CN102259628B - Method And Apparatus For Collision Detection In Vehicles - Google Patents
Method And Apparatus For Collision Detection In Vehicles Download PDFInfo
- Publication number
- CN102259628B CN102259628B CN201110124872.5A CN201110124872A CN102259628B CN 102259628 B CN102259628 B CN 102259628B CN 201110124872 A CN201110124872 A CN 201110124872A CN 102259628 B CN102259628 B CN 102259628B
- Authority
- CN
- China
- Prior art keywords
- signal
- oscillator signal
- vehicle
- collision
- time domain
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 69
- 238000001514 detection method Methods 0.000 title claims abstract description 30
- 238000012545 processing Methods 0.000 claims abstract description 22
- 238000012546 transfer Methods 0.000 claims description 22
- 238000006243 chemical reaction Methods 0.000 claims description 15
- 238000000926 separation method Methods 0.000 claims description 12
- 238000001914 filtration Methods 0.000 claims description 10
- 238000001228 spectrum Methods 0.000 claims description 10
- 230000006870 function Effects 0.000 description 29
- 238000004458 analytical method Methods 0.000 description 28
- 230000008569 process Effects 0.000 description 19
- 230000005284 excitation Effects 0.000 description 12
- 238000010586 diagram Methods 0.000 description 7
- 230000001133 acceleration Effects 0.000 description 5
- 230000005540 biological transmission Effects 0.000 description 5
- 238000004590 computer program Methods 0.000 description 4
- 238000010276 construction Methods 0.000 description 4
- 230000009466 transformation Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 3
- 238000005086 pumping Methods 0.000 description 3
- 229920002994 synthetic fiber Polymers 0.000 description 3
- 206010019133 Hangover Diseases 0.000 description 2
- XAGFODPZIPBFFR-UHFFFAOYSA-N aluminium Chemical compound [Al] XAGFODPZIPBFFR-UHFFFAOYSA-N 0.000 description 2
- 229910052782 aluminium Inorganic materials 0.000 description 2
- 239000004411 aluminium Substances 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 230000015572 biosynthetic process Effects 0.000 description 2
- 230000006378 damage Effects 0.000 description 2
- 238000013461 design Methods 0.000 description 2
- 239000000284 extract Substances 0.000 description 2
- 230000007274 generation of a signal involved in cell-cell signaling Effects 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- OKTJSMMVPCPJKN-UHFFFAOYSA-N Carbon Chemical compound [C] OKTJSMMVPCPJKN-UHFFFAOYSA-N 0.000 description 1
- 229920000049 Carbon (fiber) Polymers 0.000 description 1
- 241000221931 Hypomyces rosellus Species 0.000 description 1
- 229910000831 Steel Inorganic materials 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 229910052799 carbon Inorganic materials 0.000 description 1
- 239000004917 carbon fiber Substances 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000005336 cracking Methods 0.000 description 1
- 230000001186 cumulative effect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 230000004069 differentiation Effects 0.000 description 1
- 238000002592 echocardiography Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 239000000463 material Substances 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- VNWKTOKETHGBQD-UHFFFAOYSA-N methane Chemical compound C VNWKTOKETHGBQD-UHFFFAOYSA-N 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000010355 oscillation Effects 0.000 description 1
- 230000000704 physical effect Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 239000007787 solid Substances 0.000 description 1
- 239000010959 steel Substances 0.000 description 1
- 238000003786 synthesis reaction Methods 0.000 description 1
- 238000012731 temporal analysis Methods 0.000 description 1
- 238000000700 time series analysis Methods 0.000 description 1
Images
Classifications
-
- B—PERFORMING OPERATIONS; TRANSPORTING
- B60—VEHICLES IN GENERAL
- B60R—VEHICLES, VEHICLE FITTINGS, OR VEHICLE PARTS, NOT OTHERWISE PROVIDED FOR
- B60R21/00—Arrangements or fittings on vehicles for protecting or preventing injuries to occupants or pedestrians in case of accidents or other traffic risks
- B60R21/01—Electrical circuits for triggering passive safety arrangements, e.g. airbags, safety belt tighteners, in case of vehicle accidents or impending vehicle accidents
- B60R21/013—Electrical circuits for triggering passive safety arrangements, e.g. airbags, safety belt tighteners, in case of vehicle accidents or impending vehicle accidents including means for detecting collisions, impending collisions or roll-over
- B60R21/0136—Electrical circuits for triggering passive safety arrangements, e.g. airbags, safety belt tighteners, in case of vehicle accidents or impending vehicle accidents including means for detecting collisions, impending collisions or roll-over responsive to actual contact with an obstacle, e.g. to vehicle deformation, bumper displacement or bumper velocity relative to the vehicle
Landscapes
- Engineering & Computer Science (AREA)
- Mechanical Engineering (AREA)
- Air Bags (AREA)
- Traffic Control Systems (AREA)
- Measurement Of Mechanical Vibrations Or Ultrasonic Waves (AREA)
Abstract
The invention provides a method for collision detection in vehicles, comprising a step of providing a vibration signal in the frequency range, wherein the vibration signal is a superposition of signal components rightly by an impact on the vehicle caused and signal components excited by the impact on the vehicle structure. Furthermore, the method comprises a step of applying a non-linear processing instructions provided on the vibration signal to obtain a processed vibration signal. Also, the method includes a step of transferring the processed vibration signal in the time domain with a quefrency to obtain a time-domain vibration signal. The method further comprises a step of separating the time domain vibration signal to obtain first signal components with a low quefrency and second signal components with a high quefrency. At last, the invention also includes information about the impact on the vehicle and / or the impact of the vehicle using the first and / or second signal component.
Description
Technical field
The present invention relates to method for the impact detection of vehicle, for device and the corresponding computer program of the impact detection of vehicle.
Background technology
Preferably use the acceleration information about vehicle for the orthodox method of impact detection, in order to the collision to object of the collision of this vehicle or this vehicle detected.This mode is subject to following restriction, is being arranged on before sensor on vehicle can react, and at first the chassis of this gross vehicle must surpass predetermined acceleration rate threshold.
DE 10 2,004 038 984A1 show the device for impact detection, and this device has at least one first sensor, analyze the signal of this first sensor, with at least one second, the authenticity examination of the signal of the sensor of collision detection.
Summary of the invention
By the present invention proposes improved method and the improved device of a kind of impact detection for vehicle as described as independent claims, a kind of corresponding computer program has been proposed in addition under this background.Provided favourable design plan in each dependent claims and following specification sheets.
The present invention has realized a kind of method of the impact detection for vehicle, the method has the step that the oscillator signal in frequency domain is provided, wherein, described oscillator signal means the stack with the signal section generated in vehicle structure by this collision by the signal section that collision directly caused to described vehicle; The method also have by the Nonlinear Processing rule application to described provided oscillator signal in order to obtain the step of treated oscillator signal; The method also has that the oscillator signal that this is treated is converted to the inverted frequency time domain in order to obtain the step of time domain oscillator signal; The method also has this time domain oscillator signal is separated into to first signal component with low inverted frequency value and has the step of the secondary signal component of high inverted frequency value; And the method also have by use this first and/or the secondary signal component detection about the step of the information of the collision to the collision of this vehicle and/or this vehicle.
Describe a kind of standard time in seconds with inverted frequency, use this standard time when the signal by treated in frequency domain is converted to time domain.Determine time domain based on inverted frequency.This standard time or inverted frequency are not corresponding to the actual time but corresponding to the virtual time determined by conversion, commonly using when carrying out cepstrum analysis as technical personnel.Component of signal with low inverted frequency value is regarded as having the component of signal of low inverted frequency value for other component of signals of time domain oscillator signal.These first signal components usually form one group of a plurality of value in succession, and itself and one group have the high inverted frequency value, value equally in succession separately.
The present invention has also realized a kind of device of the impact detection for vehicle in addition, this device has following characteristics: for the device of oscillator signal that frequency domain is provided, wherein said oscillator signal means the stack with the signal section encouraged in vehicle structure by this collision by the signal section that collision directly caused to described vehicle; For by the Nonlinear Processing rule application to described provided oscillator signal in order to obtain the device of treated oscillator signal; For by this, treated oscillator signal is converted to the inverted frequency time domain in order to obtain the device of time domain oscillator signal; For separating of described time domain oscillator signal in order to obtain the first signal component with low inverted frequency value and the device with secondary signal component of high inverted frequency value; And for passing through use one or more described first and/or the secondary signal component detection device about the information of the described collision of the described collision to described vehicle and/or described vehicle.
In addition, this invention has also realized having the computer program of program code, for carry out preceding method when on control setup, implementing this program.In particular, this computer program can be stored on the machine sensible carrier such as semiconductor memory, harddisk memory or optical memory, and be used to, when program is implemented on the device of control setup or correspondingly design, carry out the method according to aforementioned embodiments.
The present invention is based on following understanding, understanding according to the component of signal of the low frequency (low inverted frequency) of power cepstrum and/or high frequency (high inverted frequency), can realize the very good differentiation of object to colliding part and/or the collision velocity of vehicle, as the information about collision.In particular, in future, come into the market more, in the vehicle with the lightweight construction such as carbon fiber or aluminium, the high and low frequency signal section occurs when the element fracture of these lightweight constructions, it can be applied to differentiating different impact site and/or the speed of object to vehicle.Can for example by transfer function, cause the low frequency signal part in this analysis, this low frequency signal is partly produced because of the caused high frequency pumping of fracture by vehicle structure.This in the situation that the typical signal curve while knowing object to the predetermined impingement position of vehicle and/or projected angle of impact caused following may, the signal section (signal section of low inverted frequency) with low inverted frequency value can be determined about object to the actual impact site of vehicle and/or the indication of projected angle of impact.Signal section that can be by having low inverted frequency value and the relatively realization of predetermined inverted frequency characteristic should be determined.
An important aspect of the present invention utilize exactly the dissociative excitation event with for example, to the transfer function of this actuation event (can be acceleration signal and/or solid acoustic signals), and utilize resolution subsequently.
According to the embodiment of the present invention, in the windowing of described incoming signal in carrying out time domain during step is provided and subsequently the described incoming signal through windowing is converted into to frequency domain, in order to obtain the described oscillator signal in frequency domain.The described incoming signal through windowing to the conversion of frequency domain can be Fourler transform, discrete Fourier transformation, Fourier series or continuous fourier transform.Same wavelet transform is also possible.Oscillator signal in frequency domain can be frequency spectrum and in the intensity of single frequency shown in time domain.This embodiment of the present invention provides following advantage, and the method by the technical maturation in numerical value ground can obtain the signal that can process well in frequency domain.
According to other embodiment of the present invention, in step is provided, can determine the incoming signal in time domain portions of the spectrum energy and subsequently by this portions of the spectrum LPF, in order to obtain the oscillator signal in frequency domain.The frequency band of the incoming signal in the time domain that can obtain by previous filtering, by a time period to the cumulative or integration of the signal strength in this frequency band, carry out the determining of energy of the portions of the spectrum of the incoming signal in time domain.Can realize the signal smoothly formed by single value by LPF subsequently.This embodiment of the present invention provides following advantage, can obtain the signal in frequency domain in very simple mode thus.
Other embodiment according to the present invention can be provided by the logarithmetics of provided oscillator signal in applying step, in order to obtain treated oscillator signal.The logarithmetics that provided oscillator signal is provided can be converted to the part of the single stack in oscillator signal the addition of part by the multiplication of part, thus the single part of the signal simply in separation frequency domain.Logarithmetics be computation-intensive but be very accurately.
In other embodiment of the present invention, the comparison of the value that described provided oscillator signal also can be provided in described applying step and value from predetermined look-up table, and according to distributing to value described look-up table and that store, carry out the described value that described provided oscillator signal is provided by predetermined value, in order to obtain described treated oscillator signal.Substitute as the logarithmetics of Nonlinear Processing rule and also can with limited calculated performance as soon as possible, process in real time the process strict to time requirement by look-up table.Made up the less loss of detailed logarithmetics on accuracy with respect to provided signal by shorter signal processing time.
According to an embodiment of the invention can executive signal in switch process conversion, especially discrete fourier inverse transform, in order to obtain described time domain oscillator signal.The discrete fourier inverse transform corresponding to the signal synthesis by frequency domain to the signal in time domain.By the correction that should be used for advantageously realizing the information in treated oscillator signal of aforesaid logarithmetics or the Nonlinear Processing rule based on look-up table, this has just realized the simple analysis of signal section.Signal in the time domain produced has international unit second, and its correct explanation is as the frequency through conversion.
According to an embodiment of the invention, in switch process, can treated oscillator signal be converted to the time domain oscillator signal by using bank of filters.In the situation that avoid being converted to by the treated oscillator signal numerical value ground from frequency domain, exactly the time domain oscillator signal, and, in the situation that the simple method based on bank of filters of passing through produced thus substitutes transform method accurately thereupon, can avoid expensive computation process and reduce signal processing time.Can window function be applied to described time domain oscillator signal in separating step according to other embodiment of the present invention, in order to obtain, there is the first signal component of low inverted frequency value and the secondary signal component with high inverted frequency value.Separate by signal that with simple mode and method, the different scope of signal to be separated from each other be possible.In the time domain oscillator signal, different information is comprised in different inverted frequency values.Yet the analysis of the different information of single inverted frequency value needs the separation of time domain oscillator signal.
According to other embodiment of the present invention, can carry out the signal filtering of described time domain oscillator signal in described separating step, there is the first signal component of low inverted frequency value and the secondary signal component with high inverted frequency value in order to obtain.By using filter function, with the separation that method completes the time domain oscillator signal simply, cheaply, be possible.Because the signal range of transfer function and excitation on the inverted frequency scale mutually away from, so the intersection in zone or that cause by the separation of inverted frequency value, fade out (Ausblendung) central area exert an influence to obtained result hardly.
According to another embodiment of the present invention, in detecting step, the first signal component with low inverted frequency value can be compared with predetermined inverted frequency curve, wherein said predetermined inverted frequency curve means the information about colliding part and/or crash type, and the secondary signal component that wherein has a high inverted frequency value means the information about impact strength by the peaked quantity in the inverted frequency interval and/or peaked intensity.Can pass through the emulation of the collision on virtual vehicle or pass through the record during standardized collision, for example the record under laboratory condition obtains predetermined inverted frequency curve.At this, this predetermined inverted frequency curve means definite colliding part of vehicle and/or definite collision velocity.In comparison procedure, the component of signal with low inverted frequency value can be compared with the value of predetermined inverted frequency curve according to the conformability of maximum possible, in order to extract about the information of colliding part and crash type and realize thus locating as much as possible accurately and classification of collision.The component of signal that has low inverted frequency value at this can be assigned to such inverted frequency curve, and this inverted frequency curve determines that to relevant the actual of inverted frequency curve the value of (that is, based on measuring) is the most similar.Can obtain in addition the peaked ratio of the component of signal with high inverted frequency value.In the inverted frequency interval, the maxim of high quantity can also be assigned to high impact strength, and in contrast, in relevant inverted frequency interval, the maxim of low quantity means low impact strength.With it similarly, have the peaked sum of the signal section of high inverted frequency value by analysis, the information obtained about impact strength is possible.Information based on about impact strength, colliding part and/or crash type (mode), can activate the active safety element in vehicle, for example personnel's constrained system or passenger restraint system targetedly.Similarly, vehicle in the relevant zone of collision, definite promptly opening circuit of Vehicular system be also possible.
In addition, the present invention has also realized a kind of control setup (that is, device), and this control setup has unit, and it is constructed in order to carry out or realize the step according to method of the present invention.Can also be fast and solve efficiently the task as basis of the present invention by usining this enforcement modification of the present invention that control setup is form.
The electric apparatus that at this, control setup/device can be interpreted as to the processes sensor signal and export accordingly control signal.This control setup can have can be with the interface of hardware and/or software mode structure.In the structure with hardware mode, described interface can be for example the part of so-called ASIC system, the various functions that described ASIC system comprises control setup.Yet also possible that, described interface is distinctive, integrated circuit or is comprised of separate member at least in part.In the structure with software mode, described interface can be software module, and described software module for example is present on the microcontroller that also has in addition other software modules.
The accompanying drawing rule
Below according to appended accompanying drawing, exemplarily explain in detail the present invention.In accompanying drawing:
Fig. 1 shows for the block scheme impact detection of vehicle, the first embodiment foundation method of the present invention;
Fig. 2 shows and is produced to the collision of vehicle by object in the situation of signal, the schematic diagram of the signal distortion by transfer function;
Fig. 3 show according to for the impact detection of vehicle, according to the figure of the example results after the step of application Nonlinear Processing rule of the embodiment of method of the present invention;
Fig. 4 show according to for the impact detection of vehicle, according to the figure of the example results after the method step that oscillator signal is back to the time domain oscillator signal time domain from frequency domain of the embodiment of method of the present invention;
Fig. 5 shows for the diagram of circuit impact detection of vehicle, another embodiment foundation method of the present invention;
Fig. 6 shows according to an embodiment of the invention, for the diagram of circuit of the method for the simplification of the impact detection of vehicle; And
Fig. 7 shows the block scheme of the embodiment as device of the present invention.
The specific embodiment
Identical or similar element means by identical or similar Reference numeral in the accompanying drawings, wherein no longer repeat specification.In addition, a lot of features that the explanation of the figure in accompanying drawing, figure and claim comprise combination.It will be clear to someone skilled in the art that and can consider individually these features or it is combined into to combination other, that do not describe in detail at this at this.In this subsequently external explanation, by using different sizes and size to be set forth the present invention, wherein the present invention can not be understood to be limited to this size and size.In addition, can implement steps of a method in accordance with the invention repeatedly and with the order different from described order.If embodiment be included between First Characteristic and Second Characteristic " and/or " connect, this can be interpreted as, according to an embodiment, this embodiment had both had First Characteristic, there is again Second Characteristic, and according to another embodiment, this embodiment only has First Characteristic, is exactly only to have Second Characteristic.
According to the expert, estimate, the new-type vehicle structure designed with lightweight construction obviously is different from the traditional signal that distortion was produced by steel beam structure aspect collision alarm.Synthetic material and playing an important role by carbon fiber-reinforced synthetic material in lightweight construction.This kind of synthetic material structural damage can cause brittle fracture process, and this has encouraged the vehicle structure such as the support of aluminium chassis.This framework also has until definite transmission performance of crash sensor.
This excitation for example can comprise velocity information, and this transfer function has the potential information about the variation (distortion, hardening, shorten) of the structure in colliding part and collision.
For this reason advantageously; extract and be included in the information in excitation and in transfer function and it is analyzed from accident signal, so that for example the crash type based on identified, impact strength and/or colliding part trigger the personal protection means in vehicle and optimally protect thus Vehicular occupant.Yet this kind of improved analysis possibility need to be with respect to the transfer function dissociative excitation, and it can not be carried out simply, because no matter be that signal section or the signal section of transfer function encouraged is present in a signal in combination.
Can in cepstrum analysis, see the mode of setting up for signal analysis, as its by B.P.Bogen, M.J.R.Healy and J.W.Tukey at " The Quefrency Alanysis of Time Series for Echoes:Cepstrum; Pseudo-Autocovariance; Cross-Cepstrum; and Saphe Cracking " Proceedings of the Symposiuman Time Series Analysis (Hrsg:M.Rosenblatt), Kapitel 15, S.209-243.New York:Wiley, propose in 1963 like that.For this mode of signal analysis separating for excitation and transfer function for example.Cepstrum analysis is mainly used in voice coding and speech analysis.Cepstrum analysis for example is applied to distinguishing in the field of seismaesthesia signal of earthquake with respect to atom bomb test in addition.For impact detection, advantageously the use of method, the especially cepstrum analysis of dissociative excitation and transfer function is not known in background technology.Therefore a kind of improved mode of the separation and extraction for the collision alarm relevant information has at length been proposed below background technology based on this.
Fig. 1 shows for the block scheme impact detection of vehicle, embodiment foundation method 90 of the present invention.Show the step that the oscillator signal in 100 frequency domains is provided.Carry out subsequently Nonlinear Processing rule application 102 to provided oscillator signal in order to obtain the step of treated oscillator signal.Carry out subsequently treated oscillator signal conversion 104 to time domain so that the step of acquisition time domain oscillator signal.Carry out subsequently the time domain oscillator signal is separated to 106 steps for first signal component and secondary signal component.Finally carry out by detect the step of 108 information by the first and second component of signals.
In step 100, the oscillator signal in frequency domain is provided, by the signal section of collision with by collision, the signal section encouraged in vehicle structure forms this oscillator signal.Two signal sections superpose.Oscillator signal in frequency domain means the output signal received in time domain.In order to obtain the oscillator signal in frequency domain, for example can carry out the discrete analysis of the oscillation such as discrete Fourier transformation.
In step 102, the oscillator signal provided is had for example processing of the processing rule of logarithmetics of nonlinear characteristic in frequency domain.Obtain thus treated oscillator signal.
In step 104, by using processing rule, treated oscillator signal is converted back to time domain by frequency domain.By described step, the especially application of Nonlinear Processing rule and inverse transform subsequently, the time domain obtained is an abstract time domain rather than real time domain.This time domain for example is described to inverted frequency in speech processes.This time domain can be interpreted as a kind of form of frequency, and it can be analyzed and makes the possibility that becomes of inferring especially as basic signal.The result of inverse transform is the time domain oscillator signal.
In step 106, the time domain oscillator signal is separated into one or more first signal components and one or more secondary signal component.These one or more first signal representation in components have part lower value, the time domain oscillator signal in abstract time domain (inverted frequency).These one or more secondary signal representation in components have part higher value, the time domain oscillator signal in abstract time domain (inverted frequency).
In step 108, analyze two component of signals, at least analyze in two component of signals, in order to obtain the information about collision.At this, use one or more first signal components, in order to obtain the information to colliding part or the collision angle of vehicle about the performance of vehicle and state or object, and use one or more secondary signal components, in order to obtain about actual collision, especially about the information of collision velocity.
Can in the other system of vehicle, further use the information obtained now in step 108.For example can control passenger restraint system by obtained information like this.
Fig. 2 shows the schematic diagram of the signal stack caused to the collision of vehicle by object, and the signal stack causes and produced the signal superposeed by two signal sections.Figure 200 is settled side by side and is connected by arrow with 202.
Figure 200 means the signal section directly caused by collision.Figure 20 2 is illustrated in the transfer function of object to vehicle in the zone of the colliding part of vehicle.In Figure 200, at the abscissa subscript, understand time t and understand the amplitude x (t) that depends on time t at the ordinate subscript.The amplitude curve of describing in Figure 200 means collision or caused fracture process in the material of the vehicle structure with light-duty thus.In Figure 20 2, the abscissa subscript understand frequencies omega and the ordinate subscript understand gain/transmission rate for this frequency (
) H (ω).The line of describing in Figure 20 2 means transfer function or the transfer function of vehicle structure, and the filter effect with this by vehicle structure, by the caused signal section of collision become distortion or hangover.
Diagram in Fig. 2 shows the model of " signal generation " in vehicle or in vehicle structure, and this " signal generation " caused by the physical property of vehicle structure.In practice, the difference due to the frequency transmission rate of vehicle structure produces, incoming signal of to be analyzed (t) as the collision alarm s of the method proposed at this.Distortion or the alienation of pumping signal x (t) have been caused thus.Therefore, for the further analysis of signal, advantageously, two of separation signal parts as far as possible exactly, by fracture process and the signal section that produced by the caused hangover of vehicle structure.
It for the new mode of analyzing current signal, is the analysis of high-frequency signal part.The signal of the energy correlation in the simple and popular analysis generation of baseband signal and definite frequency band.This resolution is carried out in the feature space of the combination around the new feature expansion, and wherein, new feature is meaned by the first and second signal sections.
Subsequently, according to Rabiner & " Cepstrum Analysis " (http://cobweb.ecn.purdue.edu/～ee649/notes/cepstrum.html, Kapitel4.5.2-3) of Juang is described in detail aforesaid way.For mode described herein, the signal model for fracture process in vehicle substitutes the example from the speech analysis field of discussing in described works.
Basis as mode described herein, based on following hypothesis, have according to the diagram of Fig. 2 the linear filtering model produced for the signal according to pumping signal x (t) and transfer function H (ω), by signal, produce and generate output or the initial signal s (t) processed for further signal.Can this relation table be shown to following formula by formula:
S (t)=x (t) * h (t), wherein, this pass ties up in frequency domain can be expressed as following formula:
S(ω)＝X(ω)·H(ω)
The purpose of the mode proposed at this is separating of X (ω) and H (ω), so that according to excitation (that is, fracture process) with transmit performance (that is, the reaction of vehicle structure to fracture process) analysis collision information.
Application examples as the Nonlinear Processing rule, obtains following relation as logarithmetics thus:
log|S(ω)＝log[|X(ω)|·|H(ω)|]＝log|X(ω)|+log|H(ω)|
This means log|S (ω) | there are two plus items:
-transfer function H (ω) generates " low frequency wave " (resonance of vehicle structure)
-excitation X (ω) generates " high frequency waves " (fracture process)
According to following relation, can separate this " ripple " by (instead) Fourler transform:
IDFT[log|S(ω)|]＝IDFT[log|H(ω)|]+IDFT[log|X(ω)|]
By logarithmetics, the multiplication of signal section is converted to the addition of individual signals part at this, can realizes thus the separation of simplifying significantly of individual signals component.
Fig. 3 shows the figure of the example results after the step that practices the Nonlinear Processing rule.In Figure 30 0, at the abscissa subscript, understand that take the frequency that Hz is unit understands the frequency intensity LOG|F (ω) of logarithm at the ordinate subscript |
2.Additionally, also indicated the window width 1/T as the second amount on abscissa.Line 302 means the averaged curve about the intensity of the oscillator signal of a frequency limit.Line 304 means the frequency intensity of accurate Calculation, and it has the curve of mound shape by windowing.
Therefore, the curve 302 that has carried out the frequency spectrum of logarithmetics has been shown in Fig. 3.When signal produces, only with addition ground, the signal section of multiplication ground stack is shown with superposeing thus.Be easier to subsequently thus again separate two signal sections or component.
Fig. 4 show according to for the impact detection of vehicle the method step of the time domain oscillator signal during implementing that oscillator signal is back to (inverted frequency) time domain from frequency domain of embodiment of method after Figure 40 0 of example results.For example can realize that this returns by IDFT (discrete fourier inverse transform), in this conversion, calculate power cepstrum.At this, power cepstrum is corresponding to function IDFT[log|S (ω) |].In Figure 40 0, understand the strength rating J of discrete fourier inverse transform at the ordinate subscript in the clear abstract time [t] (inverted frequency) of abscissa subscript
2[LOG|F (ω) |
2].Select name " inverted frequency " according to frequency.As shown in Figure 4, excitation and transfer function are further superposeed by addition ground, yet can be separated well.The part 402 of the figure in Figure 40 0 is illustrated in the transmission characteristic of the vehicle after the excitation by being caused to the caused fracture process in vehicle structure of the collision of vehicle by object.The part 404 of another of the figure in Figure 40 0 means by the caused original oscillating part of collision.Abstract time t (inverted frequency) on abscissa though unit for second be not the real time.Unit is millisecond, yet can not be understood as time shaft.The abstract time more is appreciated that a quefrency.Also can derive thus, part 402 is positioned in corresponding in the zone of lower frequency and curved portion 404 is positioned in the zone corresponding to upper frequency.Part 402 for example comprises about the information at the position of collision with about the information of the deformation and failure of the vehicle structure at colliding part subsequently.Part 404 comprises for example about providing about the type of collision and the information of seriousness.For the further application of obtained information, should these two parts of separate consideration.Be necessary it disconnected from each other for this reason.
If will encourage with transmission and separate now, will carry out so so-called " power cepstrum liftering " (Cepstral liftering).The higher coefficient obtained by this liftering function (higher inverted frequency value) is corresponding to excitation, and lower coefficient (lower inverted frequency value) is corresponding to the value of transfer function.
Can power cepstrum be introduced to segregator with two plus items now, the outstanding inverted frequency T that this segregator analysis for example encourages with and amplitude.Low inverted frequency characterizes to be transmitted.Can be using low inverted frequency directly as the cepstral coefficients analysis, or convert it to spectrum domain, for example, as level and smooth transfer function.
Fig. 5 shows for the block scheme impact detection of vehicle, another embodiment foundation method of the present invention.In step 502, pending signal s (t) is carried out to windowing.In step 504, thus obtained signal is carried out to discrete Fourier transformation subsequently, and then, in step 506, consequent signal is carried out to logarithmetics.In step 508, the signal obtained in this mode is carried out to the discrete fourier inverse transform subsequently.Then in the step 510 of power cepstrum liftering, obtained signal is separated into to the first scope 512 and the second scope 514 and in step 516, single scope is classified in order to detect to the feature of the collision of vehicle subsequently.Therefore from the step 516 of classification, produce for distinguishing the resolution feature 518 of different crash typees.
In windowing 502, by continuous but obviously limited window the signal s (t) provided by applicable device is provided in time, in order to can carry out for example discrete Fourier transformation 504 to the value in window.By Fourler transform 504, signal s (t) is converted to frequency domain and available as frequency spectrum.Step by the logarithmetics 506 as the Nonlinear Processing rule changes this frequency spectrum, in order to can carry out improved separation or be extracted in the signal section in the signal in frequency domain.In step 508, the signal through changing is back to again time domain in form by the discrete fourier inverse transform, and this time domain does not mean the real time certainly.Not equal to this time domain more should be interpreted as the standard time (inverted frequency) and it means frequency.Discrete fourier inverse transform 508 provides so-called power cepstrum, and it means the intensity about the signal of inverted frequency.The unit of account of inverted frequency is second.It is very simple that the signal that will superpose in output signal s (t) in power cepstrum is separated from each other.Realized this kind of separation in liftering step 510, will there is the signal section of lower inverted frequency and the signal section of high inverted frequency and be separated in this step.In embodiment described herein, there is self-oscillation performance or the transfer function of the signal section of lower inverted frequency corresponding to vehicle, and the signal section of high inverted frequency is corresponding to the direct impact of collision or excitation.Therefore, in step 516, analyze discretely this two component of signals, and use this result together (or separately), for differentiating 518 or distinguish collision process and collision consequence.
Fig. 6 show according to another embodiment of the present invention, for the diagram of circuit of the method for the simplification of the impact detection of vehicle.The bank of filters that there is band logical 604 by use, for example in structure-borne sound sensor 602, in the device processing signals s (t) for receiving signal s (t) and the analysis 606 to energy content by treated signal leading.In addition subsequently by the obtained signal leading about energy content to LPF 608, in order to obtain the oscillator signal of frequency domain.Can in the peculiar hardware of the sensor of structure-borne sound sensor or acceleration pick-up, implement said units.Then, in the alternative step of signal value, by the substitution value from non-linear look-up table, change the signal in frequency domain, wherein in signal value to be read in and the nonlinear relationship between substitution value are stored in look-up table (Look-Up-Table).In bank of filters again after filtering 612, in separating 614 step, select single passage for the low frequency signal part partly is separated with high-frequency signal and extracts low frequency signal part 616 at those passages from high-frequency signal part 618.In classification 620, check two signal sections (or only having wherein single signal section) and use and differentiate feature 622, the other safety device such as safety air bag or belt pretensioner for control example as vehicle according to differentiating feature 622.
In other words, according to the embodiment according to Fig. 6, be in this case device reception and/or ready signal s (t) structure-borne sound sensor 602 or acceleration pick-up, for receiving.By filtering, integration and again filtering signal s (t) is transformed into to frequency domain from time domain.By this way, can analyze simply now the intensity of the frequency of individually calling.In step 610, replace the signal value of the signal in the frequency domain obtain by the value from non-linear look-up table, wherein in this look-up table, can mean substituting corresponding to the signal value of logarithmetics.The signal of logarithmetics or the filtering again 612 of alternative signal makes and separates simply subsequently 614 signal sections with high and low inverted frequency value and become possibility by the numerical value from non-linear look-up table, this again filtering corresponding to signal in frequency domain the inverse transform to time domain.Realize this separation 614 at this with two subdivisions: the value 616 with low inverted frequency means the information that the transfer function by vehicle causes, and has the information that the value 618 of high inverted frequency means corresponding to the direct impact of collision.
In order to make it more clear: the second filtering 612 is similar to the Fourier inversion in cepstrum analysis.The input value of filter is at the output signal of current system time, frequency bands that process by step 606-610, in 604.
In classification 620 steps subsequently, analyze two inverted frequency subdivisions.Obtain the information about the distortion in collision velocity and structure and destruction analyzing or classify in 620.This information is exactly for distinguishing the resolution feature 622 of different crash typees.
The complexity of the numerical value that the embodiment of last described foundation method of the present invention is low based on it is particularly suitable for, in vehicle, still can be implemented or be implemented in vehicle with higher processing capacity especially soon with acceptable speed with lower processing capacity.According to the embodiment shown in Fig. 6, as one man by the function that can implement simply, substituted the function of numerical value or circuit engineering complexity.The modification of the simplification of signal analysis is the form for the simplification of the aforementioned manner of signal analysis, does not need to carry out DFT, therefore can easily realize.Simultaneously, substitute so-called Fourier analysis or conversion in the aforementioned embodiment by applicable bank of filters.Be similar to current SBS sensor, calculate the energy in single passage.In Fig. 6, grey illustrates sensor, is for example to have relevant functional structure-borne sound sensor.The formation of logarithm is by being used non-linear look-up table to be substituted by the exchange of values function.Process by means of the second bank of filters and substitute the IFFT function by signal.In " liftering " step, realize the selection of single passage to be analyzed, i.e. the selection of high and low inverted frequency and affiliated signal section.Classification is carried out in the similar analysis that can export by passage.
Can for example realize thus the classification of crash type, by the value of first signal part or curve and/or secondary signal, partly value or curve are compared with predetermined value or predetermined signal curve.Can for example by the record of the numerical value in theoretical calculating or the test accident under experiment condition, depict these predetermined signal curves, and the reference using it as the numerical curve under definite crash type.Then, can so determine the crash type in fact occurred, this crash type will be identified as occurred crash type, when the signal under occurred crash type or value curve with measured in the step 620 or 516 of classifying or handled signal curve is the most similar.From the high-frequency signal part, can for example by the height that the definite of ratio or the maxim (" spike ") in high inverted frequency scope occur, realize speed or the impact site of object to vehicle.
The present invention has also realized a kind of device for the impact detection at vehicle 700, as its in Fig. 7 as shown in block scheme like that.This device 700 has the device 720 (for example interface) of the oscillator signal for frequency domain is provided, wherein, described oscillator signal means directly by the caused signal section of the collision to vehicle and stack by the signal section encouraged in this collision vehicle structure.In addition this device 700 also comprise that be connected with the device 720 for providing, for by the Nonlinear Processing rule application to the oscillator signal that provides in order to obtain the device 740 of treated oscillator signal.This device 700 also comprises for inverted frequency, treated oscillator signal being converted to the time domain with inverted frequency in order to obtain the device 760 of time domain oscillator signal.In addition this device 700 also comprise that couple with the device 760 for conversion, for separating of the time domain oscillator signal in order to obtain the device 780 there is the first signal component of low inverted frequency value and to there is the secondary signal component of high inverted frequency value.Finally this device 700 also comprises for passing through use one or more first and/or the secondary signal component detection device 790 about the information of the collision of the collision to vehicle and/or vehicle.
Claims (12)
1. for the method (90,500,600) of the impact detection of vehicle, there are following steps:
Oscillator signal in (100,502-504,604-608) frequency domain is provided, and wherein, described oscillator signal means the stack to the signal section that collision directly caused with the signal section encouraged in vehicle structure by described collision of described vehicle by object;
By Nonlinear Processing rule application (102,506,610) to described provided oscillator signal, in order to obtain treated oscillator signal;
With inverted frequency, described treated oscillator signal is changed to (104,508,612) to time domain, in order to obtain the time domain oscillator signal;
Separate (106,510,614) described time domain oscillator signal, in order to obtain, there is the first signal component of low inverted frequency value and the secondary signal component with high inverted frequency value; And
By using described first and/or secondary signal component detection (108,516,620) information to the collision of the collision of described vehicle and/or described vehicle about described object, wherein about described object, to the information of the collision of described vehicle, the transfer function by described vehicle causes, and the direct impact by described collision causes about the information of the collision of described vehicle.
2. method according to claim 1, it is characterized in that, described incoming signal (s (t)) windowing (502) in being implemented as time domain during step is provided and subsequently by the described conversion of the incoming signal through windowing (504) to frequency domain, in order to obtain the oscillator signal in described frequency domain.
3. method according to claim 1, it is characterized in that, incoming signal in time domain (s (t)) is separated to (604) for portions of the spectrum and determines energy value of (606-608) described portions of the spectrum during step is provided described, in order to obtain the oscillator signal in described frequency domain.
4. according to the described method of any one in aforementioned claim, it is characterized in that, carry out the logarithmetics of described provided oscillator signal in described applying step, in order to obtain described treated oscillator signal.
5. according to the described method of any one in claims 1 to 3, it is characterized in that, the comparison of the value of described provided oscillator signal and value from predetermined look-up table is provided in described application (610) step, and according to distributing to value described look-up table and that store, carry out the described value that described provided oscillator signal is provided by predetermined value, in order to obtain described treated oscillator signal.
6. method according to claim 1, is characterized in that, executive signal conversion in described conversion (508) step, in order to obtain described time domain oscillator signal.
7. method according to claim 6, is characterized in that, the signal conversion of carrying out in described conversion (508) step is the discrete fourier inverse transform.
8. method according to claim 1, is characterized in that, in described conversion (612) step, by the use bank of filters, described treated oscillator signal is converted to described time domain oscillator signal.
9. method according to claim 1, it is characterized in that, in described separation (510) step, window function is applied to described time domain oscillator signal, there is the described first signal component of low inverted frequency value and the described secondary signal component with high inverted frequency value in order to obtain.
10. method according to claim 1, it is characterized in that, carry out the signal filtering of described time domain oscillator signal in described separation (612) step, in order to obtain, there is the described first signal component of low inverted frequency value and the described secondary signal component with high inverted frequency value.
11. method according to claim 1, it is characterized in that, in described detection (518, the described first signal component that 620) will have low inverted frequency value in step is compared with predetermined inverted frequency curve, and wherein said predetermined inverted frequency curve means the information about colliding part and/or crash type; The described secondary signal component that wherein has high inverted frequency value means the information about impact strength by the peaked quantity in the inverted frequency interval and/or peaked intensity.
12. the device (700) for the impact detection of vehicle has following characteristics:
For the device (720) of oscillator signal of frequency domain is provided, wherein said oscillator signal means the stack to the signal section that collision directly caused of described vehicle and the signal section that encouraged in vehicle structure by described collision by object;
Be used for the Nonlinear Processing rule application to described provided oscillator signal, in order to obtain the device (740) of treated oscillator signal;
For with inverted frequency, described treated oscillator signal being converted to time domain, in order to obtain the device (760) of time domain oscillator signal;
For separating of described time domain oscillator signal, in order to obtain the first signal component with low inverted frequency value and the device (780) with secondary signal component of high inverted frequency value; And
For by use described first and/or the secondary signal component detection about described object the device (790) to the information of the described collision of the described collision of described vehicle and/or described vehicle, wherein about described object, to the information of the described collision of described vehicle, the transfer function by described vehicle causes, and the direct impact by described collision causes about the information of the described collision of described vehicle.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| DE201010028845 DE102010028845A1 (en) | 2010-05-11 | 2010-05-11 | Method of detecting vehicle crash, involves directing processed oscillation signal to time domain by performing quefrency analysis to form time domain oscillation signal from which signal components with high/low quefrencies are separated |
| DE102010028845.4 | 2010-05-11 |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN102259628A CN102259628A (en) | 2011-11-30 |
| CN102259628B true CN102259628B (en) | 2014-01-08 |
Family
ID=44859734
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201110124872.5A Expired - Fee Related CN102259628B (en) | 2010-05-11 | 2011-05-11 | Method And Apparatus For Collision Detection In Vehicles |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN102259628B (en) |
| DE (1) | DE102010028845A1 (en) |
Families Citing this family (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102014001258A1 (en) * | 2014-01-30 | 2015-07-30 | Hella Kgaa Hueck & Co. | Device and method for detecting at least one structure-borne sound signal |
| CN104807538A (en) * | 2015-02-12 | 2015-07-29 | 无锡市崇安区科技创业服务中心 | Vehicle collision noise characteristic detection system |
| CN108216101B (en) * | 2016-12-13 | 2019-08-27 | 大陆汽车电子（连云港）有限公司 | Vehicle collision signal processing method and processing device |
| CN108582070A (en) * | 2018-04-17 | 2018-09-28 | 上海达野智能科技有限公司 | robot collision detecting system and method, storage medium, operating system |
| DE102019218069A1 (en) * | 2019-11-22 | 2021-05-27 | Zf Friedrichshafen Ag | Device and method for recognizing and classifying an opponent in an accident |
| DE102020205580A1 (en) * | 2020-05-04 | 2021-11-04 | Robert Bosch Gesellschaft mit beschränkter Haftung | Method for determining a type of collision of a vehicle |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102005026188A1 (en) * | 2005-06-06 | 2006-12-07 | Daimlerchrysler Ag | Control method for an accident prevention device in a vehicle |
| CN200964090Y (en) * | 2006-11-15 | 2007-10-24 | 刘加友 | Automobile collision presentiment device |
| EP2045144A1 (en) * | 2007-10-02 | 2009-04-08 | Continental Automotive GmbH | Method for identifying the collision side in a frontal impact |
| CN101596891A (en) * | 2008-06-02 | 2009-12-09 | 现代自动车株式会社 | The collision detecting system that is used for vehicle |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102004038984A1 (en) | 2004-08-10 | 2006-02-23 | Robert Bosch Gmbh | Device for crash detection |
-
2010
- 2010-05-11 DE DE201010028845 patent/DE102010028845A1/en not_active Withdrawn
-
2011
- 2011-05-11 CN CN201110124872.5A patent/CN102259628B/en not_active Expired - Fee Related
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| DE102005026188A1 (en) * | 2005-06-06 | 2006-12-07 | Daimlerchrysler Ag | Control method for an accident prevention device in a vehicle |
| CN200964090Y (en) * | 2006-11-15 | 2007-10-24 | 刘加友 | Automobile collision presentiment device |
| EP2045144A1 (en) * | 2007-10-02 | 2009-04-08 | Continental Automotive GmbH | Method for identifying the collision side in a frontal impact |
| CN101596891A (en) * | 2008-06-02 | 2009-12-09 | 现代自动车株式会社 | The collision detecting system that is used for vehicle |
Also Published As
| Publication number | Publication date |
|---|---|
| DE102010028845A1 (en) | 2011-11-17 |
| CN102259628A (en) | 2011-11-30 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN102259628B (en) | Method And Apparatus For Collision Detection In Vehicles | |
| CN110316131B (en) | Vehicle collision protection method and device | |
| KR20090060434A (en) | Method and apparatus for triggering personal protective means of a vehicle | |
| CN111724599B (en) | Method, device, equipment and medium for obtaining safety driving behavior evaluation data | |
| CN101516685B (en) | Device and method for actuating passenger protection means | |
| CN1980819A (en) | Device for triggering a vehicle safety device | |
| EP1982869A1 (en) | Method of characterizing a vehicle crash event based on shear waves in the vehicle frame | |
| US7158017B2 (en) | Method for controlling a safety system in a vehicle | |
| CN115116232B (en) | Voiceprint comparison method, device and equipment for automobile whistling and storage medium | |
| CN103241207A (en) | Method and device for determining a collision speed on the collision of a vehicle | |
| CA2473556A1 (en) | Omni-directional crash sensor | |
| CN105320812A (en) | Method for detecting tie-in strip and expansion adhesive materials in CAE pedestrian protection head collision | |
| JP3778833B2 (en) | Collision type discrimination device | |
| CN104870943A (en) | Redundant signal capture | |
| CN117452496A (en) | Target detection and identification method for seismoacoustic sensor | |
| KR101219826B1 (en) | A pulse control apparatus evaluating air bag control unit for car and method thereof, a evaluating air bag control unit and pulse processing method thereof, a system evaluating air bag control unit for car and method thereof | |
| KR20120103286A (en) | Method and apparatus of detecting event of interest related to vehicles using audio signals | |
| CN108382400A (en) | Method for determining the time of casualty parameter for evaluating motor vehicle accident situation | |
| CN100374325C (en) | Arrangement for controlling retaining means | |
| JP2024040220A (en) | vehicle | |
| US6681199B2 (en) | Peak time detecting apparatus and method | |
| CN1989027B (en) | A device for activating vehicle safety equipment | |
| JP2001294118A (en) | Vehicle collision determination device | |
| WO2018167897A1 (en) | Biometric information detection device and biometric information detection program | |
| CN109191909A (en) | A virtual resource adjustment method caused by temporary parking on expressway |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20140108 Termination date: 20170511 |