WO2019244379A1

WO2019244379A1 - Road surface condition determination method and road surface condition determination device

Info

Publication number: WO2019244379A1
Application number: PCT/JP2018/047157
Authority: WO
Inventors: 啓太石井; 剛真砂; 嵩人後藤
Original assignee: 株式会社ブリヂストン
Priority date: 2018-06-22
Filing date: 2018-12-21
Publication date: 2019-12-26
Also published as: JP2019218023A

Abstract

As a result of the present invention, a chronological waveform having a time length TK that is within the range of 13T/40 < TK < 19T/40 and includes a stepping point time tf, a kicking point time tk, or a ground contact center point time tc can be cut out from a chronological waveform for detected tire vibration, a window function can be applied to a cut-out waveform being this cut-out chronological waveform, a cut-out waveform for each time window can be extracted, and a feature vector Xi for each time window can be calculated, when calculating a kernel function KA from the feature vector Xi for each time window and a reference feature vector YASVJ being a feature vector for each time window found for each road surface state calculated beforehand, said kernel function KA being calculated after: windowing chronological waveforms for tire vibration detected by an acceleration sensor, for time T, by using a windowing means; extracting the chronological waveform for tire vibration for each time window; and calculating the feature vector Xi for each time window.

Description

Road surface condition determination method and road surface condition determination device

The present invention relates to a method and an apparatus for determining a road surface state using only data of a time-series waveform of tire vibration during running.

Conventionally, as a method of determining a road surface state using only data of a time series waveform of a tire vibration during traveling, a time window calculated from a time series waveform extracted by multiplying a time series waveform of a tire vibration by a window function is used. A method has been proposed in which a road surface state is determined using a kernel function calculated from a characteristic amount for each road surface and a reference characteristic amount, which is a characteristic amount for each time window, obtained in advance for each road surface state.
The reference feature amount is obtained by machine learning (SVM) using, as learning data, a feature amount for each time window calculated from a time series waveform of tire vibration previously obtained for each road surface condition (for example, see Patent Document 1). ).

JP 2014-35279 A

However, although the time expansion / contraction is an operation necessary for comparing the acquired time-series waveforms, there is a problem that the calculation amount is large, the calculation time is long, and the processing becomes very heavy. .

The present invention has been made in view of the conventional problems, and provides a road surface state determination method and a road surface state determination device that can ensure the road surface state determination accuracy even when the amount of time expansion / contraction calculation is reduced. The purpose is to:

The present invention provides a step (a) of detecting a vibration of a tire during running, a step (b) of extracting a time-series waveform of the detected tire vibration, and a step of adding a predetermined time width to the time-series waveform of the tire vibration. (C) extracting a time-series waveform for each time window by multiplying by the window function, (d) calculating a feature amount from the time-series waveform for each time window, and calculating in the step (d). Calculating a kernel function from the calculated feature values for each time window and a reference feature value selected from the feature values for each time window calculated from the time series waveform of the tire vibration obtained in advance for each road surface condition ( e) and a step (f) of determining the state of the running road surface based on the value of the identification function using the kernel function, wherein the stepping on the time-series waveform of the tire vibration is performed. point Time to t _f, the trailing time of t _{k points,} the time t _f and the time t _k time of grounding the center point is an intermediate time and t _c, when the period of the time series waveform is T In the step (c), the time length T _K including any one of the times t _f , t _k, or t _c from the time-series waveform of the detected tire vibration is 13T / 40 <. A time-series waveform in the range of T _K <19T / 40 is cut out, a window function is applied to the cut-out waveform, which is the cut-out time-series waveform, to extract a time-series waveform for each time window. In the step (e), The time length _{Tc, which} is cut out from the time-series waveform of the tire vibration determined in advance for each time window and the road surface condition, is in the range of 13T / 40 <T _K <19T / 40. Reference selected from feature values for each time window calculated from reference cutout waveform The feature is that a kernel function is calculated from the feature amount.
This makes it possible to reduce the number of reference features used for calculating the kernel function K (X, Y), thereby increasing the calculation speed while ensuring the road surface state determination accuracy.

Note that, as the feature vector X _i , a vibration level of a specific frequency band of a cutout waveform for each time window extracted by applying the window function, a time-varying variance of a vibration level of the specific frequency band, and the cutout waveform , One or more, or all, of the cepstrum coefficients. Further, the vibration level of the specific frequency band is obtained by passing a frequency spectrum of a cutout waveform for each time window extracted by applying the window function or a cutout waveform for each time window extracted by applying the window function through a bandpass filter. It can be obtained from the obtained time-series waveform.
Further, if the kernel function is a global alignment kernel function, a dynamic time warping kernel function, or an operation value of the kernel function, it is possible to improve the determination accuracy of the road surface state.

Further, the present invention is a road surface condition determination device that detects the vibration of a running tire and determines the condition of a road surface on which the tire is running, the device being disposed on an air chamber side of an inner liner portion of a tire tread portion. Tire vibration detecting means for detecting the vibration of the running tire, the time of the stepping point in the time series waveform of the tire vibration is t _f , the time of the kicking point is tk _{, the} time t _f and the time the time of grounding the center point is an intermediate time t _k and t _c, when the period of the time series waveform is T, from the time-series waveform of the detected tire vibration, the time t _f, t _k, Alternatively, a waveform extracting means for extracting a time-series waveform having a time length T _K including any time of t _c in the range of 13T / 40 <T _K <19T / 40; When a window function is applied to an extracted waveform A windowing means for extracting a time-series waveform for each inter-window, and a feature quantity having a component of a vibration level of a specific frequency or a feature quantity having a function of the vibration level as a component in the extracted cut-out waveform for each time window. The feature amount calculation means to be calculated and the time length T _K extracted from the time series waveform of the tire vibration for each road surface condition which is obtained in advance for each road surface condition is 13T / 40 <T _K <19T / 40. Storage means for storing a reference feature value selected from feature values for each time window calculated from a reference cutout waveform that is a time-series waveform in a range and a Lagrange undetermined multiplier corresponding to the reference feature value, Kernel function calculation means for calculating a kernel function from the feature quantity for each time window calculated by the feature quantity calculation means and the reference feature quantity stored in the storage means, and a value of an identification function using the kernel function Road surface condition determining means for determining the road surface condition based on
By adopting such a configuration, it is possible to obtain a road surface state determination device that can ensure the road surface state determination accuracy even when the amount of time expansion / contraction calculation is reduced.

The summary of the present invention does not list all necessary features of the present invention, and a sub-combination of these features may also be an invention.

It is a functional block diagram of a road surface condition discriminating apparatus according to the present embodiment. It is a figure showing an example of a mounting position of an acceleration sensor. It is a figure showing an example of a time series waveform of a tire vibration, and how to cut out a waveform. It is a figure showing the method of calculating a feature vector from a time series waveform of tire vibration. It is a schematic diagram which shows an input space. It is a figure which shows the road surface feature vector of a DRY road surface and the road surface feature vector of a WET road surface in an input space. FIG. 4 is a diagram illustrating a method for calculating a GA kernel. 4 is a flowchart illustrating a road surface state determination method according to the present invention. FIG. 4 is a diagram illustrating a relationship between a cutout width of a waveform and determination accuracy. FIG. 9 is a diagram illustrating another example of how to cut out a waveform.

Embodiment FIG. 1 is a diagram illustrating a configuration of a road surface state determination device 10 according to the present embodiment.
The road surface condition determination device 10 includes an acceleration sensor 11 as a tire vibration detection unit, a waveform cutting unit 12, a windowing unit 13, a feature vector calculation unit 14, a storage unit 15, a kernel function calculation unit 16, a road surface A state determination unit 17 is provided for performing two road surface determinations as to whether the road surface on which the tire 20 is traveling is a DRY road surface or a WET road surface.
Each unit from the waveform extracting unit 12 to the road surface state determining unit 17 is composed of, for example, computer software and a memory such as a RAM.
As shown in FIG. 2, the acceleration sensor 11 is disposed integrally at a substantially central portion of the inner liner portion 21 of the tire 20 on the tire air chamber 22 side, and detects vibration of the tire 20 due to input from a road surface. The tire vibration signal output from the acceleration sensor 11 is, for example, amplified by an amplifier, converted into a digital signal, and sent to the waveform cutout unit 12.

The waveform cutout unit 12 cuts out a portion used for determination of a road surface condition from the time series waveform of the tire vibration detected by the acceleration sensor 11 and sends the portion to the windowing unit 13.
A diagram showing an example of FIG. 3 (a) time-series waveform of tire vibration, time-series waveform of tire vibration has a large peak in the vicinity of point P _k out kicking the neighboring depression point P _f, and, regions R _k after kicking after previous depression front region R _f where land portion of the tire 20 contacts the ground, the land portion of the tire 20 is separated from the road surface, and, grounding land portion of the tire 20 is grounded on the road surface in the region R _s, different vibrations may appear by the road surface condition. Below, the area of to the area R _k after kicking from depression before area R _f of the road surface area. On the other hand, the area before the stepping-in area _Rf and the area after the kicking-out area _Rk (hereinafter referred to as “out-of-road area”) are hardly affected by the road surface, so that the vibration level is small and the road surface information is low. Not included.
Here, the period of the time-series waveform of the vibration, which is the time for the tire 20 to make one rotation, is represented by T. The period T may, for example, as shown in FIG. 3 (b), the time interval between temporally adjacent two depression point P _f, or may be calculated from the time interval between the two kick-out point P _k.
In this example, from the time series waveform of the tire vibration, the time length T _K including the time t _{f at the} depression point P _f is 13T / 40 <T _K <19T / 40 (T: the time series waveform of the tire vibration A time series waveform in the range of (period) is cut out, and the cut out waveform which is the cut out time series waveform is sent to the windowing means 13. Hereinafter, the above T _{K is referred} to as a cutout width.

FIG. 3B shows an example of how to extract a time series waveform of tire vibration.
In the figure, t _f is the time of the depressed point P _f , t _k is the time of the kick-out point P _k , and t _c is the time of the ground contact center point which is an intermediate time between the time t _f and the time t _k .
Conventionally, a waveform (T _K = 3T / 4) corresponding to 3T / 8 before and after the time t _c of the ground contact center is cut out, and the feature vector X _i calculated from the waveform (waveform of the road surface area) and the road surface The road surface condition was determined by comparing the reference feature vector Y _ASV of A with the GA kernel.
On the other hand, in the present example, the cutout width T _K including the time t _f of the stepping point P _f starting from the time 3 T / 8 before the time t _c of the ground contact center point is 13 T / 40, for example. , 3T / 8, 19T / 40, etc., and cut out waveforms, which are cut out waveforms, are sent to the windowing means 13.
In the time series waveform of FIG. 3B, when T _K = 13T / 40, the end point is the time of the depressed point t _f , and when T _K = 3T / 8, the end point is the ground contact point. the time t _c. In addition, in the case of the 19T / 40 is, it is later than the time t _k of the end point is kick-out point P _k.

As shown in FIG. 4, the windowing means 13 windows the extracted waveform with a predetermined time width (also referred to as a time window width) ΔT, and extracts a waveform of the tire vibration for each time window. To the feature vector calculation means 14.
In the figure, T _K is a cutout width, and here, T _K = 3T / 8.
As shown in FIG. 4, the feature vector calculation means 14 calculates a feature vector X _i (i = 1 to N; N is a time series for each extracted time window for each of the extracted waveforms of the extracted time windows. The number of waveforms) is calculated.
As described above, when the cutout width T _K is reduced, the number of time windows is reduced, and the number of feature vectors X _i is also reduced.
In this example, the time series waveforms of the tire vibration are used as the feature vectors X _{i to be} calculated, and the band-pass filters of 0-0.5 kHz, 0.5-1 kHz, 1-2 kHz, 2-3 kHz, 3-4 kHz, and 4-5 kHz, respectively. The vibration level (power value of the filtered wave) a _ik (k = 1 to 6) of the specific frequency band obtained by passing through each of them was used. Feature vectors, the number of _{_{X i = (a i1, a}} i2, a i3, a i4, a i5, a i6) , the feature vector X _i is the N.
Figure 5 is a schematic diagram showing the input space of feature vectors X _i, each axis represents the vibration level a _ik of a specific frequency band, which is a feature quantity, each point representing a feature vector X _i. The actual input space is a seven-dimensional space when combined with the time axis because the number of specific frequency bands is three. However, the figure is expressed in two dimensions (the horizontal axis is a ₁ and the vertical axis is a ₂ ).
In the figure, a set of feature vectors X _i when the group C travels the DRY road, when group C be the set of _i 'is the feature vector X of when traveling the WET road', and the group C If the tire can be distinguished from the group C ′, it can be determined whether the road on which the tires are traveling is a DRY road surface or a WET road surface.

The storage means 15 stores a DW identification model for identifying a DRY road surface and a WET road surface, which has been obtained in advance.
The DW identification model includes a reference feature vector Y _AK (y _jk ), which is a reference feature amount for separating a DRY road surface and a WET road surface by an identification function f (x) representing a separation hyperplane, and a reference feature vector Y _AK ( y _jk ) and a Lagrange multiplier λ _A.
The reference feature vectors Y _AK (y _jk ) and λ _A are usually the tire vibrations obtained by running a test vehicle equipped with a tire equipped with an acceleration sensor at various speeds on a DRY road surface and a WET road surface. The road surface feature vector Y _A (y _jk ), which is a feature vector for each time window calculated from the time-series waveform of the above, is obtained by learning using as input data.
In the example, the feature vector for each time series time was calculated from the cut waveforms cut from the waveform windows of tire vibration as a road surface feature vector Y _A (y _jk), seeking the reference feature vector Y _AK (y _jk) and lambda _A Therefore, the number of reference feature vectors Y _AK (y _jk ) also decreases.
Note that the tire size used for learning may be one type or a plurality of types.
The subscript A of the reference feature vector Y _AK (y _jk ) indicates DRY or WET.
The subscript j (j = 1 to M) indicates the window number of the time-series waveform extracted for each time window, and the subscript k indicates a vector component (k = 1 to 6). That is, _{_{y jk = (a j1, a}} j2, a j3, a j4, a j5, a j6)).
When the global alignment kernel function is used as in this example, the reference feature vector Y _AK (y _jk ) is the number of dimensions of the vector y _i (here, 6 × M (M; number of windows)). Of the matrix.
Hereinafter, the road feature vector Y _A a (y _jk) and the reference feature vector Y _AK (y _jk), respectively, referred to as Y _A, Y _AK.

The method of calculating the road surface feature vector Y _A is the same as the feature vector X _j described above, for example, if the reference feature vectors Y _D of DRY road, cut waveform cut out from the time-series waveform of tire vibrations when traveling along DRY road Is windowed with a time width ΔT, a cut-out waveform is extracted for each time window, and a DRY road surface feature vector Y _D is calculated for each of the extracted cut-out waveforms for each time window. Similarly, the WET road surface feature vector Y _W is calculated from a cut-out waveform for each time window when traveling on a WET road surface.
The reference feature vector Y _AK is a feature vector selected as a support vector by a support vector machine (SVM) using the DRY road surface feature vector Y _D and the WET road surface feature vector Y _W as learning data.
Here, it is important that the time width ΔT has the same value as the time width ΔT when the feature vector _Xj is obtained. If the time width T is constant, the number M of cut-out waveforms of the time window differs depending on the tire type and the vehicle speed. That is, the number M of time window cutout waveform of the reference symptom vector Y _A does not necessarily coincide with the number N of time windows of the cutout waveform of the feature vector X _j. For example, a tire species is the same, if slower than the vehicle speed when the vehicle speed when determining the feature vector X _j to determine the road feature vector Y _A is a M <N when M> N, and the fast .

Figure 6 is a conceptual diagram showing a DRY road feature vector Y _D and WET road feature vector Y _W in the input space, black circles in the figure is DRY road, open circles are WET road.
As described above, although a DRY road feature vectors Y _D also WET road feature vector Y _W also matrices, for explaining how to determine the decision boundary of the group, in FIG. 6, DRY road feature vectors Y _D and WET The road surface feature vector Y _W is shown as a two-dimensional vector.
Group identification boundaries generally do not allow linear separation. Thus, by using the kernel method, the road surface feature vectors Y _V and Y _W are mapped to a high-dimensional feature space by a non-linear mapping φ to perform linear separation, so that the road surface feature vectors Y _D and Y _W are obtained in the original input space. Non-linear classification is performed.
In order to distinguish between a DRY road surface and a WET road surface, a margin is provided for an identification function f (x) that is a separating hyperplane that separates the DRY road surface feature vector Y _Dj and the WET road surface feature vector Y _Wj . The DRY road surface and the WET road surface can be accurately distinguished.
The margin refers to the distance from the separation hyperplane to the nearest sample, and the separation hyperplane that is the identification boundary is f (x) = 0. The DRY road surface feature vectors Y _Dj are all in the region of f (x) ≧ + 1, and the WET road surface feature vectors Y _Wj are all in the region of f (x) ≦ −1.

Next, using the data set X = (x ₁ , x ₂ ,..., X _n ) and the belonging class z = {1, −1}, the optimal identification function f (x) = w for identifying the data. Find ^T φ (x) -b. Here, w is a vector representing a weight coefficient, and b is a constant.
The data is DRY road feature vectors Y _Dj and WET road feature vector Y _Wj, with data DRY road indicated by chi ₁ belongs class z = 1 is the figure, WET indicated by z = -1 is chi ₂ This is road surface data. f (x) = 0 is the identification boundary, and 1 / || w || is the distance between the road surface feature vector Y _Aj (A = D, W) and f (x) = 0.
The discriminant function f (x) = w ^T φ (x) -b is optimized using, for example, the Lagrange undetermined multiplier method. The optimization problem is replaced by the following equations (1) and (2).

Here, α and β are indexes of a plurality of learning data. Further, λ is a Lagrange multiplier, and the road surface feature vector Y _Aj with λ = 0 is vector data that is not involved in the identification function f (x) (not a support vector).
Here, by replacing the inner product φ ^T (x _α ) φ (x _β ) with the kernel function K (x _α , x _β ), the discriminant function f (x) = w ^T φ (x) −b can be nonlinearized.
Note that φ ^T (x _α ) φ (x _β ) is an inner product after x _α and x _β are mapped to a high-dimensional space by a mapping φ.
The Lagrange multiplier λ can be obtained by using an optimization algorithm such as the steepest descent method or SMO (Sequential Minimal Optimization) for the above equation (2).
Thus, without asking the inner product phi ^T a _{_{(x α) φ (x β}} ) directly, the kernel function K (x α, x _{_β)} if so replace the, there is no need to determine directly the inner product of high dimensional , Can greatly reduce the calculation time.

In this example, a kernel function K (x _α, x _β) as was used global alignment kernel function (GA kernel).
As shown in FIG. 7 and the following equations (3) and (4), the GA kernel K (x _α , x _β ) is a local kernel κ _ij (indicating the similarity between the feature vector x _α and the feature vector x _β ) x _αi , x _βj ) can be directly compared with time series waveforms having different time lengths using a function composed of the sum or the sum of the products. The local kernel κ _ij (x _αi , x _βj ) is obtained for each window of the time interval T.
FIG. 7 shows an example in which a GA kernel is obtained for a feature vector x _αi having three time windows and a feature vector x _β having two time windows.
In this example, since the cutout width is T _K = 3T / 8, the number of windows is half that of the conventional case where the cutout width is T _K = 3T / 4.

Here, || x _αi −x _βij || is a distance (norm) between feature vectors, and σ is a constant.

The kernel function calculating means 16 calculates the feature vector X _i calculated by the feature vector calculating means 14, the reference feature vector Y _{DK of the} DRY road surface stored in the storage means 15 and the reference feature vector Y _{WK of the} WET road surface. , DRYGA kernel K _D (X, Y _DK ) and WETGA kernel K _W (X, Y _WK ).
DRYGA kernel K _D (X, Y _DK), in the above formula (3) and (4), and a feature vector X _i calculated feature vector x _i.alpha the feature vector calculating section 14, DRY road feature vector x _beta local kernel _{_{_{κ ij (X i, Y DKj}}} ) when formed into a reference feature vector Y _DKj the sum or function consisting of total product of, WETGA kernel K _W (X, Y _WK) is the feature vector x _beta WET road _{Is a} function consisting of the sum or total product of the local kernels κ _ij (X _i , Y _WKj ) when the reference feature vector Y _WKj is By using these GA kernels K _D (X, Y _DK ) and K _W (X, Y _WK ), it is possible to directly compare time-series waveforms (cut-out waveforms) having different time lengths.
As described above, even when the number n of cut-out waveforms of the time window when the feature vector X _i is obtained is different from the number m of the cut-out waveforms of the time window when the road surface feature vector Y _{Aj is} obtained, it can be obtained a similarity between feature vectors X _i and the reference feature vector Y _ASVj.

In the road surface state determination means 17, the value of the identification function f _DW (x) using the kernel function K _D (X, Y _DK ) and the kernel function K _W (X, Y _WD ) shown in the following equation (5) The road surface condition is determined based on the above.

Here, N _DKV in the number of reference feature vectors Y _DKj of DRY road, N _WK is the number of reference feature vectors Y _Wkj the WET road.
In this example, the identification function f _DW is calculated, and if f _DW > 0, the road surface is determined to be a DRY road surface, and if f _DW <0, the road surface is determined to be a WET road surface.

Next, a method of determining the state of the road surface on which the tire 20 is traveling using the road surface state determination device 10 will be described with reference to the flowchart of FIG.
First, to detect the tire vibration generated by the input from the road surface on which the tire 20 is traveling by the acceleration sensor 11 (step S10), and from the time-series waveform of the detected tire vibration, the time t _k points out kick A time-series waveform including a cutout width T _K of T _K = 3T / 8 is cut out (step S11).
Then, the cut-out waveform, which is a time-series waveform of the cut-out tire vibration, is windowed with a preset time width ΔT to obtain a cut-out waveform for each time window. Here, the number of cut-out waveforms for each time window is set to m (step S12).
Next, a feature vector X _i = (x _i1 , x _i2 , x _i3 , x _i4 , x _i5 , x _i6 ) is calculated for each of the extracted time-series waveforms in each time window (step S13). In this example, the time width T is 3 msec. The number of feature vectors X _i is six.
Each of the components x _i1 to x _i6 (i = 1 to 6) of the feature vector X _i is the power value of the filtered wave of the time series waveform (cutout waveform) of the tire vibration, as described above.
Next, from the calculated feature vector X _i and the reference feature vector Y _AKj of the DRY road surface and the WET road surface recorded in the storage means 15, the reference feature vector Y _DK of the DRY road surface and the reference feature of the WET road surface are provided. A vector Y _WK is extracted (step S14), and a local kernel κ _ij (X _i , Y _AKj ) is calculated from the reference feature vectors Y _DK and Y _WK and the feature vector X _i, and then the local kernel κ _ij ( The sum of X _i , Y _AKj ) is obtained, and the GA kernel function K _A (X, Y _AK ) is calculated (step S15).
The GA kernel function K _D (X, Y _DK ) where A = D is a GA kernel function for a DRY road surface, and the GA kernel function K _W (X, Y _WK ) where A = W is a GA kernel function for a WET road surface. .
Then, an identification function f _DW (x) using the GA kernel function K _D of the DRY road surface and the GA kernel function K _W of the WET road surface is calculated (step S16). If f _DW > 0, the road surface is the DRY road surface And if f _DW <0, it is determined that the road surface is a WET road surface. (Step S17).
Since the cut-out waveform is as small as about 50% of the cut-out width of the road area waveform, the amount of data used for calculating the kernel function K (X, Y) can be reduced, and as a result, the calculation time can be shortened.

[Example]
The time calculated from the time series waveform (cut-out waveform) of the tire vibration when traveling on the DRY road surface and the WET road surface, in which the support vector of the DRY road surface and the support vector of the WET road surface are determined in advance. Road surface data, which is a feature of each window, was obtained as learning data by machine learning (SVM).
Specifically, as shown in Table 1 below, the used road surface data is divided into those for training (for Train) and those for test (for Test), and the support vector for the DRY road surface and the support vector for the WET road surface are After the determination, the support vector of the DRY road surface and the boundary surface of the support vector of the WET road surface were determined. At this time, the hyperparameters C and σ of the support vector machine were C = 2 and σ = 125, respectively.
At this time, the number of support vectors was 415 at the maximum.

FIG. 9 is a diagram showing the relationship between the cutout width and the discrimination accuracy. “All waveforms” indicates that the cutout waveform is the entire waveform (100%) of the road surface area, that is, the cutout width _Tk is 3T / 4. Refers to the case.
"Until after kick" is waveforms up just after the point out kick from depression front region R _f, is cut width T _k is 19T / 40 (63%).
The “ground center” is the first half of the entire waveform, and the cutout width T _k is 3T / 8 (50%).
"Until stepping" is a time-series waveform of depression front region R _f, is cut width T _k is 13T / 40 (43%).
As is clear from the figure, if 2 road discrimination DRY / WET, cut width T _k is even 13T / 40, it was confirmed that it is possible to obtain a sufficient determination accuracy.

Although the present invention has been described with reference to the embodiment and the examples, the technical scope of the present invention is not limited to the scope described in the embodiment. It will be apparent to those skilled in the art that various changes or improvements can be made to the embodiment. It is apparent from the appended claims that embodiments with such changes or improvements can be included in the technical scope of the present invention.

For example, in the above embodiment, the time length T _K including the time t _f of the depression point P _f is in the range of 13 T / 40 <T _K <19 T / 40 from the time series waveform of the tire vibration. cut out waveform has a cutout waveform is this cut-out time series waveform, FIG. 10 (a), the (b), the including time t _k of the kick-out point P _k, the time length T _K is, and the time-series waveform in the range of _{13T / 40 <T K <19T} / 40, the time t _k at time t _f and trailing point P _k of the depression point P _f of the time at which the ground center point of the intermediate Even when a time-series waveform including the time t _{c and} having a time length T _{K in} the range of 13T / 40 <T _K <19T / 40 is cut out and used as a waveform, the kernel function K is obtained while ensuring the road surface state determination accuracy. The calculation time of (X, Y) can be shortened.

Further, in the above-described embodiment, two road surface determinations as to whether the road surface on which the tire 20 is traveling are the DRY road surface or the WET road surface are performed using the DW identification model. Is used, it is possible to determine whether the road surface on which the tire 20 is traveling is a DRY road surface, a WET road surface, a SNOW road surface, or an ICE road surface.
Here, if A, A ′ = DRY, WET, SNOW, ICE (A ≠ A ′), the AA ′ identification model is an identification function f _{AA ′} (x ), A road surface feature vector Y _AK and a Lagrangian multiplier λ _{AA ′} , which are reference feature amounts for separation by A), and an A ′ road surface feature vector Y _A′K Lagrangian multiplier λ _{A ′ A.}
The reference feature values Y _ASV and λ _A are the values of tire vibration obtained by running a test vehicle equipped with a tire equipped with an acceleration sensor at various speeds on DRY, WET, SNOW, and ICE road surfaces. The road surface feature vector Y _A (y _jk ), which is a feature vector for each time window calculated from the series waveform, is obtained by learning using as input data.
The data of the A road surface, the data belonging to z = 1 indicated by chi ₁ in FIG. 6, data A 'road is data belonging to z = -1 shown by chi _2.

It should be noted that a Lagrange multiplier λ _A corresponding to the reference feature vector Y _AK exists for each identification model. For example, three Lagrangian multipliers λ _DW , λ _DS , and λ _DI corresponding to the DRY road surface feature vector Y _DK have different values. The same applies to other road surface feature vectors Y _WK , Y _SK , and Y _IK .
The method of calculating the GA kernel function K _A (X, Y _AK ) is the same as in the embodiment, and the GA kernel function K _D (X, Y _DK ) where A = D is the GA kernel function of the DRY road surface, and A = W A certain GA kernel function K _W (X, Y _WK ) is a GA kernel function for a WET road surface, a GA kernel function K _S (X, Y _SK ) where A = S is a GA kernel function for a SNOW road surface, and a GA where A = I The kernel function K _I (X, Y _IK ) is a GA kernel function for the ICE road surface.
The determination of the road surface state is performed using six identification functions f _{AA ′} (x) shown in the following equations (6) to (11).

As described above, if the discriminant function is f _{AA ′} (x), since the data on the road A is data belonging to z = 1 and the data on the road A ′ is data belonging to z = −1, From the six discriminant functions f _{AA ′} , the road surface can be determined as follows.
If f _DW > 0, f _DS > 0, f _DI > 0, it is determined that the road surface is a DRY road surface.
If f _DW <0, f _WS > 0, f _WI > 0, it is determined that the road surface is a WET road surface.
If f _DS <0, f _WS > 0, f _SI > 0, it is determined that the road surface is a SNOW road surface.
If f _DI <0, f _WI <0, f _SI <0, it is determined that the road surface is an ICE road surface.

Further, in the above-described embodiment, the tire vibration detecting means is the acceleration sensor 11, but other vibration detecting means such as a pressure sensor may be used. In addition, the acceleration sensor 11 may be installed at another location such as one at a position separated from the center of the tire in the width direction by a predetermined distance in the width direction, or may be installed in a block.
Further, in the embodiment, a feature vector X _i and the power value x _ik of filtration wave, variance, when the power value x _ik of filtration wave _{(log [x ik (t)} 2 + x ik ( t-1) ² ]) may be used. Alternatively, a feature vector X _i, Fourier coefficients a vibration level of a particular frequency band when the Fourier transform of the tire vibration time series waveform or may be cepstral coefficients. The cepstrum is obtained by assuming the waveform after Fourier transform as a spectrum waveform and performing Fourier transform again, or assuming that an AR spectrum is a waveform and obtaining an AR coefficient (LPC Cepstrum). Since the shape of the spectrum can be characterized without being affected, discrimination accuracy is improved as compared with the case where a frequency spectrum obtained by Fourier transform is used.
In the above embodiment, the GA kernel is used as the kernel function, but a dynamic time warping kernel function (DTW kernel) may be used. Alternatively, a GA kernel and a DTW kernel operation value may be used.

In summary, it can be described as follows. That is, the present invention provides a step (a) of detecting a vibration of a tire during running, a step (b) of extracting a time-series waveform of the detected tire vibration, and a step of extracting a predetermined time-series waveform of the tire vibration. (C) extracting a time-series waveform for each time window by applying a window function of a time width, (d) calculating a feature amount from the time-series waveform for each time window, and (d) The kernel function is calculated from the feature amount for each time window calculated in the above and the reference feature amount selected from the feature amount for each time window calculated from the time series waveform of the tire vibration obtained in advance for each road surface condition. A time-series waveform of the tire vibration, wherein the method further comprises a step (e) and a step (f) of determining a state of the running road surface based on a value of the identification function using the kernel function. In Time the t _f of only write point, time, t _k of the out-point _kick, the time t _f and the time t _k time of grounding the central point, which is an intermediate of the time of the t _c, the period of the time-series waveform When T is set, in the step (c), the time length T _K including any one of the times t _f , t _k, or t _{c is} _obtained from the time-series waveform of the detected tire vibration, A time-series waveform in the range of 13T / 40 <T _K <19T / 40 is cut out, and a time-series waveform for each time window is extracted by applying a window function to the cut-out waveform, which is the cut-out time-series waveform, In (e), the time length _Tc cut out from the feature amount for each time window and the time series waveform of the tire vibration obtained in advance for each road surface condition is 13T / 40 <T _K <19T / 40. Select from feature values for each time window calculated from reference cutout waveforms in the range A kernel function is calculated from the reference feature amount to be performed.
This makes it possible to reduce the number of reference features used for calculating the kernel function K (X, Y), thereby increasing the calculation speed while ensuring the road surface state determination accuracy.

Further, the present invention is a road surface condition determination device for detecting vibration of a tire during traveling and determining a condition of a road surface on which the tire travels, the device being disposed on an air chamber side of an inner liner portion of a tire tread portion. Tire vibration detecting means for detecting the vibration of the running tire, the time of the stepping point in the time series waveform of the tire vibration is t _f , the time of the kicking point is tk _{, the} time t _f and the time the time of grounding the center point is an intermediate time t _k and t _c, when the period of the time series waveform is T, from the time-series waveform of the detected tire vibration, the time t _f, t _k, Alternatively, a waveform extracting means for extracting a time-series waveform having a time length T _K including any time of t _c in the range of 13T / 40 <T _K <19T / 40; Apply a window function to an extracted waveform A windowing means for extracting a time-series waveform for each inter-window, and a feature quantity having a component of a vibration level of a specific frequency or a feature quantity having a function of the vibration level as a component in the extracted cut-out waveform for each time window. The feature amount calculation means to be calculated and the time length T _K extracted from the time series waveform of the tire vibration for each road surface condition which is obtained in advance for each road surface condition is 13T / 40 <T _K <19T / 40. Storage means for storing a reference feature value selected from feature values for each time window calculated from a reference cutout waveform that is a time-series waveform in a range and a Lagrange undetermined multiplier corresponding to the reference feature value, A kernel function calculating unit that calculates a kernel function from the feature amount for each time window calculated by the feature amount calculating unit and the reference feature amount stored in the storage unit, and a classification function using the kernel function. Characterized in that it comprises a road surface condition judging means for judging road surface condition based on.
By adopting such a configuration, it is possible to obtain a road surface state determination device that can ensure the road surface state determination accuracy even when the amount of time expansion / contraction calculation is reduced.

10 road surface condition determination device, 11 acceleration sensor,
12 waveform cutting means, 13 windowing means,
14 feature vector calculation means, 15 storage means,
16 kernel function calculating means, 17 road surface state determining means,
20 tires, 21 inner liner part, 22 tire chamber.

Claims

Detecting a vibration of the running tire (a), extracting a time-series waveform of the detected vibration of the tire (b), and applying a window function of a predetermined time width to the time-series waveform of the tire vibration. (C) extracting a time-series waveform for each time window by multiplying, (d) calculating a feature amount from the time-series waveform for each time window, and (E) calculating a kernel function from the characteristic amounts of the above and a reference characteristic amount selected from the characteristic amounts for each time window calculated from the time series waveform of the tire vibration obtained in advance for each road surface condition; (F) determining the state of the road surface on which the vehicle is traveling based on the value of the identification function using the kernel function;
In the road surface condition determination method provided with
The tire vibration when the time of depression point in the sequence waveform t f, kick t k time points out, the time t f and the time of intermediate time in which a ground center point of the time t k and t c, When the period of the time series waveform is T,
In the step (c),
From the time series waveform of the detected tire vibration, the time length T K including any one of the times t f , t k, or t c is in a range of 13T / 40 <T K <19T / 40. , And extract a time-series waveform for each time window by applying a window function to the cut-out waveform, which is the extracted time-series waveform,
In the step (e),
A feature amount of the time for each window, cut out from the time-series waveform of tire vibration obtained in advance for each road surface condition, the reference time length T c is in the range of 13T / 40 <T K <19T / 40 A road surface state determination method characterized by calculating a kernel function from a reference feature amount selected from feature amounts for each time window calculated from a cut-out waveform.
The feature quantity is
The vibration level of a specific frequency band of the cut-out waveform for each time window extracted by applying the window function,
Time-varying dispersion of the vibration level of the specific frequency band,
And any one, or a plurality, or all of the cepstrum coefficients of the cut-out waveform,
The vibration level of the specific frequency band is obtained by passing a frequency spectrum of a cutout waveform for each time window extracted by applying the window function or a cutout waveform for each time window extracted by applying the window function through a band-pass filter. The road surface state determination method according to claim 1, wherein the vibration level is a vibration level in a specific frequency band obtained from the time series waveform.
The method according to claim 1, wherein the kernel function is a global alignment kernel function, a dynamic time warping kernel function, or an operation value of the kernel function.
A road surface state determination device that detects vibration of a running tire and determines a state of a road surface on which the tire runs,
Tire vibration detection means arranged on the air chamber side of the inner liner portion of the tire tread portion to detect the vibration of the running tire,
The tire vibration when the time of depression point in the sequence waveform t f, kick t k time points out, the time t f and the time of intermediate time in which a ground center point of the time t k and t c, When the period of the time series waveform is T,
From the time series waveform of the detected tire vibration, the time length T K including any one of the times t f , t k, or t c is in a range of 13T / 40 <T K <19T / 40. A waveform extracting means for extracting a time-series waveform in
Windowing means for applying a window function to the cut-out waveform, which is the cut-out time-series waveform, to extract a time-series waveform for each time window,
A feature amount calculating unit that calculates a feature amount having a component of a vibration level of a specific frequency in the cut-out waveform for each extracted time window or a feature amount having a function of the vibration level as a component,
The time length T K extracted from the time-series waveform of the tire vibration for each road surface condition obtained in advance for each road surface condition is a time-series waveform having a range of 13T / 40 <T K <19T / 40. Storage means for storing a reference feature amount selected from the feature amount for each time window calculated from a certain reference cutout waveform and a Lagrange undetermined multiplier corresponding to the reference feature amount,
Kernel function calculation means for calculating a kernel function from the feature quantity for each time window calculated by the feature quantity calculation means and the reference feature quantity stored in the storage means;
A road surface state determination device for determining a road surface state based on a value of the identification function using the kernel function.