from the determined time of exceeding the threshold by the sensor signal;

wherein the threshold value is set in dependence on the sensor state signal. Thus, the threshold value of the sensor (corresponding to the sensitivity of the sensor) can be adjusted, for example, to the (depending on the degree of contamination) performance of the sensor, so that even with dirty sensors accurate distance values can be measured. It is advantageous if, for different distance values, the amplitude of the sensor signal is compared with different threshold values. In an advantageous development, furthermore, the threshold value determined in dependence on the sensor state signal is stored in correlation with environment-dependent parameters, in particular temperature, air humidity and / or location. Using these stored threshold values, the threshold value can thus be adapted to the currently present environmental parameters.

In a further advantageous embodiment of the invention, the following further step is provided:
Determining the current geographical position of the vehicle;
wherein determining the plurality of time and / or location dependent distance values in dependence on the determined current geographical position of the vehicle is performed. This makes it possible to prevent determination of the distance values in environments which are unsuitable for this purpose (eg highway).

In a further advantageous embodiment of the invention, the following further step is provided:
Determining the current geographical position of the vehicle;
wherein the statistical evaluation of the stored plurality of distance values in dependence on the determined current geographical position of the vehicle is performed. Thus, for example, a weighting of the distance values as a function of the determined current geographical position of the vehicle is possible.

In An advantageous development of the invention comprises the distance measuring device Further, a position detecting device for determining the geographical Position of the vehicle, and the evaluation evaluates the distance values dependent on from the geographical determined by the position determining device Position of the vehicle statistically.

In An advantageous development of the invention is also a warning signal generator provided, depending on from the sensor state signal outputs a warning signal indicating the state of the sensor. Thus, the driver can be signaled, for example if the sensor is defective or degraded.

• – einen Sensor, der ein mit dem Abstand des Fahrzeugs (1) zur Fahrwegbegrenzung korrelierendes Sensorsignal ausgibt;
• – eine programmgesteuerten Einrichtung, die den Sensor aktiviert oder deaktiviert und anhand des vom Sensor ausgegebenen Sensorsignals Parklücken entlang eines vom Fahrzeug passierten Fahrweges erkennt;
• – einen Warnsignalgeber, der ein Hinweissignal ausgibt, falls die programmgesteuerten Einrichtung anhand des Sensorsignal eine Parklücke erkennt; und
• – eine Positionsermittlungsvorrichtung zur Ermittlung der geographischen Position des Fahrzeuges, die ein von der geographischen Position des Fahrzeuges abhängiges Positionssignal ausgibt,

wobei die programmgesteuerte Einrichtung den Sensor in Abhängigkeit vom von der Positionsermittlungsvorrichtung ausgegebenen Positionssignal aktiviert oder deaktiviert.An inventive parking assistance system of a vehicle, in particular of a motor vehicle, for outputting parking instructions, contains:

• A sensor that is one with the distance of the vehicle ( 1 ) outputs sensor signal correlating to the travel limit;
• A program-controlled device which activates or deactivates the sensor and recognizes parking spaces along a route traveled by the vehicle based on the sensor signal output by the sensor;
• - A warning signal generator, which outputs a warning signal if the program-controlled device detects a parking space based on the sensor signal; and
• A position-determining device for determining the geographical position of the vehicle, which outputs a position signal dependent on the geographical position of the vehicle,

wherein the program-controlled device activates or deactivates the sensor in dependence on the position signal output by the position-determining device.

One A significant advantage of this aspect of the invention is that under Inclusion of the geographical location of the vehicle, an improved search of parking spaces or display of parking space information possible is. For example, it is possible the search for parking spaces (ie the operation of the sensor) in geographical positions where parking is not possible is to be stopped.

It is advantageous if the position-determining device is used as a satellite-supported position-determining device, in particular as a GPS system. Such a GPS system is already provided in many vehicles and can thus be used as a position detecting device be integrated into the parking assistance system.

The Invention will be described below with reference to the schematic figures The drawings specified embodiments explained in more detail. Show it attended:

1 a schematic diagram of a vehicle with a distance measuring device according to an embodiment of the present invention;

2 the waveform (the envelope) of a sensor signal of a distance sensor as a function of time;

3A a scene detected by the distance sensor of a vehicle while driving (parking search);

3B the scene in 3A corresponding determined distance values as a function of the distance traveled with the sensor intact;

3C the scene in 3A corresponding determined distance values as a function of the distance covered in the case of a defective or soiled sensor;

4 a flowchart showing a method according to an embodiment of the invention;

5A a histogram showing the statistical distribution of the determined distance values with intact sensor; and

5B a histogram showing the statistical distribution of the determined distance values for defective or contaminated sensor.

description the embodiments

In all figures of the drawing are the same or functionally identical elements - if nothing else is stated - with the same reference numerals have been provided.

1 shows a schematic diagram of a vehicle with a distance measuring device according to an embodiment of the present invention.

In 1 is a motor vehicle 1 shown schematically. At a vehicle front 2 are distance sensors 3 arranged. At a vehicle rear 4 are also distance sensors 5 arranged. On a left side of the vehicle 6 are lateral distance sensors 8a and 8b intended. On a right side of the vehicle 7 are lateral distance sensors 9a and 9b intended. The distance sensors are used to measure distances to obstacles in the vehicle environment. The distance sensors 3 . 5 . 8th . 9 are formed in the present embodiment as ultrasonic sensors. However, you can also measure distances based on another measuring principle, eg with optical signals or with radar signals. Furthermore, video sensors are possible which determine a distance from a recorded image information. The distance sensors 3 . 5 . 8th . 9 supply their sensor signals via a data bus 10 to a program-controlled device 11 (For example, a microprocessor, microcontroller or the like) with a memory 18 in the vehicle 1 , The program-controlled device 11 determined by the distance sensors 3 . 5 . 8th . 9 supplied sensor signals distances to obstacles in the vehicle environment and the location of these obstacles in the vehicle environment. To accurately determine the location of the obstacles, the program-controlled device can 11 also use the principle of triangulation, whereby the distance values determined by the various sensors are compared with one another.

Furthermore, the program-controlled device 11 designed to determine a suitable parking space and optionally to determine a driving trajectory in this parking space. In this sense, the program-controlled device is used 11 also as a parking assistant. It also preferably determines expenses to the driver. For the output is the program-controlled device 11 connected to a warning buzzer that serves as an indicator 12 and / or as a speaker 13 can be trained. The ad 12 In particular, it is executed as a screen of a navigation display in the vehicle. Furthermore, instructions can also be issued via an indicator in a combination instrument, via a head-up display or via LED displays which are additionally to be mounted on the instrument panel.

With the help of the ad 12 or the speaker 13 For example, notes may be output that tell the driver, for example, that the vehicle has just passed a sufficiently large parking space.

To determine a movement or the speed of the vehicle is the program-controlled device 11 preferably via a data bus 14 , which is designed in particular as a CAN bus, with at least one speed sensor 15 and a gear sensor 17 connected. In a preferred embodiment, the speed sensor is 15 designed as a wheel speed sensor that measures a wheel movement of the vehicle. If a wheel movement is detected, the current speed of the vehicle is determined on the basis of the wheel rotation and the circumference of the wheel and the course of time. From the current Ge speed of the vehicle can in turn be closed in conjunction with the passage of time on the distance covered.

Also provided is a navigation system 16 (Positioning device), which via a data line with the program-controlled device 11 connect is. The navigation system 16 may in particular be designed as a satellite-based system, in particular as a GPS system (Global Positioning System), and uses for this purpose, signals emitted by satellites, for determining the position of the vehicle by means of triangulation. The thus determined position data sent to the program-controlled device 11 are issued to the driver on the monitor 12 to display the current position, eg on a map section of the vehicle environment. The navigation system 16 may also determine a route to a destination entered by the driver (ie, perform a navigation) and the driver via the display 12 and / or the speaker 13 Give directions to reach the destination.

2 shows the waveform 20 the envelope of a sensor signal output by one of the distance sensors of a distance sensor as a function of time. A transmitted by the distance sensors transmit pulse (ultrasonic pulse) is sent out in each measurement cycle. The reflected sound from an obstacle is received by the distance sensor, and converted by an ultrasonic transducer of the distance sensor into an electrical sensor signal. 2 shows a typical waveform of the envelope of this electrical sensor signal with the amplitude A over the time axis T, with a so-called initial crosstalk 21 , An echo pulse reflected by an obstacle 22 occurs at a time T1 on which this echo pulse 22 has a certain period of time until a further time T2. The times T1 and T2 are by means of a settable threshold 24 defined, which corresponds to a certain amplitude value. In this so-called echo signal method corresponds to the time from the transmission of the signal to the occurrence of the echo pulse 22 the distance of the thus scanned object from the transmitter (distance sensor). The ultrasonic pulse emitted by the distance sensor propagates at a speed of sound which is only slightly dependent on the temperature and the weather conditions and can therefore be considered as known. The time from the emission of the ultrasonic pulse to the occurrence of the echo pulse 22 at time T1, the double time required by an ultrasonic pulse (due to the round-trip time) thus corresponds to the distance between the ultrasonic sensor and the obstacle. By evaluating the waveform 20 can the program-controlled device 11 Thus, on the basis of time T1, the distance from an obstacle located in the measuring field (also called measuring lobe) of the distance sensor is determined.

The time difference between the times T1 and T2 is called the pulse length 23 and may, in particular in conjunction with the amplitude of the echo pulse 22 , Give indications of the nature of the obstacle. Thus, a small pulse length and amplitude indicate a small height of the measured object. In conjunction with suitable plausibility measures, a classification (ie classification into object types) of the obstacle can thus also be made. In other words, the waveform can be used to infer whether the object is a car, a truck, a curb, a wall, a tree, and so on. A plausibility check can take place, for example, by checking the object types classified as passing by at a parking space during the subsequent parking process. Thus, for example, the parking assistant can first classify a parking space limiting obstacle as a curb. If one of the rear sensors then approaches the obstacle during the parking process up to a very short distance (eg 15 cm), then this rear sensor should no longer detect it because the measuring lobe of the rear sensor spreads over the curb. However, if the rear sensor nevertheless detects an obstacle at this distance in this situation, then this is an indication that the obstacle is not a curb as originally suspected, but rather a higher obstacle, such as a wall. In other words, the original hypothesis that the obstacle is a curb has proved to be implausible. In this situation, the parking assistant may issue a warning signal that causes the driver to cancel the parking operation and / or a new parking trajectory may be determined.

The program-controlled device 11 Determined distance values or objects can be used for the parking assist function (parking assistant) of the vehicle 1 be used. This is based on 3 explained in more detail.

3A shows one of the lateral distance sensor 9a of the vehicle 1 scene captured while driving (parking search). The vehicle 1 moves on a street along which several vehicles 31 (Track limits) are parked. Between two of these vehicles 31 there is a parking space 30 , The distance sensors 9a and 9b move along the trajectory represented by a dashed line 32 ,

3B is a schematic representation of the scene in 3A correspondingly determined distance values A as a function of the distance traveled W with an intact and clean distance sensor 9a , The determined distance values are used to identify the obstacles, possibly with the help of a plausibility check as described above. As in 3B shown in the representation from left to right (ie along the direction of travel) initially two vehicles 34 , then a curb at a distance 35 (Track boundary), and finally another vehicle 34 identified. The distance values and / or the identified objects are stored in memory 18 saved.

3C is a schematic representation of the scene in 3A correspondingly determined distance values as a function of the distance traveled with dirty distance sensor 9a , In the 3C become the vehicles 31 falsely as curbs 36 identified. Such a misidentification may be due, for example, to the fact that due to the pollution, an echo pulse 22 is recorded, which has relatively short and / or a small amplitude, which is characteristic of low obstacles. Furthermore, the corners of the vehicles, which generate a parabolic distance course when the sensor is clean, are not detected due to their lower reflectivity than the sides of the vehicle. Only the vehicle pages themselves are detected. The consequence of such a misinterpretation may be that the parking assist system assumes that parts of the vehicle (rear, front end) are allowed to protrude over the obstacle without collision and therefore issue false warning signals or omit the issuing of warning signals altogether.

Between the second and the third vehicle 31 In the direction of travel is limited by a curb parking space. However, the curb (or the curbs assigned distance values) is perceived only fragmentary, so that a correct identification of the curb does not take place.

In an embodiment of the invention the distance values or the identified objects are in memory stored and subjected to a statistical evaluation, based the condition (e.g., intact, soiled, defective, etc.) of the proximity sensor can be determined.

4 shows a flowchart showing an exemplary method according to an embodiment of the invention.

In step S1, a sensor signal is generated. For this purpose, the distance sensor sends 9a an ultrasonic signal, which is reflected by the obstacle. The received echo signal is converted into an electrical signal, which is the program-controlled device 11 is supplied.

In step S2, the determination of a distance value on the basis of in 2 sketched procedure. The program-controlled device compares this 11 the amplitude of the waveform 20 with a threshold 24 , When the amplitude of the waveform 20 at time T1, the threshold 24 exceeds, then closes the program-controlled device 11 that there is an obstacle in the measuring field. From the time period T1, the program-controlled device closes 11 then the distance of this obstacle to the distance sensor 9a , The associated distance value is stored in memory 18 saved.

In the following steps S3 and S4, those in memory 18 stored distance values statistically evaluated. This can be done in many ways. In the present embodiment, the relative frequency of different pitch ranges is determined. This is based on the 5 explained in more detail.

5A is a histogram showing the statistical distribution of the measured distance values for an intact, clean distance sensor 9a shows. For this purpose, the distance values are divided into different distance ranges (eg 0-50 cm, 50-100 cm, 100-150 cm, etc.). Each time a new distance value is measured, it is determined in which distance range that distance value falls and a corresponding counter is incremented. In the in 5A shown distribution Against comparatively many distance values in the distance ranges 0-50 cm and 50-100 cm ago. This is because in the street along which the vehicle 1 moves, comparatively many vehicles 31 are parked, the distance to the distance sensor during the passage is usually not more than 1 m. Also, there are comparatively many distance values in the distance ranges 300-350 cm. This distance corresponds to the curb 33 , which in passing this distance to the distance sensor 9a located. In the distance between 100 and 300 cm, however, fewer obstacles are detected. This in 5A The histogram shown thus shows a typical distribution of distance values measured with an intact, clean sensor along a street with parked cars.

5B on the other hand shows a histogram, which represents the statistical distribution of the determined distance values with defective or polluted sensor. With this distance sensor obstacles in the distance up to 50 cm still can be perceived, obstacles in the distance of 50-100 cm however Only limited, and obstacles at a distance of more than 100 cm are no longer perceived.

In the present embodiment, the statistical evaluation is now carried out by forming the mean value (average value or median) of the distance values in step S3, comparing this mean value with a predetermined comparison value in step S4, and the generated sensor state signal depending on whether the average value is greater or less than the predetermined comparison value. If the mean value of the distance values is very low, then this is an indication that the distance sensor 9a possibly, because of pollution, no longer recognizes obstacles further afield. Specifically, the median of the distance values is determined in step S3, and the determined median is compared with the comparison value 0.5 m in step S4. If the determined median is less than the comparison value 0.5 m, then the procedure jumps to step S5, in which a sensor state signal is generated, which signals that the distance sensor 9a is dirty and / or defective. If the determined median is greater than the comparison value, then the procedure jumps to step S6, in which a sensor state signal is generated, which signals that the distance sensor 9a not mischievous and / or defective. Thereafter, the procedure jumps back to step S1.

In other words, it is determined in step S4 whether the proportion of histogram values in the 5A and 5B in the range 0-0.5 is greater than 50%. If so, then the procedure jumps to step S5 and otherwise jumps to step S6. Im in 5A As shown, the proportion of the distance values greater than 0.5 m is about 80% and is thus greater than 50%, for which reason the procedure jumps from step S4 to step S6. Im in 5B By contrast, in the example shown, the proportion of the distance values greater than 0.5 m is only about 20% and is thus less than 50%, for which reason the procedure jumps from step S 4 to step S 5. By statistical evaluation of the determined distance values can thus on the state of the distance sensor 9a getting closed.

The statistical evaluation can also be done in other ways. For example, in a second embodiment, by comparing the statistical distribution of the determined distance values with an expected distribution to the state of the distance sensor 9a closed. For this purpose, the distance of the statistical distribution of the determined distance values from an expected standard distribution is calculated. This can be done, for example, by forming the mean value of the differences of the determined relative frequencies to the comparison values of the standard distribution or similar statistical methods. A suitable standard distribution can be determined, for example, empirically and in memory 18 be stored. It is also possible, depending on the geographical position of the vehicle one of several in memory 18 stored standard distributions (this is explained in more detail in connection with the eighth embodiment).

In step S4, the mean value thus formed is compared with a suitable threshold value. If the average is above the threshold, then the procedure branches to step S5, where a sensor state signal is generated, which signals that the proximity sensor 9a is dirty and / or defective. If the average value is not above the threshold, then the procedure branches to step S6, where a sensor state signal is generated, which signals that the distance sensor 9a not dirty and / or broken. Thereafter, the procedure jumps back to step S1.

The statistical evaluation can also be applied to identified objects. In a third embodiment, objects (car, truck, curb, tree, house, etc.) are identified from the distance values in an additional method step. From the relative or even the absolute frequency of these identified objects can also be on the state of the distance sensor 9a getting closed. For this purpose, similar to the first or the second embodiment, comparisons with, possibly dependent on the geographical position, expectation values can be used. Unusual waveforms can be detected not only in the case of insufficient detection performance, but also in the case of objects whose characteristic pattern is still unknown to the system. If one parks in front of or behind such an object, then the position of the object can be measured or verified, for example, with the front and / or rear sensors. If the determined position agrees well with the position determined by the side sensors, insufficient detection performance of the side sensors is unlikely. It is possible to ignore such an object in the statistics. Furthermore, it is also possible to adopt such a new signal curve, possibly only after repeated detection and verification, in the system as a new standard curve.

In a fourth embodiment, the time sequence of the determined distance values is analyzed. Like from the 3B can be seen, adjacent distance values correlate very strongly with each other. In other words, when using the distance sensor 9a output sensor signal is determined a certain distance value, then the There is a high probability that the distance value determined in the next measurement cycle will differ only slightly or not at all. An exception to this is the edges of objects such as trucks or house walls; but these occur relatively rarely. In this embodiment, therefore, the autocorrelation of the determined sequence of distance values is determined and from this the condition of the distance sensor 9a inferred. For example, the value of the autocorrelation is when driving past the parking space 30 in 3A with dirty distance sensor 9a (please refer 3C ) lower than with a clean distance sensor 9a (please refer 3B ).

In a fifth embodiment, the timing of the two is determined by the different distance sensors 9a and 9b Analyzed output signals detected sensor signals. As in 1 shown, are the distance sensors 9a and 9b at a known distance on the right side of the vehicle 7 of the vehicle 1 , When the vehicle drives past an obstacle located on the right side of the vehicle, the obstacle first enters the measuring field of the distance sensor 9a one and after a certain (dependent on the vehicle speed) period in the field of view of the distance sensor 9b one. A comparison of the on the distance sensors 9a and 9b based distance values therefore leaves, taking into account the speed sensor 15 measured vehicle speed, a conclusion on the state of the distance sensors 9a and 9b to. For this, the cross-correlation is due to that of the distance sensors 9a and 9b output sensor signals formed. When properly used, the cross-correlation should have a maximum which corresponds to the current vehicle speed. If the cross-correlation has no such maximum, then this is an indication that at least one of the two distance sensors 9a and 9b is defective.

In a sixth embodiment, the weighting of the distance values or objects on which the statistical evaluation is based is dependent on that of the speed sensor 15 measured instantaneous vehicle speed. At high vehicle speeds, the measured distance values are subject to a higher inaccuracy or uncertainty than at lower vehicle speeds. Therefore, it makes sense to provide the distance values, which were measured at a lower vehicle speed, in the statistical evaluation with a greater weight. It is also possible distance values measured above a certain vehicle speed (eg 25 km / h) are not considered in the statistical evaluation.

In a seventh embodiment a weighting of the statistical evaluation is used Distance values or objects depending on whether one in the vehicle provided parking assistant is turned on. In general, one will Use of the parking assistant only if the A comparatively low speed vehicle (e.g., 15 or 20 km / h). Furthermore, the parking assistant usually only turned on in certain environments, for example, typically in environments where there are several objects along the road (especially cars) interrupted by more or less large parking spaces are parked. Especially this is the case when the assistant is switched on manually by the driver to. Therefore leads a higher one Weighting of those distance values or objects that are switched on with Parking assistants were measured, to a higher accuracy of the determined Sensor status signal.

The statistical characteristics of the measured distance values on which the statistical analysis is based are highly dependent on the vehicle environment. So that shows in 5A Histogram shown is a distribution of distance values, as is typical for measurements along a densely populated road. However, in other vehicle environments, such as driving on rural roads outside the city or on highways, this distribution will be different. In particular, fewer measured values for small distances are to be expected in this case. Also, the relative or absolute frequency of the detected objects (car, truck, curb, tree, house, etc.) used in the third embodiment also differs depending on the geographical position of the vehicle.

In an eighth embodiment, therefore, the detection of the sensor state is performed location-dependent. This is one of the navigation system 16 outputted position signal is used to obtain suitable comparison values (eg a suitable standard distribution) for step S4 in FIG 4 select.

This is done first by using the navigation system 16 output position signal, the position of the vehicle 1 on one example in memory 18 stored navigation map determined. This navigation map is divided into different geographic regions, each of which is assigned a specific type of road (eg city street, highway, motorway, residential area, etc.). Through this comparison with the navigation system 16 output position signal can thus be detected in what kind of vehicle environment the vehicle is currently. On the basis of this information, a suitable standard distribution (or suitable comparison values) can be derived from several in memory 18 stored standard distributions (or comparison values) are selected.

In an alternative embodiment of the eighth embodiment, the detection takes place the sensor state only in certain vehicle environments. For example Is it possible, to carry out the detection of the sensor condition only in areas where typically parked, e.g. in residential areas, and the capture sensor state in unsuitable vehicle environments (e.g. on highways).

The use of the geographical position of the vehicle obtained via the navigation system 1 Incidentally, it is not only sensible for the decision whether a detection of the sensor state (ie a sensor test) should be performed or not, but also for the decision whether the parking space measurement is to be performed or not, in other words, whether the parking assistant should be turned on or not. In a ninth embodiment, therefore, there is an automatic activation of a parking assistant of the vehicle 1 depending on the geographical position of the vehicle 1 , For this purpose, as in the eighth embodiment, that of the navigation system 16 issued position signal used to first detect the nature of the vehicle environment. If it is detected that the vehicle 1 In an environment where parking is typically done, eg in a residential area, the parking assistant is automatically turned on. This has the advantage that the driver does not have to turn on the parking assistant manually, but can concentrate fully on the driving.

In an advantageous development of the ninth embodiment, the information about the geographical position of the vehicle is used to prevent an activation of the parking assistant in areas where parking is impossible. If the driver of the vehicle 1 Attempts to activate the parking assistant in a vehicle environment in which parking can not be parked (eg on a motorway, in the no-parking area, at intersections), or the vehicle 1 gets into this vehicle environment with activated parking assistant, then this is based on the with the navigation system 16 detected geographic position of the vehicle detected, the activation of the parking assistant is suppressed (or the parking assistant is at least temporarily disabled) and a corresponding warning that parking at this position is not possible, is issued with the warning signal generator. This has the advantage that alleged parking spaces (eg parking spaces in the ban) are not displayed.

The information about the geographical position of the vehicle 1 can also be used in other ways for the detection of the sensor status (sensor test) or the automatic activation / deactivation of the parking assistant. Thus, in a tenth embodiment, the detection of the sensor state or the parking assistant is automatically activated in the vicinity of positions in which it has already been parked at an earlier time. These are all geographical positions in which the vehicle 1 parked once or in which the parking assistant has already been manually activated, in memory 18 saved. Approaching the vehicle 1 one of these stored positions (eg up to 300 m), then eg the parking assistant can be activated automatically. This has the advantage that in a situation in which the driver's desire for a parking space measurement is likely, switching on the parking assistant is automatic, so that the driver can concentrate fully on the driving. It is also possible to consider for the automatic activation of the parking assistant only geographical positions, in the vicinity of the vehicle 1 already executed a certain number (eg at least 10) of parking operations. Such filtering increases the likelihood that automatic activation of the parking assistant will only occur in areas where the driver actually wishes to park.

In an eleventh embodiment of the invention is taken into account on which vehicle side of the vehicle 1 a parking operation is possible. This is based on the navigation system 16 certain information about the geographical position determines on which side a parking operation is possible. For example, if the system recognizes that the vehicle is in a one-way street, then it may indicate that parking is possible on either side of the vehicle. In this case, both the distance sensors 8th on the left side of the vehicle 6 as well as the distance sensors 9 on the right side of the vehicle 7 used for parking space detection on the respective sides of the vehicle. In streets with oncoming traffic, however, the parking space detection on the right side of the vehicle 7 with the distance sensors 9 be limited. Furthermore, it is also possible to activate in a one-sided parking ban only the sensors on the opposite side of the parking prohibition. The with the navigation system 16 The information obtained about the type of road can also be used in a similar way for the detection of the sensor state. For example, it is possible to detect the state of the left-side distance sensors 8th to restrict to the case that the vehicle 1 through a one-way street, since in this case the measured signals are not distorted by oncoming vehicles.

In a twelfth embodiment of the invention, the parking assistant is automatically activated when navigating with the aid of the navigation system 16 ends. This is done by the navigation system 16 determined current position compared with the previously entered by the driver as a destination position. If the current position of the vehicle is in a certain radius (eg 300 m) from the target position, the parking assistant is automatically switched on. This has the advantage that the driver at the destination does not have to turn on the parking assistant manually, but can concentrate fully on the driving. In an advantageous embodiment of this embodiment, the parking assistant is only turned on when the navigation ends in a street in which parking is possible (eg in a residential area).

in the Below is another concrete embodiment of a method explained according to the present invention.

6A shows an example of intact and clean distance sensor 9a actually measured distance values over a road section according to the schematic representation in 3A , It shows 6A the determined distance values A (in cm) as a function of the distance traveled W (in cm) with an intact and clean distance sensor 9a , The distances A and the path W are applied at different scales on the vertical and the horizontal axis.

The two objects in the foreground (ie with a smaller distance) have a typical curve for the class "car" (almost straight course over 4 to 6 m with front and rear subsequent curvature) and are thus correctly detected and as cars 34 classified. Also between the two cars 34 lying object has a typical for the class "curb" curve and is thus correctly detected and classified so curb.

6B is a histogram which shows a frequency distribution of the curvature k of the distance values (or the curve or graph defined by the distance values) 6A shows. The curvature of a curve indicates the change in direction of the curve per unit length. The curvature values of a curve are between 0 and 1, with the curvature of a straight line being equal to zero. Im in 6B shown histogram is a maximum for the range (0.0-0.05), which is the range of the absolute or nearly straight sections in 6A equivalent. However, there are also comparatively many values in the range of 0.05-0.35, which is typical for the in 3A illustrated vehicle environment in which comparatively many "curvy" or curved areas at the front and rear ends of the roadside vehicles are detected.

7A shows an example of a defective distance sensor 9a actually measured distance values over a road section according to the schematic representation in 3A , The two cars in the foreground (ie with a smaller distance) do not show the curve typical for their class, but rather a characteristic for the class "curb" curve course (relatively straight course without subsequent curvature) and are thus falsely as curbs 36 classified. The curb 35 between the passenger cars (or the distance values assigned to the curb), however, is correctly detected and classified. However, much less distance values are measured in this example than in 6A shown example, and the curvy sections of the cars are hardly captured.

7B is a histogram which shows the distribution of the curvature k of the distance values (or the graph defined by the distance values) in 7A shows. Also in the 7B The histogram shown is a maximum for the range (0.0-0.05). However, this maximum is greater than the maximum of the histogram in 6B because in 7A there are more straight areas than in 6A , Furthermore, the histogram of 7B hardly any values with a curvature k> 0.2, since these curvatures are detected only insufficiently by the defective sensor.

The statistical evaluation of the stored plurality of distance values is carried out in this example now such that it is determined how the ratio between the proportion of below a threshold (predetermined comparison value) values of the determined distance values (see 6A . 7A ) derived histogram of the curvature k ( 6B . 7B )) is the proportion of the values above this threshold value. In the present example, this threshold is as in 6B and 7B shown, set to 0.2. If this ratio is greater than a predetermined ratio (in this example 50%: 50%), then the program-controlled device generates 11 a sensor state signal indicating that the sensor 9a is defective, and otherwise a sensor condition signal indicating that the sensor 9a not defective.

In the example of 6B the proportion of histogram values smaller than 0.2 is approximately 40% and the proportion of histogram values greater than 0.2 approximately 60%. Since this ratio is 40/60 smaller than the predetermined ratio (comparison ratio) 50/50, the program-controlled device generates 11 a sensor state signal indicating that the sensor 9a not defective.

In the example of 7B the proportion of histogram values smaller than 0.2 is about 90% and the proportion of histogram values greater than 0.2 about 10%. Since this ratio is 90/10 greater than the predetermined ratio 50/50, the program-controlled device generates 11 a sensor state signal indicating that the sensor 9a is defective.

The Decision regarding of the relationship the proportions of the values above and below the threshold and thus the generation of the sensor state signal is performed only when a sufficient number of distance values has been determined, ie For example, after the statistics for a certain period of time (e.g., 1 minute) or over one certain distance (e.g., 500 m) has been updated.

Thus, a statistical evaluation of the with the help of the sensor 9a determined distance values to the condition of the sensor 9a close and a defect of the sensor 9a to recognize. Information about a condition of the gauge 9a will continue to be used to alert the driver in the event of a faulty or low-power sensor 9a with the help of the ad 12 and / or the threshold characteristics (sensitivity definition curves) of the sensor 9a to adapt to its performance.

Although the present invention by way of example by way of preferred embodiments It is not limited to this, but in many ways and modifiable.

For example, the navigation system 15 via the program-controlled device 11 with the monitor 12 connected. However, it is also possible to use the navigation system 15 directly with the monitor 12 connect to.

Further In the method described above, the indication of the sensor state on only two different states limited. However, it is also possible For example, with the help of a suitable indicator, the degree of contamination to identify.

Further it is also possible the methods of the various embodiments and examples with each other to combine.

Claims (22)

Method for detecting the sensor status of a vehicle ( 1 ) mounted sensor ( 8a . 8b . 9a . 9b ) for measuring the distance of the vehicle ( 1 ) to a track boundary ( 31 . 33 ), in particular a lateral track boundary ( 31 . 33 ) comprising the steps of: - generating one with the distance of the vehicle ( 1 ) to the track boundary ( 31 . 33 ) correlating sensor signal ( 20 ) through the sensor ( 8a . 8b . 9a . 9b ); - Determining a plurality of time and / or location dependent distance values from the generated sensor signal and storing the determined distance values; - Statistical evaluation of the stored plurality of distance values; and - generating a correlated with the result of the statistical evaluation sensor state signal. Method according to claim 1, characterized in that that the statistical evaluation of the distance values forming a Mean of the plurality of distance values includes, this mean with is compared with a predetermined comparison value, and the generated Sensor state signal depends on whether the mean is greater or is less than the predetermined comparison value. Method according to claim 1 or 2, characterized that the statistical evaluation of the distance values determine the relative frequencies of the plurality of distance values includes these relative frequencies are compared with predetermined comparison values, and the generated Sensor state signal depends on whether the relative frequencies bigger or are smaller than the predetermined comparison values. Method according to at least one of the preceding claims, characterized in that the statistical evaluation of the distance range determines the determination of a frequency distribution of curvature values (k) of one of the distance values and during the determination of the distance values from the vehicle ( 1 ) and the generated sensor state signal depends on whether the ratio of the fraction of the below-threshold curvature values (k) in the frequency distribution to the proportion of the above-threshold curvature values (k) in the frequency distribution is greater or less than one predetermined comparison ratio. Method according to at least one of the preceding claims, characterized in that the statistical evaluation of the distance values comprises the classification of the page boundary into different object types and the determination of the frequencies of the respective object types, these frequencies of the respective object types are compared with predetermined comparison values, and the generated sensor state signal thereof depends on whether the relative frequencies are greater or less than those above certain comparative values. Method according to at least one of the preceding Claims, characterized in that the statistical evaluation of the distance values a Comparison of a temporal sequence of distance values with specified standard courses includes. Method according to at least one of the preceding Claims, characterized in that the statistical evaluation of the distance values the Forming an autocorrelation of the distance values. Method according to at least one of the preceding claims, characterized in that at least two sensors ( 8a . 8b . 9a . 9b ), in particular ultrasonic sensors, on the vehicle ( 1 ), and that statistically evaluating the distance values comprises forming a cross-correlation of the time sequence of the distance values determined from the sensor signals generated by the at least two sensors. Method according to at least one of the preceding Claims, characterized in that the statistical evaluation of the distance values a Weighting of the distance values as a function of the vehicle speed includes. Method according to at least one of the preceding claims, characterized in that the statistical evaluation of the distance values is a weighting of the distance values as a function of whether the vehicle ( 1 ) Einparkassistent is switched on, includes. Method according to at least one of the preceding claims, characterized in that in response to the sensor state signal, a warning signal is output, which determines the state of the sensor ( 8a . 8b . 9a . 9b ). Method according to at least one of the preceding claims, characterized in that the determination of the distance values comprises the following substeps: - comparing an amplitude of the sensor signal ( 20 ) with a threshold ( 24 ); Determining a time of exceeding the threshold value ( 24 ) by the sensor signal ( 20 ); Determining a distance value from the determined time of exceeding the threshold value ( 24 ) by the sensor signal ( 20 ); where the threshold value ( 24 ) is set in response to the sensor state signal. A method according to claim 12, characterized in that for different distance values, the amplitude of the sensor signal ( 20 ) with different thresholds ( 24 ) is compared. Method according to claim 12 or 13, characterized in that the threshold value determined in dependence on the sensor state signal ( 24 ) is stored in correlation with environment-dependent parameters, in particular temperature, humidity and / or location. Method according to at least one of the preceding claims, characterized in that the following further step is provided: determining the current geographical position of the vehicle ( 1 ); wherein determining the plurality of time and / or location dependent distance values in dependence on the determined current geographical position of the vehicle ( 1 ) is carried out. Method according to at least one of claims 1 to 14, characterized in that the following further step is provided: determining the current geographical position of the vehicle ( 1 ); wherein the statistical evaluation of the stored plurality of distance values as a function of the determined current geographical position of the vehicle ( 1 ) is carried out. A method according to claim 16, characterized in that the statistical evaluation of the distance values, a weighting of the distance values as a function of the determined current geographical position of the vehicle ( 1 ). Distance measuring device for a vehicle ( 1 ) for measuring the distance of the vehicle ( 1 ) of a track boundary ( 31 . 33 ), - with a sensor ( 8a . 8b . 9a . 9b ), one with the distance of the vehicle ( 1 ) to the track boundary ( 31 . 33 ) correlating sensor signal ( 20 ) outputs; With a distance value determination device ( 11 ) containing a plurality of time- or location-dependent distance values from that of the sensor ( 8a . 8b . 9a . 9b ) generated sensor signal ( 20 ) determined; - with a memory ( 18 ) in which the distance values determined by the distance value determining device are stored; - with an evaluation device ( 11 ), which are stored in memory ( 18 ) statistically evaluates stored plurality of distance values and the sensor state of the sensor ( 8a . 8b . 9a . 9b ) is determined on the basis of the result of the statistical evaluation, and generates a sensor state signal correlating with the sensor state. Distance measuring device according to claim 18, characterized in that further comprises a position detecting device ( 16 ) for determining the ge ographic position of the vehicle ( 1 ) is provided, and the evaluation device ( 11 ) the distance values as a function of that of the position-determining device ( 16 ) determined geographical position of the vehicle ( 1 ) statistically evaluated. Distance measuring device according to claim 17 or 18, characterized in that further comprises a warning signal generator ( 12 . 13 ) is provided which outputs a warning signal in response to the sensor state signal, which indicates the state of the sensor ( 8a . 8b . 9a . 9b ). Parking assistance system of a vehicle ( 1 ), in particular a motor vehicle, for issuing parking instructions, - with a sensor ( 8a . 8b . 9a . 9b ), one with the distance of the vehicle ( 1 ) to the track boundary ( 31 . 33 ) correlating sensor signal ( 20 ) outputs; - with a program-controlled device ( 11 ), the sensor ( 8a . 8b . 9a . 9b ) is activated or deactivated and used by the sensor ( 8a . 8b . 9a . 9b ) output sensor signal ( 20 ) Parking spaces along one of the vehicle ( 1 ) passed track recognizes; - with a warning signal generator ( 12 . 13 ), which outputs a warning signal if the program-controlled device ( 11 ) based on the sensor signal ( 20 ) detects a parking space; and - with a position-determining device ( 16 ) for determining the geographical position of the vehicle ( 1 ), which depends on the geographical position of the vehicle ( 1 ) outputs dependent position signal, wherein the program-controlled device ( 11 ) is designed so that it the sensor ( 8a . 8b . 9a . 9b ) depending on the position determining device ( 16 ) Output position signal activated or deactivated. Parking assistance system according to claim 21, characterized in that the position-determining device ( 16 ) is designed as a satellite-based position-determining device, in particular as a GPS system.