CN103235891A - Road identification system and method based on vehicle vertical vibration system identification - Google Patents

Abstract

The invention discloses a road identification system and method based on vehicle vertical vibration system identification. The road identification system based on the vehicle vertical vibration system identification is mainly composed of a sensor signal acquisition module, a sensor signal processing module, a system identification module, a self-adaptive state observation module, a road space process generating module, a road characteristic parameter extraction module and a road classifier. The sensor signal processing module, the system identification module, the self-adaptive state observation module, the road space process generating module, the road characteristic parameter extraction module and the road classifier are integrated in an ECU (electronic control unit) chip and communicated with one another through a bus, and the sensor signal acquisition module is subjected to data transmission with the ECU chip through a wire harness. The road identification system and method based on the vehicle vertical vibration system identification aims at providing a method and a measuring system which are capable of estimating road types conveniently, rapidly and relatively accurately to solve the defect that an existing road identification technology cannot adapt to the changes of vehicle system parameters.

Description

Road surface recognition system and method based on the System Discrimination of vehicle vertical vibration

Technical field

The invention belongs to the recognition technology field, road surface in the Vehicle Engineering, be specifically related to a kind of vertical vibration signal when utilizing vehicle to travel and carry out the identification of Vehicular vibration systematic parameter, the system and method for the walking along the street face of going forward side by side identification.

Background technology

Along with the continuous development of automotive engineering, people are also more and more higher to the requirement of vehicle performance.Under such overall situation, semi-active suspension technology, Active suspension technology, ABS technology are arisen at the historic moment.These new technologies have all improved comfortableness and the security of vehicle to a certain extent.If can go out the type on road surface according to the driving states signal recognition of vehicle, just can optimize the steering logic in the above-mentioned new technology largely, thereby further improve the performance of vehicle.

Chinese scholars has been done a large amount of research at the road surface recognition methods that is used for the real vehicle control system, has obtained certain achievement.Chinese patent CN102289674A and US Patent No. 4651290 all disclose a kind of method of identifying based on the road surface of acceleration signal, respectively acceleration signal frequency analysis and statistical study have been carried out, but this method is used for different cars, all must carry out a large amount of tests demarcates, even same car, load change or suspension system change the result that (Active suspension can be regulated damping or stiffness parameters automatically under different driving cycles) all can influence identification.Moustapha Doumiati is in the control meeting of the ACC(U.S.) a kind of method of utilizing Kalman filtering to estimate Uneven road of proposition, can estimate the elevation on road surface comparatively simply, but this method do not possess the adaptability when the Vehicular system parameter changed.

Summary of the invention

The object of the present invention is to provide a kind of simple and efficiently and can estimate method and the measuring system thereof of road surface types comparatively exactly, effectively improve existing automobile-used road surface recognition technology and can not adapt to the weak point that the Vehicular system parameter changes.

The present invention is achieved through the following technical solutions for solving the problems of the technologies described above:

A kind of road surface recognition system based on the System Discrimination of vehicle vertical vibration, mainly by collecting sensor signal module I, sensor signal processing module II, System Discrimination module III, self-adaptation state observation module IV, space, road surface course generation module V, road surface characteristic parameter extraction module VI and road surface sorter VII are formed, described sensor signal processing module II, System Discrimination module III, self-adaptation state observation module IV, space, road surface course generation module V, road surface characteristic parameter extraction module VI and road surface sorter VII are integrated in the ECU chip X, and carry out each other communication by bus, described collecting sensor signal module I is carried out data transmission by wire harness and ECU chip X.

Described collecting sensor signal module I comprises body-acceleration sensor 1, suspension moves stroke sensor 2, vehicle speed sensor 3 and A/D converter 4, described body-acceleration sensor 1 is used for the vertical acceleration signal of collection vehicle vertical direction, and be installed on the vehicle front overhang vibration damper on the installed surface, the moving stroke sensor 2 of suspension adopts angular displacement sensor, the one end is connected on the subframe, the other end is connected in wheel, measure the moving stroke of suspension with the variation of its angular displacement, vehicle speed sensor 3 generally all can have ready-made on middle-and-high-ranking car, can directly read speed information from the CAN bus, also can obtain from meter panel of motor vehicle, described A/D converter 4 has the high conversion rate characteristic.

Described sensor signal processing module II, major function is the sensor signal that collects to be carried out filtering remove noise pollution, do the processing in early stage of data simultaneously for System Discrimination module III, because System Discrimination module III requires vehicle body vertical velocity and absolute displacement signal to import as it, this just requires acceleration signal is done Integral Processing, because the acceleration signal that measures exists unpredictable zero point drift and noise pollution, and rate signal and displacement signal that direct integral is obtained do not have value, so added the function of integral filtering in sensor signal processing module II.

Described System Discrimination module III, major function is the systematic parameter of real-time identification vehicle vertical vibration, for the realization of self-adaptation state observation module IV provides the basis.

Described self-adaptation state observation module IV is to be based upon on the basis of System Discrimination module III and to realize, can adapt to the parameter creep of different loads state, different rows turner condition Vehicular system.

Space, described road surface course generation module V is that the output signal of self-adaptation state observation module IV and vehicle speed sensor 3 is done an integration, to realize converting the road surface time history to space, road surface course.

Characteristic parameter extraction is done in the signal output of described road surface characteristic parameter extraction module VI road pavement space course generation module V, to realize representing its road surface characteristic with one or several characteristic index.

The function of described road surface sorter VII is the characteristic parameter output definition threshold value for road surface characteristic parameter extraction module VI, thereby distinguishes different road surface types very easily.

A kind of road surface recognition methods based on the System Discrimination of vehicle vertical vibration may further comprise the steps:

(1) gathers vehicle body vertical acceleration signal, the moving stroke signal of suspension and vehicle speed signal by collecting sensor signal module I;

(2) the above-mentioned signal that collects is passed through sensor signal processing module II, with the filtering of realization signal and the integrating function of acceleration signal;

(3) the vehicle body acceleration signal that sensor signal processing module II is obtained, vehicle body vertical velocity signal, the moving stroke signal of vehicle body vertical deviation signal and suspension is passed to System Discrimination module III and is carried out System Discrimination, with real-time variation of monitoring out vehicle vertical vibration systematic parameter;

(4) systematic parameter that identification obtains according to step (three) is adjusted the systematic parameter in the self-adaptation state observation module IV in real time;

(5) the moving stroke signal of vehicle body acceleration signal, vehicle body vertical deviation signal and suspension is passed to self-adaptation state observation module IV, to realize the identification of road pavement input;

(6) the road surface time history signal and the vehicle speed signal that pick out in the step (five) are passed to space, road surface course generation module V, to realize that the road surface time history is to the conversion of space course;

(7) space, the road surface course signal that step (six) is obtained is passed to road surface characteristic parameter extraction module VI, to extract the parameter that can characterize road surface characteristic;

(8) in the sorter VII of road surface, set threshold value (setting of threshold value can be demarcated by the test on the different road surfaces) for the road surface characteristic parameter, the parameter that step (seven) obtains is passed through road surface sorter VII, from the classification of final realization road pavement.

Described collecting sensor signal module I, sensor signal processing module II, System Discrimination module III and self-adaptation state observation module IV are nucleus modules of the present invention, its job step also is the core procedure of the road surface recognition methods based on the System Discrimination of vehicle vertical vibration of the present invention, provides the detailed introduction of the job step of these several modules below respectively.

(A) collection vehicle vehicle body vertical vibration acceleration, the moving stroke of suspension and vehicle speed signal

The installation site of the moving stroke sensor 2 of acceleration transducer 1 and suspension as shown in Figure 3, wherein acceleration transducer 1 is contained on the vibration damper on the mount pad, the moving stroke sensor 2 of suspension is angular displacement sensor, the one end is connected on the subframe, the other end is connected in wheel, measures the moving stroke of suspension with the variation of its angular displacement.Vehicle speed signal generally can directly read from the CAN bus on middle-and-high-ranking car, also can obtain from meter panel of motor vehicle.

(B) processing of sensor signal

Here it mainly is acceleration transducer signals that the signal that relates to is handled, because System Discrimination module III requires vehicle body vertical velocity and absolute displacement signal to import as it, this just requires acceleration signal is done Integral Processing, because the acceleration signal that measures exists unpredictable zero point drift and noise pollution, and rate signal and displacement signal that direct integral is obtained do not have value.The present invention introduces the wave filter of band integrating function in signal processing module, the transport function of wave filter is as follows:

${\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="HDA00003144073400022.png,HDA00003144073400032.png"\; state="\{\{state\}\}">G</figure-callout>}}_1=\frac{1}{{({\mathrm{<figure-callout\; id="2"\; label="S"\; filenames="HDA00003144073400011.png,HDA00003144073400022.png"\; state="\{\{state\}\}">S</figure-callout>}}^2+{2\mathrm{\&xi;}}_0{\mathrm{\&omega;}}_{0}S+{\mathrm{\&omega;}}_0^2)}^2},$ ${\mathrm{<figure-callout\; id="2"\; label="G"\; filenames="HDA00003144073400011.png,HDA00003144073400022.png"\; state="\{\{state\}\}">G</figure-callout>}}_2=\frac{S}{{({\mathrm{<figure-callout\; id="2"\; label="S"\; filenames="HDA00003144073400011.png,HDA00003144073400022.png"\; state="\{\{state\}\}">S</figure-callout>}}^2+{2\mathrm{\&xi;}}_0{\mathrm{\&omega;}}_{0}S+{\mathrm{\&omega;}}_0^2)}^2}$

Wherein, G ₁, G ₂Be respectively the integration filter of asking absolute displacement and speed, ω ₀Be the cutoff frequency of wave filter, and ω ₀Satisfy ω ₀＜＜ω ₁, ω ₁Be the natural frequency of vehicle body, ξ ₀Be damping ratio.

(C) vehicle vertical vibration System Discrimination

Owing to the parameter of active/semi-active suspension system can change along with the variation of load, load distribution and driving cycle, the present invention has introduced vehicle vertical vibration systematic parameter real-time identification module.

Here with the example of least square method of recursion identification 1/4 car vibrational system parameter (spring carried mass and ratio of damping) as an illustration, single-wheel vehicle vertical vibration system simplification is become as shown in Figure 4.

To the sprung mass row equation of motion:

Making the moving stroke of suspension is x, and the stroke of nonspring carried mass and spring carried mass is respectively x ₁, x ₂, x=x ₂-x ₁, have:

$M{\stackrel{\&CenterDot;\&CenterDot;}{x}}_2+C_s\stackrel{\&CenterDot;}{x}{K}_{s}x=0$

Wherein, M is spring carried mass, and Cs is the suspension shock-absorber damping, and Ks is bearing spring rigidity;

Thereby

$-K_{s}x=\begin{array}{c}\left[M{C}_s\right]\left[\begin{array}{c}{\stackrel{\&CenterDot;\&CenterDot;}{x}}_2\\ \stackrel{\&CenterDot;}{x}\end{array}\right]\end{array}$

For the sake of simplicity, following formula can be write as

y＝θ ^Tψ

Wherein

Then can use Recursive Least Squares, estimate the θ value in the following formula.Concrete algorithm for estimating is as follows

$\stackrel{\&CenterDot;}{\widehat{\mathrm{\&theta;}}}=P(y-{\widehat{\mathrm{\&theta;}}}^T\mathrm{\&psi;})\mathrm{\&psi;}$

$\stackrel{\&CenterDot;}{P}=-P\frac{\mathrm{\&psi;}{\mathrm{\&psi;}}^{\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HDA00003144073400022.png,HDA00003144073400032.png"\; state="\{\{state\}\}">T</figure-callout>}}}{1+{\mathrm{\&psi;}}^T\mathrm{P\&psi;}}P$

P is one 2 * 2 symmetric matrix in the formula, the covariance that representative is estimated,

Be the minimum variance estimate of θ, P and

All want initialize.

(D) foundation of self-adaptation state observer

In conjunction with Fig. 4, this vibrational system also remains m, Kt after the System Discrimination through step (three), the several parameters of Ks are undetermined, but in fact these several quantitative changeizations are very little in motion for vehicle, can ignore its variation, them as constant, whole like this vibrational system parameter has just been determined, set up the state observer of the face input state of leading the way below, with the example that is established as of optimum state observer (Kalman filter), describe here.

In conjunction with Fig. 4, the selection mode variable It is as follows to get the discrete system dynamic equation:

x _k+1＝G _kx _k+b _m,k

z _k＝H _kx _k+b _s,k

In the formula:

Be the state vector of system, and original state is 0;

Be the output vector of system, wherein: x _{2, k}-x _{1, k}: suspension moves stroke, can be recorded by height sensor;

Vehicle body acceleration can be recorded by acceleration transducer; X2, k: the vehicle body absolute displacement can obtain 2 integrations of acceleration signal; x _{0, k}: the road surface elevation;

b _{M, k}And b _{S, k}Be respectively process error and measuring error vector, being assumed to be average here is 0, and incoherent white noise.

State matrix G and output matrix H are as follows:

$G=E+\mathrm{\&Delta;t}\left(\begin{array}{cccccc}0& 1& 0& 0& 0& 0\\ -\frac{\mathrm{Kt}}{M}& -\frac{\mathrm{Cs}}{M}& \frac{\mathrm{Kt}}{M}& -\frac{\mathrm{Cs}}{M}& 0& 0\\ 0& 0& 0& 1& 0& 0\\ \frac{\mathrm{Ks}}{m}& \frac{\mathrm{Cs}}{m}& -\frac{\mathrm{Ks}+\mathrm{Ku}}{m}& -\frac{\mathrm{Cs}}{m}& \frac{\mathrm{Ku}}{m}& 0\\ 0& 0& 0& 0& 0& 1\\ 0& 0& 0& 0& 0& 0\end{array}\right)$

$H=\left(\begin{array}{cccccc}1& 0& -1& 0& 0& 0\\ -\frac{\mathrm{Kt}}{M}& -\frac{\mathrm{Cs}}{M}& \frac{\mathrm{Kt}}{M}& \frac{\mathrm{Cs}}{M}& 0& 0\\ 1& 0& 0& 0& 0& 0\end{array}\right)$

Wherein, E is unit matrix, and Δ t is the sampling time.

According to above-mentioned equation design Kalman filter, the acceleration signal that sensor is recorded and moving stroke signal are as input.Can estimate the curve of road surface time history.

In addition, the detailed process of space, the described road surface of step (five) course generation is:

For the road surface time history that will obtain in the step (four) converts space, road surface course to, to use the signal of vehicle speed sensor here.For reducing the operand of ECU, the present invention introduces trigger pip Flag, and has only T in the regulation unit interval period T ₀In time period, space, road surface course generation module V triggers, as shown in Figure 5.When trigger pip Flag is 1, reads vehicle speed signal v, and begin to calculate S=S+v Δ t, simultaneously with the road surface time history data storage of calculating; If S 〉=S ₀The time, utilize method of interpolation that the road surface time history of storage is changed into the space course, and then S is put 0; If S＜S ₀, then store the road surface time history in the S distance section, and return and continue to read vehicle speed signal v.When trigger pip Flag is 0, do not carry out reading and changing of data.

The detailed process of the described road surface characteristic Parameter Extraction of step (six) is:

The method that eigenwert is extracted has a lot, can estimate the energy of the signal that the road surface comprises with the power Zymography in different frequency range, also the primary band in-scope of road surface signal input can be estimated with the single order zero crossing method of passing through, even statistical study such as root-mean-square value, maximal value can be directly got to signal.It is worthy of note that road surface characteristic parameter extraction module VI should adopt same time trigger with space, road surface course generation module V.

The detailed process of the foundation of the described road surface of step (seven) sorter is:

At the different characteristic parameter extracting method that step (six) adopts, can set up different road surface sorters.For example pass through method at the single order zero crossing, the road surface sorter only need provide 4 frequency range: 0-4Hz, 4-8Hz, 8-12Hz, more than the 12Hz, and the feature dominant frequency parameter that obtains in the determining step (six) can directly draw the road surface classification results in which frequency range.

Beneficial effect of the present invention is: can be simple and efficient and can estimate road surface types comparatively exactly, can be used on the different vehicles, and do not need to carry out a large amount of tests and demarcate, and possess the adaptability that load change and Vehicular system parameter change.

Description of drawings

Fig. 1 is the theory diagram of the road surface recognition system based on the System Discrimination of vehicle vertical vibration of the present invention;

Fig. 2 is the main flow chart of the road surface recognition methods based on the System Discrimination of vehicle vertical vibration of the present invention;

Fig. 3 is the layout synoptic diagram of each sensor among the present invention;

Fig. 4 is the 1/4 Vehicular vibration system that the present invention is used for System Discrimination and road surface estimation;

Fig. 5 is the process flow diagram that the present invention is used for the road surface time history is converted to the space course.

Among the figure:

I, collecting sensor signal module; II, sensor signal processing module; III, System Discrimination module;

IV, self-adaptation state observation module; V, space, road surface course generation module; VI, road surface characteristic parameter extraction module;

VII, road surface sorter; X, ECU chip;

1, body-acceleration sensor; 2, the moving stroke sensor of suspension; 3, vehicle speed sensor; 4, A/D converter.

Embodiment

The present invention will be further described below in conjunction with accompanying drawing.

The present invention is achieved through the following technical solutions for solving the problems of the technologies described above:

Fig. 1 is the theory diagram of the road surface recognition system based on the System Discrimination of vehicle vertical vibration of the present invention, as seen it is mainly by collecting sensor signal module I, sensor signal processing module II, System Discrimination module III, self-adaptation state observation module IV, space, road surface course generation module V, road surface characteristic parameter extraction module VI and road surface sorter VII are formed, described sensor signal processing module II, System Discrimination module III, self-adaptation state observation module IV, space, road surface course generation module V, road surface characteristic parameter extraction module VI and road surface sorter VII are integrated in the ECU chip X, and carry out each other communication by bus, described collecting sensor signal module I is carried out data transmission by wire harness and ECU chip X.

Described collecting sensor signal module I comprises body-acceleration sensor 1, suspension moves stroke sensor 2, vehicle speed sensor 3 and A/D converter 4, Fig. 3 is the layout synoptic diagram of each sensor among the present invention, described body-acceleration sensor 1 is used for the vertical acceleration signal of collection vehicle vertical direction, and be installed on the vehicle front overhang vibration damper on the installed surface, the moving stroke sensor 2 of suspension adopts angular displacement sensor, the one end is connected on the subframe, the other end is connected in wheel, measure the moving stroke of suspension with the variation of its angular displacement, vehicle speed sensor 3 generally all can have ready-made on middle-and-high-ranking car, can directly read speed information from the CAN bus, also can obtain from meter panel of motor vehicle, described A/D converter 4 has the high conversion rate characteristic.

Described sensor signal processing module II, major function is the sensor signal that collects to be carried out filtering remove noise pollution, do the processing in early stage of data simultaneously for System Discrimination module III, because System Discrimination module III requires vehicle body vertical velocity and absolute displacement signal to import as it, this just requires acceleration signal is done Integral Processing, because the acceleration signal that measures exists unpredictable zero point drift and noise pollution, and rate signal and displacement signal that direct integral is obtained do not have value, so added the function of integral filtering in sensor signal processing module II.

Described System Discrimination module III, major function is the systematic parameter of real-time identification vehicle vertical vibration, for the realization of self-adaptation state observation module IV provides the basis.

Described self-adaptation state observation module IV is to be based upon on the basis of System Discrimination module III and to realize, can adapt to the parameter creep of different loads state, different rows turner condition Vehicular system.

Space, described road surface course generation module V is that the output signal of self-adaptation state observation module IV and vehicle speed sensor 3 is done an integration, to realize converting the road surface time history to space, road surface course.

Characteristic parameter extraction is done in the signal output of described road surface characteristic parameter extraction module VI road pavement space course generation module V, to realize representing its road surface characteristic with one or several characteristic index.

The function of described road surface sorter VII is the characteristic parameter output definition threshold value for road surface characteristic parameter extraction module VI, thereby distinguishes different road surface types very easily.

Fig. 2 is the main flow chart of the road surface recognition methods based on the System Discrimination of vehicle vertical vibration of the present invention, may further comprise the steps:

(9) gather vehicle body vertical acceleration signal, the moving stroke signal of suspension and vehicle speed signal by collecting sensor signal module I;

(10) the above-mentioned signal that collects is passed through sensor signal processing module II, with the filtering of realization signal and the integrating function of acceleration signal;

The vehicle body acceleration signal that (11) obtain sensor signal processing module II, vehicle body vertical velocity signal, the moving stroke signal of vehicle body vertical deviation signal and suspension is passed to System Discrimination module III and is carried out System Discrimination, with real-time variation of monitoring out vehicle vertical vibration systematic parameter;

(12) systematic parameter that identification obtains according to step (three) is adjusted the systematic parameter in the self-adaptation state observation module IV in real time;

(13) pass to self-adaptation state observation module IV with the moving stroke signal of vehicle body acceleration signal, vehicle body vertical deviation signal and suspension, to realize the identification of road pavement input;

(14) pass to space, road surface course generation module V with the road surface time history signal and the vehicle speed signal that pick out in the step (five), to realize that the road surface time history is to the conversion of space course;

(15) pass to road surface characteristic parameter extraction module VI with space, the road surface course signal that step (six) obtains, to extract the parameter that can characterize road surface characteristic;

(16) set threshold value (setting of threshold value can be demarcated by the test on the different road surfaces) for the road surface characteristic parameter in the sorter VII of road surface, the parameter that step (seven) obtains is passed through road surface sorter VII, from the classification of final realization road pavement.

Described collecting sensor signal module I, sensor signal processing module II, System Discrimination module III and self-adaptation state observation module IV are nucleus modules of the present invention, its job step also is the core procedure of the road surface recognition methods based on the System Discrimination of vehicle vertical vibration of the present invention, provides the detailed introduction of the job step of these several modules below respectively.

(A) collection vehicle vehicle body vertical vibration acceleration, the moving stroke of suspension and vehicle speed signal

The installation site of the moving stroke sensor 2 of acceleration transducer 1 and suspension as shown in Figure 3, wherein acceleration transducer 1 is contained on the vibration damper on the mount pad, the moving stroke sensor 2 of suspension is angular displacement sensor, the one end is connected on the subframe, the other end is connected in wheel, measures the moving stroke of suspension with the variation of its angular displacement.Vehicle speed signal generally can directly read from the CAN bus on middle-and-high-ranking car, also can obtain from meter panel of motor vehicle.

(B) processing of sensor signal

Here it mainly is acceleration transducer signals that the signal that relates to is handled, because System Discrimination module III requires vehicle body vertical velocity and absolute displacement signal to import as it, this just requires acceleration signal is done Integral Processing, because the acceleration signal that measures exists unpredictable zero point drift and noise pollution, and rate signal and displacement signal that direct integral is obtained do not have value.The present invention introduces the wave filter of band integrating function in signal processing module, the transport function of wave filter is as follows:

${\mathrm{<figure-callout\; id="1"\; label="G"\; filenames="HDA00003144073400022.png,HDA00003144073400032.png"\; state="\{\{state\}\}">G</figure-callout>}}_1=\frac{1}{{({\mathrm{<figure-callout\; id="2"\; label="S"\; filenames="HDA00003144073400011.png,HDA00003144073400022.png"\; state="\{\{state\}\}">S</figure-callout>}}^2+{2\mathrm{\&xi;}}_0{\mathrm{\&omega;}}_{0}S+{\mathrm{\&omega;}}_0^2)}^2},$ ${\mathrm{<figure-callout\; id="2"\; label="G"\; filenames="HDA00003144073400011.png,HDA00003144073400022.png"\; state="\{\{state\}\}">G</figure-callout>}}_2=\frac{S}{{({\mathrm{<figure-callout\; id="2"\; label="S"\; filenames="HDA00003144073400011.png,HDA00003144073400022.png"\; state="\{\{state\}\}">S</figure-callout>}}^2+{2\mathrm{\&xi;}}_0{\mathrm{\&omega;}}_{0}S+{\mathrm{\&omega;}}_0^2)}^2}$

Wherein, G ₁, G ₂Be respectively the integration filter of asking absolute displacement and speed, ω ₀Be the cutoff frequency of wave filter, and ω ₀Satisfy ω ₀＜＜ω ₁, ω ₁Be the natural frequency of vehicle body, ξ ₀Be damping ratio.

(C) vehicle vertical vibration System Discrimination

Owing to the parameter of active/semi-active suspension system can change along with the variation of load, load distribution and driving cycle, the present invention has introduced vehicle vertical vibration systematic parameter real-time identification module.

Here with the example of least square method of recursion identification 1/4 car vibrational system parameter (spring carried mass and ratio of damping) as an illustration, single-wheel vehicle vertical vibration system simplification is become as shown in Figure 4 1/4 Vehicular vibration system.

To the sprung mass row equation of motion:

Making the moving stroke of suspension is x, and the stroke of nonspring carried mass and spring carried mass is respectively x ₁, x ₂, x=x ₂-x ₁, have:

$M{\stackrel{\&CenterDot;\&CenterDot;}{x}}_2+C_s\stackrel{\&CenterDot;}{x}+K_{s}x=0$

Wherein, M is spring carried mass, and Cs is the suspension shock-absorber damping, and Ks is bearing spring rigidity;

Thereby

$-K_{s}x=\left[M{C}_s\right]\left[\begin{array}{c}{\stackrel{\&CenterDot;\&CenterDot;}{x}}_2\\ \stackrel{\&CenterDot;}{x}\end{array}\right]$

For the sake of simplicity, following formula can be write as

y＝θ ^Tψ

Wherein

Then can use Recursive Least Squares, estimate the θ value in the following formula.Concrete algorithm for estimating is as follows

$\stackrel{\&CenterDot;}{\widehat{\mathrm{\&theta;}}}=P(y-{\widehat{\mathrm{\&theta;}}}^T\mathrm{\&psi;})\mathrm{\&psi;}$

$\stackrel{\&CenterDot;}{P}=-P\frac{\mathrm{\&psi;}{\mathrm{\&psi;}}^{\mathrm{<figure-callout\; id="1"\; label="T"\; filenames="HDA00003144073400022.png,HDA00003144073400032.png"\; state="\{\{state\}\}">T</figure-callout>}}}{1+{\mathrm{\&psi;}}^T\mathrm{P\&psi;}}P$

P is one 2 * 2 symmetric matrix in the formula, the covariance that representative is estimated,

Be the minimum variance estimate of θ, P and

All want initialize.

(D) foundation of self-adaptation state observer

In conjunction with Fig. 4, this vibrational system also remains m, Kt after the System Discrimination through step (three), the several parameters of Ks are undetermined, but in fact these several quantitative changeizations are very little in motion for vehicle, can ignore its variation, them as constant, whole like this vibrational system parameter has just been determined, set up the state observer of the face input state of leading the way below, with the example that is established as of optimum state observer (Kalman filter), describe here.

In conjunction with Fig. 4, the selection mode variable

It is as follows to get the discrete system dynamic equation:

x _k+1＝G _kx _k+b _m,k

z _k＝H _kx _k+b _s,k

In the formula:

Be the state vector of system, and original state is 0;

Be the output vector of system, wherein: x _{2, k}-x _{1, k}: suspension moves stroke, can be recorded by height sensor;

Vehicle body acceleration can be recorded by acceleration transducer; x _{2, k}: the vehicle body absolute displacement can obtain 2 integrations of acceleration signal; x _{0, k}: the road surface elevation;

b _{M, k}And b _{S, k}Be respectively process error and measuring error vector, being assumed to be average here is 0, and incoherent white noise.

State matrix G and output matrix H are as follows:

$G=E+\mathrm{\&Delta;t}\left(\begin{array}{cccccc}0& 1& 0& 0& 0& 0\\ -\frac{\mathrm{Kt}}{M}& -\frac{\mathrm{Cs}}{M}& \frac{\mathrm{Kt}}{M}& -\frac{\mathrm{Cs}}{M}& 0& 0\\ 0& 0& 0& 1& 0& 0\\ \frac{\mathrm{Ks}}{m}& \frac{\mathrm{Cs}}{m}& -\frac{\mathrm{Ks}+\mathrm{Ku}}{m}& -\frac{\mathrm{Cs}}{m}& \frac{\mathrm{Ku}}{m}& 0\\ 0& 0& 0& 0& 0& 1\\ 0& 0& 0& 0& 0& 0\end{array}\right)$

$H=\left(\begin{array}{cccccc}1& 0& -1& 0& 0& 0\\ -\frac{\mathrm{Kt}}{M}& -\frac{\mathrm{Cs}}{M}& \frac{\mathrm{Kt}}{M}& \frac{\mathrm{Cs}}{M}& 0& 0\\ 1& 0& 0& 0& 0& 0\end{array}\right)$

Wherein, E is unit matrix, and Δ t is the sampling time.

According to above-mentioned equation design Kalman filter, the acceleration signal that sensor is recorded and moving stroke signal are as input.Can estimate the curve of road surface time history.

In addition, the detailed process of space, the described road surface of step (five) course generation is:

For the road surface time history that will obtain in the step (four) converts space, road surface course to, to use the signal of vehicle speed sensor here.For reducing the operand of ECU, the present invention introduces trigger pip Flag, and has only T in the regulation unit interval period T ₀In time period, space, road surface course generation module V triggers, be illustrated in figure 5 as for the process flow diagram that the road surface time history is converted to the space course: when trigger pip Flag is 1, read vehicle speed signal v, and begin to calculate S=S+v Δ t, simultaneously with the road surface time history data storage of calculating; If S 〉=S ₀The time, utilize method of interpolation that the road surface time history of storage is changed into the space course, and then S is put 0; If S＜S ₀, then store the road surface time history in the S distance section, and return and continue to read vehicle speed signal v.When trigger pip Flag is 0, do not carry out reading and changing of data.

The detailed process of the described road surface characteristic Parameter Extraction of step (six) is:

The method that eigenwert is extracted has a lot, can estimate the energy of the signal that the road surface comprises with the power Zymography in different frequency range, also the primary band in-scope of road surface signal input can be estimated with the single order zero crossing method of passing through, even statistical study such as root-mean-square value, maximal value can be directly got to signal.It is worthy of note that road surface characteristic parameter extraction module VI should adopt same time trigger with space, road surface course generation module V.

The detailed process of the foundation of the described road surface of step (seven) sorter is:

At the different characteristic parameter extracting method that step (six) adopts, can set up different road surface sorters.For example pass through method at the single order zero crossing, the road surface sorter only need provide 4 frequency range: 0-4Hz, 4-8Hz, 8-12Hz, more than the 12Hz, and the feature dominant frequency parameter that obtains in the determining step (six) can directly draw the road surface classification results in which frequency range.

It is worthy of note that mode given here is just as a specific embodiment of the present invention, obviously, described embodiment only is the present invention's part embodiment, rather than whole embodiment.Based on the embodiment among the present invention, those of ordinary skills belong to the scope of protection of the invention not making the every other embodiment that obtains under the creative work prerequisite.

Claims (10)

1. road surface recognition system based on the System Discrimination of vehicle vertical vibration is characterized in that:

By collecting sensor signal module (I), sensor signal processing module (II), System Discrimination module (III), self-adaptation state observation module (IV), space, road surface course generation module (V), road surface characteristic parameter extraction module (VI) and road surface sorter (VII) are formed, described sensor signal processing module (II), System Discrimination module (III), self-adaptation state observation module (IV), space, road surface course generation module (V), road surface characteristic parameter extraction module (VI) and road surface sorter (VII) are integrated in the ECU chip (X), and carry out each other communication by bus, described collecting sensor signal module (I) is carried out data transmission by wire harness and ECU chip (X);

Described collecting sensor signal module (I) comprises body-acceleration sensor (1), suspension moving stroke sensor (2), vehicle speed sensor (3) and A/D converter (4), the vertical acceleration signal of described body-acceleration sensor (1) collection vehicle vertical direction; Suspension moves stroke sensor (2) and adopts angular displacement sensor, measures the moving stroke of suspension with the variation of its angular displacement; Vehicle speed sensor (3) is gathered vehicle speed signal; Described A/D converter (4) has the high conversion rate characteristic.

2. a kind of road surface recognition system based on the System Discrimination of vehicle vertical vibration according to claim 1 is characterized in that:

Described sensor signal processing module (II) is carried out filtering, is removed noise pollution the sensor signal that collects, and does the processing in early stage of data simultaneously for System Discrimination module (III); Sensor signal processing module (II) contains the integral filtering module, and the acceleration signal that measures is carried out integration and filters zero point drift and noise pollution.

3. a kind of road surface recognition system based on the System Discrimination of vehicle vertical vibration according to claim 1 is characterized in that:

The systematic parameter of described System Discrimination module (III) real-time identification vehicle vertical vibration is for the realization of self-adaptation state observation module IV provides the basis;

Described self-adaptation state observation module (IV) can adapt to the parameter creep of different loads state, different rows turner condition Vehicular system.

4. a kind of road surface recognition system based on the System Discrimination of vehicle vertical vibration according to claim 1 is characterized in that:

Space, described road surface course generation module (V) is integrated the output signal of self-adaptation state observation module (IV) and vehicle speed sensor (3), converts the road surface time history to space, road surface course;

Characteristic parameter extraction is done in the signal output of described road surface characteristic parameter extraction module (VI) road pavement space course generation module (V), represents its road surface characteristic with one or several characteristic index;

Described road surface sorter VII is the characteristic parameter output definition threshold value of road surface characteristic parameter extraction module (VI), and then distinguishes different road surface types.

5. road surface recognition methods based on the System Discrimination of vehicle vertical vibration may further comprise the steps:

(1) gathers vehicle body vertical acceleration signal, the moving stroke signal of suspension and vehicle speed signal by collecting sensor signal module (I);

(2) the above-mentioned signal that collects is passed through sensor signal processing module (II), with the filtering of realization signal and the integrating function of acceleration signal;

(3) the vehicle body acceleration signal that sensor signal processing module (II) is obtained, vehicle body vertical velocity signal, the moving stroke signal of vehicle body vertical deviation signal and suspension is passed to System Discrimination module (III) and is carried out System Discrimination, monitors out the variation of vehicle vertical vibration systematic parameter in real time;

(4) systematic parameter that identification obtains according to step (three) is adjusted the systematic parameter in the self-adaptation state observation module (IV) in real time;

(5) the moving stroke signal of vehicle body acceleration signal, vehicle body vertical deviation signal and suspension is passed to self-adaptation state observation module (IV), to realize the identification of road pavement input;

(6) the road surface time history signal and the vehicle speed signal that pick out in the step (five) are passed to space, road surface course generation module (V), to realize that the road surface time history is to the conversion of space course;

(7) space, the road surface course signal that step (six) is obtained is passed to road surface characteristic parameter extraction module (VI), to extract the parameter that can characterize road surface characteristic;

(8) set threshold value for the road surface characteristic parameter in road surface sorter (VII), wherein threshold value can be demarcated by the test on the different road surfaces, the parameter that step (seven) obtains is passed through road surface sorter (VII), from the classification of final realization road pavement.

6. a kind of road surface recognition methods based on the System Discrimination of vehicle vertical vibration according to claim 5, it is characterized in that: the sensor signal processing module (II) described in the described step (two) is introduced the wave filter of band integrating function, and the transport function of wave filter is as follows:

$G_1=\frac{1}{{(S^2+{2\mathrm{\&xi;}}_0{\mathrm{\&omega;}}_{0}S+{\mathrm{\&omega;}}_0^2)}^2},$ $G_2=\frac{S}{{(S^2+{2\mathrm{\&xi;}}_0{\mathrm{\&omega;}}_{0}S+{\mathrm{\&omega;}}_0^2)}^2}$

Wherein, G ₁, G ₂Be respectively the integration filter of asking absolute displacement and speed, ω ₀Be the cutoff frequency of wave filter, and ω ₀Satisfy ω ₀＜＜ω ₁, ω ₁Be the natural frequency of vehicle body, ξ ₀Be damping ratio.

7. a kind of road surface recognition methods based on the System Discrimination of vehicle vertical vibration according to claim 5, it is characterized in that: the variation of monitoring out vehicle vertical vibration systematic parameter in the described step (three) in real time realizes by the following method:

Utilize least square method of recursion identification 1/4 car vibrational system parameter, to the sprung mass row equation of motion:

Making the moving stroke of suspension is x, and the stroke of nonspring carried mass and spring carried mass is respectively x ₁, x ₂, x=x ₂-x ₁, have:

$M{\stackrel{\&CenterDot;\&CenterDot;}{x}}_2+C_s\stackrel{\&CenterDot;}{x}{K}_{s}x=0$

Wherein, M is spring carried mass, and Cs is the suspension shock-absorber damping, and Ks is bearing spring rigidity;

Thereby

$-K_{s}x=\begin{array}{c}\left[M{C}_s\right]\left[\begin{array}{c}{\stackrel{\&CenterDot;\&CenterDot;}{x}}_2\\ \stackrel{\&CenterDot;}{x}\end{array}\right]\end{array}$

For the sake of simplicity, following formula can be write as

y＝θ ^Tψ

Wherein

Then can use Recursive Least Squares, estimate the θ value in the following formula, concrete algorithm for estimating is as follows

$\stackrel{\&CenterDot;}{\widehat{\mathrm{\&theta;}}}=P(y-{\widehat{\mathrm{\&theta;}}}^T\mathrm{\&psi;})\mathrm{\&psi;}$

$\stackrel{\&CenterDot;}{P}=-P\frac{\mathrm{\&psi;}{\mathrm{\&psi;}}^T}{1+{\mathrm{\&psi;}}^T\mathrm{P\&psi;}}P$

P is one 2 * 2 symmetric matrix in the formula, the covariance that representative is estimated,

Be the minimum variance estimate of θ, P and

All want initialize.

8. a kind of road surface recognition methods based on the System Discrimination of vehicle vertical vibration according to claim 5, it is characterized in that: the systematic parameter of adjusting in the self-adaptation state observation module (IV) in real time in the described step (four) is namely set up the self-adaptation state observer, utilizes Kalman filter to set up the state observer of the face input state of leading the way:

The selection mode variable

It is as follows to get the discrete system dynamic equation:

x _k+1＝G _kx _k+b _m,k

z _k＝H _kx _k+b _s,k

In the formula:

Be the state vector of system, and original state is 0;

Be the output vector of system, wherein: x _{2, k}-x _{1, k}: suspension moves stroke, can be recorded by height sensor;

Vehicle body acceleration can be recorded by acceleration transducer; x _{2, k}: the vehicle body absolute displacement can obtain 2 integrations of acceleration signal; x _{0, k}: the road surface elevation;

b _{M, k}And b _{S, k}Be respectively process error and measuring error vector, being assumed to be average here is 0, and incoherent white noise;

State matrix G and output matrix H are as follows:

$G=E+\mathrm{\&Delta;t}\left(\begin{array}{cccccc}0& 1& 0& 0& 0& 0\\ -\frac{\mathrm{Kt}}{M}& -\frac{\mathrm{Cs}}{M}& \frac{\mathrm{Kt}}{M}& -\frac{\mathrm{Cs}}{M}& 0& 0\\ 0& 0& 0& 1& 0& 0\\ \frac{\mathrm{Ks}}{m}& \frac{\mathrm{Cs}}{m}& -\frac{\mathrm{Ks}+\mathrm{Ku}}{m}& -\frac{\mathrm{Cs}}{m}& \frac{\mathrm{Ku}}{m}& 0\\ 0& 0& 0& 0& 0& 1\\ 0& 0& 0& 0& 0& 0\end{array}\right)$

$H=\left(\begin{array}{cccccc}1& 0& -1& 0& 0& 0\\ -\frac{\mathrm{Kt}}{M}& -\frac{\mathrm{Cs}}{M}& \frac{\mathrm{Kt}}{M}& \frac{\mathrm{Cs}}{M}& 0& 0\\ 1& 0& 0& 0& 0& 0\end{array}\right)$

Wherein, E is unit matrix, and Δ t is the sampling time;

According to above-mentioned equation design Kalman filter, the acceleration signal that sensor is recorded and moving stroke signal can estimate the curve of road surface time history as input.

9. a kind of road surface recognition methods based on the System Discrimination of vehicle vertical vibration according to claim 5 is characterized in that: the detailed process that space, the described road surface of described step (five) course generates is:

Introduce trigger pip Flag, and have only T in the regulation unit interval period T ₀In time period, space, road surface course generation module V triggers, and when trigger pip Flag is 1, reads vehicle speed signal v, and begins to calculate S=S+v Δ t, simultaneously with the road surface time history data storage of calculating; If S 〉=S ₀The time, utilize method of interpolation that the road surface time history of storage is changed into the space course, and then S is put 0; If S＜S ₀, then store the road surface time history in the S distance section, and return and continue to read vehicle speed signal v; When trigger pip Flag is 0, do not carry out reading and changing of data.

10. a kind of road surface recognition methods based on the System Discrimination of vehicle vertical vibration according to claim 5, it is characterized in that: during the described road surface characteristic Parameter Extraction of described step (six), eigenwert is extracted and can be adopted power spectrumanalysis method, single order zero crossing to pass through method or adopt and directly signal is got root-mean-square value, peaked statistical analysis method;

During the road surface characteristic Parameter Extraction, road surface characteristic parameter extraction module (VI) should adopt same time trigger with space, road surface course generation module (V);

At the different characteristic parameter extracting method that step (six) adopts, described step (seven) can be set up different road surface sorters; Pass through method at the single order zero crossing, the road surface sorter only need provide 4 frequency range: 0-4Hz, 4-8Hz, 8-12Hz, more than the 12Hz, and the feature dominant frequency parameter that obtains in the determining step (six) can directly draw the road surface classification results in which frequency range.

Images