publication number DE102009035072 (A1)
Publication Date: 2011 -02-10
Inventor: STEPHEN ZECHA [DE]; Rasshofer RALPH Helmar [DE] + (ZECHA, STEPHAN; Rasshofer, RALPH Helmar)
Applicant: BAYERISCHE MOTOREN WERKE AG [DE]; CONTINENTAL SAFETY ENGINEERING INTERNATIONAL GMBH [DE] + (BAYERISCHE MOTOREN WERKE AKTIENGESELLSCHAFT; CONTINENTAL SAFETY ENGINEERING INTERNATIONAL GMBH)
Classification:
International: B60R21/01; B60W30/08; G08G1/16
European: G08G1/16; G08G1/16A
Application number : DE200910035072 20,090,728
Priority number (s): DE200910035072 20,090,728
The EPA is not responsible for the accuracy of data originating from other authorities than the EPO, and thus adheres especially not for its completeness, timeliness or fitness for a particular purpose. Information, please refer to the respective patent authorities.
this text translated description of DE 102009035072 (A1)
[0001] The present invention relates to a method for predicting the position or the location area and / or the movement of an object or vulnerable road users such as a pedestrian, relative to a vehicle , the current relative position of the object to the vehicle is used for prediction, which is determined by the location of the object side arranged by a vehicle-side Sende-/Empfängereinheit Sende-/Empfängereinheit.
[0002] The generic method used to impending collision by a vehicle with other vulnerable road users predict how pedestrians, cyclists or even animals such as dogs. So shall the driver of the vehicle collision, if necessary, the road users are warned in time or according to the risk of collision or heavy-counter measures such as braking or full deployment of airbags for pedestrians are encouraged in advance of an impending collision. This should be avoided with vulnerable road users and are reduced in the second place, the accident severity of collisions in the first place collision accidents.
[0003] Furthermore, the present invention device sub-units for carrying out the inventive method.
[0004] The method and apparatus for predicting the position and / or movement of an object or a pedestrian relative to a vehicle using imaging sensors have been known for some years.
[0005] For the avoidance of accidents involving pedestrians, cyclists or animals by road vehicles with imaging sensors such as cameras are equipped, which monitors the road in the direction of vehicle travel constantly. example, if a pedestrian caught by the camera, close to the vehicle dangerous, the driver is warned or initiated autonomous braking.
is [0006] disadvantage of such methods is that the road users that are obscured by the parked cars, trees or light poles, unterbrochner sight due to the vehicle from the camera can not be detected. This means that the camera-based pedestrian protection system in case of imminent collision with hidden road users or reacted too late and so does not provide reliable protection.
[0007] A solution without the above disadvantage has a technology conference in the "driver assistance systems workshop, 2008" in April 2008, already put forward procedure: "Pedestrian protection through cooperative sensors. According to this method detects the road users such as pedestrians using cooperative sensors. According to that road users an active or passive RFID transponder, its relative Position to a vehicle through the vehicle mounted transponder detection device is determined. The transponder receives signals from the detection device located in the vicinity of the vehicle transponder, and analyzes these signals. By measuring the distance signal propagation time is determined by the vehicle at any transponder or the transponder that is transmitted by pedestrians. Use of the multiple antenna system, the transponder detection device is detected and the azimuth angle of the transponder or pedestrian in relation to the longitudinal axis. This allows the current position of each transponder or pedestrian carrying the transponder with him, even with relatively simple vehicle to unterbrochner sight, be clearly established and reliable.
[0008] For predicting a future position or movement of vulnerable road users, the above method comes up with cooperative sensor technology to its limits, however. To predicate a position of the pedestrian at a future time, several successive current positions are recorded. From this plurality of current and past position values, then a future position by means of time-consuming steps of calculation are derived. This leads to an impending collision with pedestrians making in cases can not be predicted in time.
[0009] Another known Method for predicting the position and / or movement of pedestrian of using imaging sensor by evaluating the change in behavior such as viewing angle changes of the pedestrian proved to be due to the excessive time required for image processing and the uncertainty of many factors to be illiquid and also unreliable.
[0010] The present invention is to provide a method which has none of the above-mentioned disadvantages and, for the purposes of maximum effectiveness and minimal false triggering rate optimal solution for pedestrian protection.
[0011] This object is the invention by a method having the characteristic features the patent claim 1. According to the invention of the motion-related data on the road participating vulnerable objects such as pedestrians by means of sensors arranged on the object, for example, a motion sensor detected and transferred from the object side Sende-/Empfängereinheit for the on Sende-/Empfängereinheit. From this data the future positions of movement of the object to one or more time points are predicted.
[0012] The "motion-related data" in this invention include, in addition to the information about the dynamic and kinematic properties of the object, such as speed, acceleration, braking and / or rotation, advantageously and depending on the version and the information on the posture and the orientation of the object such as the diffraction angle and / or the orientation of the object to the compass.
[0013] The term "object" in this document summarizes all non-motorized or motorized weak, "vulnerable" road users, that all road users except motor vehicles with protective body together. The term "vulnerable" is because the non-motorized or motorized weak road users such as pedestrians or motor cyclists due to the lack of protective body in collisions with motor vehicles special injury risks. Among the objects Thus, all pedestrians, especially children, cyclists, wheelchair users, small cyclists, motor cyclists and animals such as dogs and bicycles. Motor bicycles, wheelchairs, roller skates, etc..
[0014] For a simple illustration of the invention will be referred to all vulnerable road users by pedestrians or cyclists. This does not mean that the present invention is limited only to pedestrians and bicyclists. For simplicity, the vulnerable road users also summarized the object Genant.
[0015] The invention is based on the idea that pedestrians, especially children (including animals) as compared to road users in vehicles have relatively high maneuverability and therefore its future position and movement is very difficult predicated.
[0016] One method, each pedestrian a static, mostly circular lounge area as a possible future position of associate and ultimately to determine risk of collision with the vehicles with the assumption that the pedestrian can move at a fixed speed in all directions, led due to the large potential common areas of the pedestrian often false collision warnings and even be hazardous to road traffic (see 4).
[0017] This is physically justified by the fact that put the essential information for predicting the future position of a pedestrian in the change of the speed of the pedestrian or its rotation rate.
[0018] In the future position of the pedestrian to predicate more accurately, the idea, in addition to the consideration of the current position of the pedestrian movement and the current state of the pedestrians involved. The current state of motion is used here in the first place by using a motion sensor mounted on the pedestrian identified. The resulting motion data, for example, give information on whether the pedestrian will stop or continue.
[0019] Advantageously, be more dynamic data such as acceleration, rotation rate of the pedestrian, the pedestrian orientation of the compass using the appropriate sensors arranged on pedestrian identified. These data provide further information on the state of motion of the pedestrian. In conjunction with the current position data can thus predict the potential future position of the pedestrian with significantly higher accuracy.
[0020] Preferably, the acceleration of the pedestrian, both in its longitudinal axis, horizontal axis and measured in the vertical axis and used for prediction. From the measured values of the longitudinal, transverse and vertical accelerations ax, ay and az of the pedestrian will be two-or three-dimensional acceleration vector a -> = (ax, ay)
[0021] According to the rotational movements of the pedestrian on the vertical axis [gamma]. or inflections or inclination of the pedestrian to cross the axis [beta]. possibly also the lateral stretching or inclination of the pedestrian is measured about the longitudinal axis and in the form of a two-or three-dimensional rotation rate vector [omega] -> = ([alpha], [beta]..)
[0022] From the motion data can be predicted and the future state of motion of the pedestrian. Thus, for example from a current rate of 1 m / s, a recent acceleration of 1 m / s <2>, a rotation rate [deg.] From 0 / s from a pedestrian and its orientation can be north to predict with a high probability that he is in a second with a speed of 2 m / s and acceleration of 1 m / s <2> and a rotation rate [deg.] from 0 / s continues to move north. Combining the latest Be wegungsdaten with the current position data, we can predict that the pedestrian north in a second with a high probability of 1.5 meters from the current position.
[0023] In the road safety is not only the pedestrians considered as a potential risk of collision with vehicles, the serious injury risks. Cyclists, wheelchair users, skaters or pets such as dogs are seen in collisions with motor vehicles as a potential victim with serious injury. These various vulnerable road users have different physical or physiological properties and hence are of each other in terms of their physical or be considered separately physical motion properties. Among the pedestrians including children, adults, older people also differ in terms of their physical movement or physical properties from one another.
[0024] Advantageously, therefore, a so-called object model erfindungemässen prediction is used. As the "object model" here is a set of parameters are understood, which is all road users (ie each vulnerable road users) are assigned and contains the following physical or physiological properties of the road user in the form of parameters. These properties include,
- the type of object
- the physiological or physical ability to move the object, including
- the weight,
- the amount,
- the maximum speed,
- the maximum acceleration,
- the maximum rate of turn,
- the movement pattern of the object, namely typical sequence of movements of a road user , for example, bending the upper body before the start running in a child.
[0025] The type of the object's information to which road users it is, namely, whether the object is an adult, a child, an old person with a walker or a child with roller skates, a bicycle, a wheelchair or pet like a dog for example.
[0026] From the type of object can be the maximum radius of motion of the object space, determine the maximum attainable speed, acceleration or rotation rate and so rude. Thus, for example in relation to the same period under the radius of motion of a pedestrian area with a walking aid is much smaller than that of a child with roller skates.
[0027] However, since the values of the maximum range of motion, maximum speed, etc. can vary greatly even within the same group of objects from object to object, further parameters in the form of physiological or physical power of movement of the object added to the set of parameters of the object model . These parameters include weight, body height, the maximum speed, maximum acceleration and / or the maximum rate of rotation of the object. If required, other parameters such as body width can be added. Taking into account the statutory scheme, further information such as chronic diseases that affect reaction or motility of the pedestrian, be added as parameters to the object model. These parameters are more accurate than those from the type of object-assessed roughly parameter values and therefore more meaningful for the assessment of physiological and physical movement of assets. Based on these parameters can, for example, an overweight different (little moving) child of a normal weight.
[0028] In addition to the power of movement of the object, the movement pattern of the object is added to the object model. The pattern of movement, namely the typical sequence of movements of a road user type, information about which follow changes in motion after motion changes currently recorded. For example, prevents a child his upper body (this changes the value of the rate of rotation about the transverse axis [beta].), It can be predicted with a high probability that the child will immediately start moving.
[0029] The parameters for the object model, when introducing the object-side Device operation are set manually. If the object-side device entity, for example, integrated in a mobile phone, so these parameters can easily enter the mobile phone keypad. If the object-side device entity, for example, integrated in a school bag, this will be integrated via an interface connected to a computer. Through keyboard on the computer, the above parameters are then adjusted. If the value of one or other parameter such as, for example, the weight of the pedestrian over time, the appropriate parameter is updated manually.
[0030] From the parameters of the object model and the actual measured motion data (depending on Version also in the time immediately preceding movement data) of the object can be determined advantageously leading indicators for predicting the future position and / or movement of any objects. Based on the current position of the object can be predicated on the basis of the Sun leading indicators of future position of the object. As leading indicators here characteristic changes in the attitude or state of motion of the object to be understood to suggest a change in movement or direction of movement.
[0031] The characteristics of a leading indicator thus obtained can be a reliable prediction of the movement of road users in the early Prädiktionsphase , while, for example, an image-based prediction of the prior art, even a significant change in behavior of road users could not identify.
[0032] Each position and / or movement of a road user at a specified future date should ideally be predicted with an associated position and movement probability. The future position is predicted, preferably in the form of a common area with a corresponding position or probability.
[0033] Preferably, the inventive method is performed using at least two device sub-units, which are separated in the vehicle and are arranged on the object. Both device sub-units each have a Sende-/Empfängereinheit communicate and exchange with each other using wireless data communication between these two Sende-/Empfängereinheiten and data. The on-board Sende-/Empfängereinheit locates the object side Sende-/Empfängereinheit using electromagnetic wave. Preferably, the object-side Sende-/Empfängereinheit an active or passive RFID transponder and a transponder onboard detection device with multi-antenna system. The object-side device sub-unit has a motion sensor, an acceleration sensor for measuring the translational acceleration of the object, a rotation rate sensor for measuring the rotational movement of the object around its vertical axis and / or a magnetic compass sensor for detecting the orientation of the object to the sky direction.
[0034] Advantageously, by the above-mentioned sensors at the location of the transponder every six kinetically relevant parameters, namely the longitudinal, transverse and vertical accelerations ax, ay and az and the rotation about the vertical axis [gamma]. To the transverse axis [beta]. or around the axis [alpha]. In terms of object motion and the orientation of the object to the compass. For simplified version can also only part of the kinetic parameters are used.
[0035] The transponder detection device in Vehicle determines the transponder located in the vehicle environment and determines the distance and azimuth wrap each transponder to the vehicle, preferably in relation to the longitudinal axis. The radial distance between the transponder to transponder detection device and the vehicle will be measured by the signal propagation time measurement. The azimuth angle can be determined using the multiple antenna system, the transponder detection device.
[0036] From the distance and azimuth angles of the vehicle-side device sub-unit's current position of each transponder, and thus each transponder carrier determined ie object relative to the vehicle. Also sends the vehicle-side device sub-unit wake-up signals to the harvested transponders or object-side device sub-units and suggests that, by means of various of the object sensors arranged transaction data such as speed, acceleration, yaw rate and / or orientation to measure the object to the direction and the measured data including the object model parameters and identification code to the vehicle-side device sub-unit to send. The vehicle-side device sub-unit authenticates the transponder by examining the identification codes and determines from the received data early indicators for the objects. Alternatively, the leading indicators and the object identified by the object-side device sub-unit and then to the vehicle-side device sub-unit be sent.
[0037] The vehicle-side device sub-unit then predicted in the leading indicators and the current positions of the objects and the driving dynamic of the vehicle, the future positions of objects relative to the vehicle at specified future dates. The vehicle dynamics data of the vehicle may, for example, the wheel speed, the steering wheel angle, brake pressure, etc., and are preferably derived from the electric system stability.
[0038] Advantageously, calculates the vehicle-side device sub-unit from the vehicle dynamics data is a driving path for the vehicle. At any given time, the device sub-unit in the driving path is calculated each an area as a recreation area of the vehicle at the predetermined time and compares it with the predicted position and the predicted common area of the object to that date. Overlap the two common areas of the vehicle and the object, then an impending collision, and accordingly recognized as measures to avoid the collision causes. If the two surface areas stay close together, so an increased risk of collision is predicted.
[0039] In a less critical case the driver is warned of a possible collision with an optical, acoustic or tactile signal. If the driver or not, the situation is critical, so for example an emergency stop can be initiated. If unavoidable collision mitigation measures to be initiated by accident. In this example, the bonnet in the direction of the impacting object may be raised in advance or, if the pedestrian protection airbags for example, be installed in the vehicle front, airbags are deployed. Here, the transaction data or the parameters of the object model, such as the type, speed, body height, the weight of the object are used for adaptive pedestrian protection measure which submitted during the course of the novel prediction of the object-side transceiver unit for on-board transmitter / receiver have been. Thus, the hood the weight or body size of the object to be lifted or adjusted adaptively, the pedestrian protection airbags to be deployed also adaptively adjusted. This will minimize the consequences of accidents and thus achieve an optimal pedestrian protection.
[0040] Thanks to the determination of the transaction data directly to the road users, the movement-related leading indicators, and during the interruption of sight between the vehicle and the road users can be determined and used for prediction. By detecting early indicators right on the road user movement characteristic changes can reliably over a long period of time and without time delay be detected. The captured early indicators are more reliable than the indirect image based on the vehicle and identified leading indicators. Thus, the inventive method a much earlier and reliable detection of impending collisions with vulnerable road users as possible, the image-based method, and thus offers an optimal solution for pedestrian protection in the sense of maximum effectiveness and minimal false triggering rate.
[0041] The object-side device part units in cell phones, clothing, particularly safety vests, shoes, bags, accessories such as handbags, belts, watches, eyewear, bicycles, bike helmets, bike computers, Scooters, roller skates, strollers, school bags, wheelchairs, crutches, walking sticks, pet collars, etc. are integrated.
[0042] The components of the object-side device entity, such as the transmission / reception unit, the acceleration, rotation rate, magnetic compass sensors, as well as the energy source such as batteries can be installed in a closed case. Alternatively, the components are well separated from each other in the same support as an article of clothing or a bicycle being distributed.
[0043] The power supply for the object-side device entity, including transponders and sensors, with built-in replaceable button batteries or guaranteed by solar cells. Alternatively you can use the generated by the movement of the carrier kinetic energy to electricity. If the object-side device sub-unit within existing portable electronic devices (eg mobile phone or iPod) is integrated, so the device sub-unit to the power of the electronics in these devices already built-in battery or batteries are supplied.
[0044] The use of the above object model, especially the parameters regarding the type of object is hidden, however, a risk that the object is verified false. to name an example, a father with his school bag of his child with an integrated object-side device sub-unit, may not be the father as an adult pedestrian at a lower maneuverability, but falsely verified as a child with a higher maneuverability. Or a pedestrian, who pushes his bicycle with a built-in object-side device sub-unit is falsely verified as a cyclist with a high potential speed. In addition to misuse cases (abuse) can occur, whereby for example, the object-side device sub-units of the carriers, for example, children simply thrown onto the street. Using the data of object-based sensors can be checked, whether it is the movement of the object-side device sub-unit, a realistic movement of a road user, is primarily a pedestrian. As a result, such cases of misuse or improper verification through the use of object-based sensors excluded and false alarms are avoided in the vehicle.
[0045] The use of object-based sensor can therefore also serve to verify the object. The sensor can be carried by the object and the availability of measurement data are used to verify a example for pedestrian or animal characteristic running motion. Thus, the sensor data in the above examples are used by the parameters of the object model object to certain . Check Return the sensor data, which are typical of a running motion of an adult pedestrian, the pedestrian is verified with the school bag of his child in spite of the parameters of the object model of the satchel integrated object-side device entity rather than as a child, but as an adult pedestrian. Similarly, the cyclist pushes his bike, using the sensor data verified correctly as a pedestrian. Do the Data of a typical litter curve suggests the discarding of the transponder, followed by movement data, indicating a quiescent state of the transponder, it can predict with a high probability that the transponder was lost or misused. This may be cases of misuse or improper object verification in the large is excluded.
[0046] The invention is further illustrated by the embodiments with the aid of figures. For simplified description of the invention, the inventive device sub-units as well as those shown in the figures, objects such as vehicles, pedestrians, streets, etc. are shown schematically and simplified. It shows the process of
[0047] 1, a street scene in plan view is presented with generalized as an impending collision between a vehicle and a pedestrian using the inventive method is predicted
[0048] 2 is a detailed description of the On-device sub-unit for carrying out the inventive method,
[0049] 3 is a detailed representation of the object-side device sub-unit for carrying out the inventive method,
[0050] 4A, 4B in two examples The inventive prediction of the position or movement of a pedestrian in relation to a vehicle,
[0051] 5 is an example of prediction of the position or movement of the pedestrian without the use of motion-related leading indicators,
[0052] 6A, 6B, in another two examples The inventive prediction the position or movement of a cyclist or a pedestrian in relation to a vehicle, and
[0053] 7, 8 each in a detail such as the inventive predicting the future position of a cyclist based on the current motion data and in conjunction with associated probabilities.
[0054] The 1 is a section of road as the 500 with a passing vehicle 300, equipped with a novel on-board device entity 120, and a pedestrian 210, an inventive object-side device sub-unit (transponder with integrated sensors 113, 114, 115, 116) 110 carries with him.
[0055] The pedestrian 210 is currently running between two parked cars in a row 411, 412 in the street. By the parked vehicle 411 approaching the line of sight 800 between the moving vehicle 300 and the pedestrian 210 is interrupted. The tracks of the vehicle 300 and 210 of the pedestrian crossing in the middle of the road 600 and there is therefore a risk of collision between two road users 210, 300
[0056] The vehicle-side device entity 120 has to perform the inventive method according Rechen-/Steuereinheit 2, a transceiver unit 122 and a 121 with a transmitting antenna 1212 and a receiving unit 1213, which consists of a multiple antenna system formed with multiple antennas 1214 is on. The transceiver 121 sends the transmit antenna 1212 interrogation signals in the vehicle environment. First you will query signals to all potential in the area located 110 transponders sent nine hundred and first The transponder 110, which are located in the vicinity of the vehicle 300 to be excited by the interrogation signals and send 110 of her hand response signals to the vehicle-side transmitting / receiving unit 121st This response signals include preferably identification number of the transponder 110, which allow the transponder 110 differ from each other clearly, and the parameters of the object model of the object 210, 220, with 110 at the transponder, and the sensor data, by various sensors integrated in the transponder 110 113, ..., were measured 116th The response signals are received by the antenna 1214 of the Multiple Antenna Systems 1213 902nd
[0057] The signal processing unit 1211 then calculates based on the signal delay between the sending 901 and receiving 902 of the signals and the propagation speed of the radial distance of each transponder 110 for vehicle 300. By comparing the signal level of the multiple antennas 1214 of the multiple antenna system in 1213 on several occasions received response signal 902 of each transponder 110 detects the signal processing unit 1211 then the azimuth angle of each transponder 110 to the vehicle longitudinal axis. found in the signal processing unit 1211 Depending on the version filtering, modulation, demodulation, and analog or Digital-/Analogumwandlung Codier-/Decodierung etc. and instead to generate the interrogation signals or from the response signals from the transponders 110 transmitted identification number, object model parameters and sensor data to . win Depending from management is carried out in this signal processing unit 1211, a first selection of the transponder 110, which, according to the distance and azimuth angle of the vehicle 300 in the area of risk. This is especially the transponder 110 and the road users 210, 220 excluded from further collision risk assessment, for example, are behind the vehicle 300 or reside in an area where no collisions between the vehicle transponder 110 and 300 are possible. Preferably, the signal processing unit 1211 instead of the authentication of the transponder 110 on the basis of data transmitted by the transponders 110 number.
[0058] The Rechen-/Steuereinheit 122, a data recovery unit 1221, a unit for tracking and Datenfusionierung 1222, a unit risk calculation 1223 and 1224 includes a decision unit to trigger the warning signals on detection of collision risk.
[0059] The data extraction unit 1221, the received data packets divided by the signal processing unit 1211 is already pre-selected transponders 110 in the Sensor data and object model data and transmits that data to the Datenfusionierungs and Tracking Unit 1222nd The Datenfusionierungs and Tracking Unit identified in 1222 then from the parameters of the object model and sensor data, the type of transponder support (ie the motorist) 210, 220 and extrapolated from this data and based on the current positions of the transponder 110 relative to the vehicle 300 potential living area of the road user, 210, 220 at predetermined future dates. Here are preferably first identified early indicators. Preferably, the predicted residence areas shall be provided with associated probabilities.
[0060] Furthermore, determines the Datenfusionierungs and tracking unit 1222 based on the driving dynamics of the components of the electronic stability system 310 transmitted data of the vehicle 300 to the driving path of the vehicle 300 and the common areas of the vehicle 300 at the same times in the future.
[0061] Optionally, the Datenfusionierungs and tracking unit determines 1222 from the sensor data of the imaging sensors on the vehicle 300, such as vehicle front camera 331 the current position data of the collected road users 210, 220 and this data compares with the use of multiple antenna system in 1213 determined position data. Thus, the system reliability is increased again.
[0062] In the risk assessment unit 1223 will be in the data fusion and tracking unit 1222 identified potential common areas of the vehicle 300 and the road users 210, 220 compared to each other. Overlap the stay of the vehicle 300 with a common area of a road user 210, 220 at a future date, predicted an imminent collision risk. Furthermore, risk assessments are carried out on possible collisions taking into account the probabilities of the predicted location areas of the market participants 210, 220, 300. It used to be optional and the vehicle dynamics data such as speed, yaw rate, or weight of the vehicle 300. these data, evaluated and assessed how likely a collision occur and how severe the collision or injury is involved in the collision of road user 210, may be 220th To estimate the collision or the severity of potential injury to the parameters of the object model and the sensor data of the object-side sensor 113, 114, 115, taken 116th
[0063] The result of the risk assessment shall be in the form of, for example collision risk values, accident severity and injury risk values to the trigger decision unit 1224th The trigger decision unit 1224 then selects the determined collision risk values, accident severity or injury risk values according to the appropriate counter measure to avoid collision or to reduce accidents. Among the measures include eg auditory, visual or tactile warning signals for the driver of the vehicle 300, adjusting the speed of the vehicle by interrupting the fuel supply, autonomous braking, setting the parameters for lifting the hood or deploying the air bag for pedestrian protection, and so on.
[0064] Optionally, the trigger decision unit 1224 generates a warning signal for the market participants 210, 200. This warning signal can, for example acoustic horn or flashing on his vehicle 300. Advantageously, however, this warning signal via the transmission antenna 1212 to the transponder 110 of the affected Road user, 210, 220 sent nine hundred and first Receives the transponder 110, such a warning signal from the vehicle 300, so does the transponder 100 by emitting an audible warning tone, for example, by vibration or the market participants 210, 220 from a dangerously approaching vehicle 300 or a risk of collision with the vehicle 300 carefully.
[0065] The object-side device entity 110 has 3, a transceiver unit according to 111, the unit with the vehicle-Sende-/Empfangs 121 on electromagnetic wave communication and data exchanges 903, 904 and is embodied in the form of a transponder, a control unit 112, in each case a motion 113, 114-Bescheunigungs, rotation rate 115, and magnetic compass sensor 116. Furthermore, the device sub-unit 110 has a memory 117 and a synchronization unit 118 to synchronize the data communication.
[0066] The Sende-/Empfangeinheit 111 includes an antenna 1111, an antenna switch 1112 for switching between a known type transmit and receive mode or transmit / receive antenna 1111, a signal receiver 1113 and a signal transmitter 1114th Preferably, the directional coupler 111 Sende-/Empfangeinheit and circulators for separating transmit and receive signals.
[0067] Using the sensors, especially inertial sensors 113, 114, 115, 116, the object-side device entity 110, the detected current motion data of the transponder carrier 210.
[0068] Depending on the design and quality requirements included the acceleration sensors 114 or the rotation rate sensor 115, the acceleration or the rotation movements of carrier 210, 220 three-, two-, or one-dimensional.
[0069] By three-dimensional measurement of acceleration is primarily the orientation of the acceleration sensor 114 and the sensors should be installed 113, 114, 115, 116 and the control unit 112 and the transceiver unit 111 in one and the same housing, then the orientation of the whole object-side device entity 110 are recorded. The detection of the direction of the acceleration sensor 114 or the device 110 is so important because the acceleration sensor 114 or the device 110 by the carrier or road users 210, 220 are continuously observed in different positions, so it provides for the same acceleration movement of carrier 210, 220 different values.
call [0070] As an example, the acceleration sensor 114 thus the object-side device sub-unit 110 is integrated into a phone that leads the pedestrian 210 with itself, then the sensor 114, for example, when calling from the pedestrian 210 held in an upright position and after the phone calls, for example, brought in a bag in a horizontal position. In these different positions, the sensor 114 itself in the uniform accelerating pedestrian 210 different acceleration rates.
[0071] Thus, the acceleration sensor delivers 114 in the upright position, a first three-dimensional acceleration value a-> 1 = (AX1, AY1, AZ1)
[0072] Although the pedestrian has 210 accelerates uniformly over the entire time, the soft values of the two-dimensional acceleration vectors - seen in pairs based on the sensitivity or measurement axes - each other strongly. So soft, the two readings ax1 and ax2 for x-axis, ay1 and a2 for y-axis, AZ1 and AZ2 from each other for z-axis.
[0073] This is due to the fact that the orientation of the acceleration sensor 114 and the axes of sensitivity of the acceleration sensor 114 has been amended over time. It is therefore necessary to update the orientation of the sensors 113, 114, 115, 116 times.
[0074] The alignment of the inertial sensors, especially 113, 114, 115, 116 is advantageously based on the acceleration sensor 114, for example, determined in a quiescent state of the pedestrian 210, wherein a vector sum of a -> from three measured Vector components ax, ay, az forms. Indicates the amount of the vector sum of a -> approximate the value of Erdschwerebeschleunigung 9.81 m / s <2> shows, the direction of the vector sum of a -> the direction of the Erdgravitationskraft. In combination with the measured values of the magnetic compass sensor 116 can be determined from the direction of Earth's gravity, the orientation of the acceleration sensor 114th
[0075] The rotational movement of the pedestrian 210 will also be depending on the design and quality requirement three, recorded two or one dimensional. The rotational movement about the vertical axis [gamma]. 210 of the pedestrian is like the lateral acceleration ay information on whether the pedestrian 210 its direction changes. The rotational movement of the pedestrian 210 to the cross-axis [beta]. , Namely the slope of the pedestrian 210 is, for example, indicate that the pedestrian will start running immediately.
[0076] The object-side device unit 110 is part of an integrated energy storage (power source) 119, which for example consists of one or more of kinetic (motion) energy of the 210 pedestrian automatically rechargeable battery cells that power. Thus, the device entity 110 are considered in the 3 as an active transponders with several integrated sensors 113, ..., 116.
[0077] The object-side device entity 110 is preferably equipped with energy-saving mode. Thus, for example, the device sub-unit 110 is placed in a low-power sleep mode as long as the unit 110 receives a signal from outside and the motion or acceleration sensor 113, 114 registered no movement or acceleration of the pedestrian 210th
[0078], the transceiver unit 111 receives signal from the outside 904, it sends a wake-up signal 1191 to the current source 119 and sensors 113, ..., 116 and other units 117, 118. Alternatively, you can "wake up" the device sub-unit 110 of the motion or acceleration sensor 113, 114. Join the movement or acceleration sensor 113, 114 pedestrian movement or acceleration of the 210, it sends wake-up signal 1192 , 1193 to the current source 119 and other units 111, 112, 115, ... 118.
[0079] The object-side device of unit 110 also has a memory 117, are stored in the identification code of the device entity 110, and parameters of the object model for the transponder carrier 210, 220. The parameters of the object model can be updated manually during operation of the transponder 110th For this purpose, the device unit 110 of an optional interface that allows the device part unit 110, for example, can exchange data with a computer.
[0080] The object-side device entity 110 is optional depending on the version equipped with a synchronization unit 118. The synchronization unit 118 serves to protect the time to synchronize the signal transmission between the vehicle-and object-side device sub-units 110, 120 in time division multiplexing to provide. Alternatively, the data communications in code division multiplex or frequency division multiplexing are accomplished. This attention to the vehicle-side device and object-part units 110, 120 corresponding components for performing the code division multiplex or frequency division multiplexing.
[0081] Figures 4A, 4B each show an example of what can be an impending collision or a non-critical situation by means of the inventive method to predict clear. The two 4A, 4B show two similar road pictures like the 1. On the road 500, a vehicle approached 300 approached in an east direction. Between two parked vehicles 411, 412 is a pedestrian 210, where the pedestrian 210 in the 4A North and 500 in the direction of the road running into road 500 and the pedestrian 210 in the 4B south or back to the road 500 from the road 500 runs away. We assume that the two pedestrian 210 moves in the two 4A, 4B with the same velocity v0 and acceleration a0. The alignment of the pedestrian 210 shows in the 4A North and the pedestrian 210 in the south in the 4B. This transaction data such as speed v0, Acceleration a0 and orientation to the direction from sensors 113, ..., 116 of the device entity 110, with 210 at the pedestrian is measured and transmitted to the vehicle-side device of unit 120. These data to calculate the device unit 120 of leading indicators on the state of motion of the pedestrian 210th Based on the current position of the pedestrian device 210, the entity identified 210 of the leading indicators and the driving dynamics of the vehicle data 300, the future positions and / or common areas of the vehicle 300 and walking 210 to one or more predetermined points in time.
[0082] Thus, each at a time t1, a Position or a common area 302 for the vehicle 300 and a position or stay portion 212 for the current in the north pedestrian predicted 210, which partially overlap 601 (see Figure 4A). The overlapping area 601 indicates an imminent collision of the vehicle 300 with the pedestrian 210th Consequently, the Rechen-/Steuereinheit sends 122 of the vehicle-side device entity 120, a control signal to the pedestrian protection system 320 and causes it to take appropriate measures.
[0083] Unlike the example in 4A are the predicted positions and / or common areas 302, 300 and 214 of the vehicle when the pedestrian 210 in the illustrated 4B Example separated. Although the pedestrian 210 dangerously close to the road 500 and the driving path 301 of the vehicle 300 is, in this case is no risk of collision, since the pedestrian 210 clearly from the road 500 and from the driving path 301 and, hence, of the predicted common area 302 of the vehicle 300 is moving away. The determined rate of rotation as early indicators suggest an intention not to 210 of the pedestrian way to turn around and run into the street 500th
[0084] The examples in Figures 4A, 4B are the advantages of the inventive method clearly. By taking into account the movement of data from the pedestrian 210, wherein the transaction data directly on the pedestrian covered 210, can predicate the future position of the pedestrian 210 to a predetermined time with high accuracy.
[0085] A vague and large predicted to enthaltsbereich 216 of the pedestrian 210 with the assumption that the pedestrian moves 210 in all directions with equal speed or acceleration, often leads to incorrect prediction of an imminent collision, as in the illustrated fifth
[0086] Although the pedestrian 210 moves away from the road 500 and obviously not collide with the vehicle 300 will be due to the lack of information about the state of motion - for example, the direction of movement 213 - 300 of the pedestrian incorrectly overlap 602 of the future recreation areas 302, 216 of the vehicle 300 and 210 of the pedestrian identified. Based on a false conflict and predicted it will be caused unnecessary even dangerous pedestrian protection measures to avoid an alleged collision.
[0087] The 4B, 5, the differences between a motion-based prediction and a prediction without pre-drawing of the transaction data of carrier 210 and thus the benefits of the novel prediction clearly show.
[0088] Figures 6A, 6B show further examples of the novel prediction. In the two 6A, 6B, respectively, a section of an east-west directed main road 500 with a mapped south of this main road 500 street which opens 510th In the main road 500 is approaching a vehicle 300 approaching from the west to east direction. From the street a cyclist travels 510 220 (6A) and is a pedestrian 210 (6B) from the south to the north into the main race 500th The sight of the vehicle 300 to 210 cyclists 220 and pedestrian cut through the trees 420th
[0089] The bicycle 220 has higher speed than the pedestrian 210th Consequently, the invention predicted common area 222 of the 220 cyclists are much greater than the Common area 218 of the pedestrian 210th The predicted common area 222 of the 220 cyclist lies partly in the same time predicted for the common area 302 of the vehicle 300. This points to an impending collision of the rider 220 with the vehicle 300. Appropriate action to avoid collision must be initiated. The predicted common area 218 of the pedestrian 210 is still way off from the living area 302 of the vehicle 300. This means that there is no risk of collision between the vehicle 300 and the pedestrian 210th The vehicle 300 may then pass unchecked.
[0090] The 7, 8, the prediction according to the invention show detail. According to the 7, the movement data of the current state of motion of the cyclist 220 as speed, acceleration, yaw rate, etc. used for predicting the future position of the cyclist's 220th Thus, for example, a possible common area 2221 of the rider 220 to a given point in time t1 at a current speed of v1 = 2 m / s, acceleration of a 1 = 0 m / s <2> and rotation rate of [omega] 1 = 0 [deg.] / s predicted. At the same speed of V1, a higher acceleration of a2 = 2 m / s <2> and the same rotation rate of [omega] 1, a relatively large common area is predicted 2222nd At a higher speed of v2 = 4 m / s, an acceleration of a 1 = 0 m / s <2> and rotation rate of [omega] 1 is again predicted an even larger lounge 2223rd
[0091] The 8 shows the predicted common areas 222A, 222B, 222C, 222d, 222E, 222f of the rider 220 with the associated probability at a given time t1 with current transaction data: v1 = 2 m / s, a2 = 2 m / s <2>, [omega] 1 = 0 [deg.] / s. For predicting the parameters of the object model, for example, the weight of the rider 220 can be used together with the bike. Accordingly, the cyclist is 220 for the time t1, for example, with a probability of 80% for 222C, each with a probability of 5% in 222B or 222d. With residual probability of 10% could be the cyclists in one of the areas 222A, 222E, or 222f are in the range predicted by the outside lounge area 222nd The area 222C is therefore predicted with such a high probability, because the parameters of the object model in particular with respect to the weight of the rider 220 on a moment of inertia suggests that prevents a rapid change in direction or abrupt braking of the bicycle.
[0092] For even more accurate prediction of the position of the cyclist, the environment information 220 can be used around the cyclist 220th The probabilities for the predicted Common areas can then be modified according to the ambient information. If, for example, in 222d, a tree, then the probability for this section 222d to reduce to zero because it is assumed that the cyclist will not drive 220 into a tree.
[0093] Some of the information environment, for example, the information on road condition, etc. of the infrastructure in the vehicle environment, to the weather. This information is retrieved from the vehicle-side device of unit 120, for example, the navigation data or reports th of weather stations.
numeral list
110: object-side device sub-unit, transponders with integrated sensors 113, 114, 115, 116
111: sending / receiving unit of the object-side device part unit 110
1111: transmitting /
1112: combiner
1113: signal receiver Rx
1114: signal transmitter Tx
112: Rechen-/Steuereinheit the object-side device part unit 110
1121: Identification auditor
1122: signal modulation and demodulation
1123: data mixer
113: motion sensor
114: accelerometer
115: yaw-rate sensor
116: magnetic compass sensor
117: memory
118: synchronization unit
119: power source
1191: wake-up signal from the transmitting / reception unit 111
1192: wakeup from the motion sensor 113
1193: wake-up signal from the acceleration sensor 114
120: Onboard device entity
121: sending / receiving unit of the On-device part unit 120
1211: signal processing unit of the vehicle-side transmission / reception unit 121
1212: transmitting antenna of the vehicle-side transmission / reception unit 121
1213: multiple antenna system of the vehicle-side transmission / reception unit 121
1214: receiving antenna of the vehicle-side transmission / reception unit 121
122: Rechen-/Steuereinheit of the On-device part unit 120
1221: data recovery unit
1222: data fusion and tracking unit
1223: Risk Assessment Unit
1224: trigger decision unit
210: Pedestrian
211, 213, 215 direction of the pedestrian 210
212, 214,: According predicted common area of the pedestrian 210 to a given future point in time t1
216: Excluding the current motion state predicted common area of the pedestrian 210 at the time t1
218: According predicted common area of the pedestrian 210 given at a future point in time t2
220: cyclists
221: direction of travel of the cyclist
220 222: According predicted stay of the rider 220 to date t2
2221: According predicted stay of the rider 220 to the time t2 at a Radfahrtgeschwindigkeit of v1 (v1 = 2 m / s), Radfahrtbeschleunigung of a1 (a1 = 0 m/s2)
2222: According predicted stay of the rider 220 to at a time t2 Radfahrtgeschwindigkeit of V1, Radfahrtbeschleunigung of a2 (a2 = 2 m/s2)
2223: According predicted stay of the rider 220 to the time t2 at a Radfahrtgeschwindigkeit from v2 (v2 = 4 m / s), Radfahrtbeschleunigung of a1 ( a1 = 0 m/s2)
222A, ..., 222f: According predicted stay of the rider 220 with associated Probabilities
300: Vehicle
301: direction of travel of the vehicle 300
302: common area of the vehicle 300 at a time t1 at a constant speed of travel
310: Electronic stability system
320: pedestrian protection control system
331: Fahrzeugfronkamera
332: image processing unit
411, 412: Park End Vehicles
420: trees
500: road to vehicle traffic
501: road marking
510: Since road that joins the road 100
600, 601, 602,: predicted collision between the vehicle 300
603 and the pedestrian 210 and cyclists
220 800 : Due to a parked vehicle 411 broken sight between the vehicle 300 and the pedestrian 210
901: signal to the transponder
902: signal from the transponder
903: signal to the vehicle
904: signal from the vehicle
QUOTES INCLUDED IN THE DESCRIPTION
[ 0094] This list of references cited by the applicant was automatically generated and is included solely to better inform the reader. The list is not part of the German patent or utility model. The DPMA is not responsible for any errors or omissions.
cited non-patent literature
[0095]
- Technology Conference: Workshop Driver Assistance Systems 2008 "in April 2008 it is transmitted procedure:" Pedestrian protection through cooperative sensors "[0007]
method and apparatus for predicting the position and / or movement of an object relative to a vehicle
Technical and scientific translations
0 comments:
Post a Comment