JP2000313327A - Road surface condition estimation device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a road surface condition estimation device hard to be influenced by a noise. SOLUTION: A vehicle speed signal which is computed in a wheel speed calculating circuit 35 from a signal output from a wheel speed sensor 10 and which includes a vibration component at the time of impulse-excited wheel cylinder hydraulic pressure in an appointed cycle is input in a frequency computing circuit 30 and an amplitude computing circuit 32 for computing a frequency which is highly correlated with several wave form data, and an amplitude which corresponds to the frequency. An attenuation rate computing circuit 34 computes an attenuation rate, and a road surface μ estimation device 14 judges a road surface μ gradient to be small when the attenuation rate is large and large when that is small on the basis of the attenuation rate computed in the attenuation rate computing circuit 34.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a road surface condition estimating apparatus, and more particularly to a method of estimating a road surface condition, for example, a vibration frequency change due to a tire change or the like by using a plurality of detected frequencies when estimating a road surface μ gradient. To a road surface state estimation device that is robust to

[0002]

2. Description of the Related Art The vibration of the wheel speed changes with respect to the road surface .mu. Gradient, and when the wheel is gripping the road surface, that is, when the road surface .mu. The vibration lasts relatively long. On the other hand, near the wheel grip limit, that is, when the road surface μ gradient is small, the vibration of the wheel speed converges relatively quickly.

Therefore, if the degree of attenuation of the wheel speed vibration is measured from the wheel speed signal, the road surface condition, for example, the road surface μ gradient can be detected.

[0004] The degree of attenuation of wheel speed vibration can be obtained from the amplitude value of the wheel speed signal. However,
If the amplitude value is directly obtained from the instantaneous value of the waveform data of the wheel speed signal, it is difficult to detect the degree of attenuation using only the instantaneous value because the waveform data contains sudden noise.

SUMMARY OF THE INVENTION The present invention has been made to solve the above-mentioned problems, and an apparatus for estimating a road surface state which is hardly affected by noise by detecting an amplitude as a characteristic parameter of a waveform by using a correlation calculation technique. The purpose is to provide.

[0006]

SUMMARY OF THE INVENTION To achieve the above object, the present invention provides a wheel speed detecting means for detecting a wheel speed and outputting a wheel speed signal; Calculates the correlation between the time-series data of the wheel speed signal and the waveform data of a plurality of predetermined frequencies having different frequencies, and calculates the resonance frequency of the wheel speed signal from the waveform data having a high correlation. In the resonance frequency calculation means, the resonance frequency calculated by the resonance frequency calculation means,
The time series data of the wheel speed signal in the state where the wheel speed signal includes the vibration component in the second cycle shorter than the first cycle and the waveform data of the frequency obtained by the resonance frequency calculation. An amplitude calculating means for calculating a correlation coefficient and calculating the amplitude of the wheel speed signal from the obtained correlation coefficient; and estimating a road surface state based on a degree of attenuation of the amplitude of the wheel speed signal calculated by the amplitude calculating means. Road surface state estimating means.

The frequency calculating means of the present invention calculates the correlation between time-series data of the wheel speed signal and a plurality of waveform data having different frequencies in a state where the wheel speed signal contains a vibration component in the first cycle. Is calculated, and the resonance frequency of the wheel speed signal is calculated from the waveform data having a high correlation. As the waveform data, sine wave and cosine wave data can be used.
The amplitude calculating means calculates, in a second cycle shorter than the first cycle, the amplitude of the wheel speed signal in a state where the wheel speed signal includes a vibration component at the frequency calculated by the frequency calculating means. I do. This amplitude can be calculated from the correlation coefficient between the time series data of the wheel speed signal and the waveform data of the vibration frequency calculated by the vibration frequency calculation means, and from the calculated correlation coefficient. As the waveform data of the amplitude calculation means,
Sine wave and cosine wave data different from the above sine wave and cosine wave can be used.

The road surface state estimating means estimates the road surface state based on the degree of attenuation of the amplitude calculated by the amplitude calculating means, for example, the attenuation rate or the attenuation ratio. As described above, the wheel vibration continues for a relatively long time when the wheel is gripping the road surface, and the vibration converges relatively quickly near the grip limit of the wheel. If the road surface condition is slippery and the degree of attenuation is slow, it can be determined that the road surface condition is not slippery.

In the present invention, it is effective to further provide a wheel cylinder oil pressure exciting means for exciting the wheel cylinder oil pressure when the amplitude of the wheel speed signal is small, and to excite the wheel vibration. As a result, a vibration component is included in the wheel speed signal. In the present invention, wheel vibration may be generated by using disturbance from the road surface without using the wheel cylinder hydraulic excitation unit.

[0010]

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram showing an embodiment of a road surface condition estimating apparatus according to the present invention.

In this embodiment, as shown in FIG. 1, a wheel is mounted so as to face a tooth of a rotating gear fixed to the wheel, and outputs an electric signal whose frequency changes in proportion to the wheel speed. A speed sensor 10 is provided. This wheel speed sensor 10 is attached to each of the four wheels of the vehicle, but only one is shown as a representative.

An output terminal of the wheel speed sensor 10 is connected to a wheel speed calculating circuit 35 for calculating a wheel speed from a vibration cycle or a vibration frequency of a vibrating electric signal, and outputs an electric signal or a numerical signal proportional to the wheel speed. Output. The output terminal of the wheel speed calculating circuit 35 includes a frequency calculating circuit 30 for calculating the frequency of the wheel speed signal using waveform data of a plurality of predetermined frequencies having different frequencies in a long cycle, and a frequency in a short cycle. An amplitude calculation circuit 32 for calculating the amplitude of the wheel speed signal with respect to the frequency calculated by the calculation circuit 30 is connected. The frequency calculation circuit 30 is connected to input the data of the calculated frequency to the amplitude calculation circuit 32.

The amplitude calculation circuit 32 is connected to a road μ gradient estimating device 14 for estimating a road μ gradient from the attenuation rate via an attenuation rate calculation circuit 34 for calculating an amplitude attenuation rate.

Further, the output of the amplitude calculation circuit 32 is input to the brake hydraulic control device 16 which outputs a control command to the hydraulic unit 18 when the vibration amplitude at the resonance frequency becomes equal to or less than a predetermined value. The output end of the brake hydraulic control device 16 is provided with the brake pedal 2 based on this operation amount.
AB for controlling the oil pressure of the wheel cylinder 20 (wheel cylinder oil pressure) by switching between the oil pressure of the master cylinder 26 corresponding to the pedal force and the high pressure source (not shown) irrelevant to the pedal force.
A hydraulic unit 18 having an S control valve is connected.
The ABS control valve includes a pressure increasing valve that increases the wheel cylinder oil pressure and a pressure reducing valve that reduces the wheel cylinder oil pressure.

If a pressure sensor 22 composed of a semiconductor pressure sensor for detecting wheel cylinder oil pressure can be attached to the wheel cylinder 20, the damping rate will be reduced.
The characteristics of the average wheel cylinder oil pressure can also be used for estimating the road surface condition. In this case, the output of the pressure sensor 22 is connected to the road surface μ gradient estimating device 14.

According to the present embodiment, the wheel speed signal output from the wheel speed calculating circuit 35 and including the vibration component when the wheel cylinder oil pressure is excited by the impulse at a predetermined cycle is calculated by the frequency calculating circuit 30 and the Amplitude calculation circuit 32
And the frequency and amplitude of the wheel speed signal are calculated as follows. In the following, the sampling cycle of measurement is 1 (ms), and the cycle of impulse excitation is 512
An example in which the sampling point is 0.512 (s) will be described.

First, the ABS control valve is controlled by the brake oil pressure control device 16 to excite the wheel cylinder oil pressure in a predetermined cycle. This impulse excitation is performed only when the amplitude of the wheel speed signal calculated as described below is equal to or less than a predetermined value. FIG. 2 shows an excitation cycle of 0.512 with an average wheel cylinder oil pressure command of 4 (MPa).
FIG. 7 (s) shows changes in wheel cylinder oil pressure and wheel speed when the impulse is excited.

FIGS. 3 and 4 show changes in the wheel speed vibration component and the average vibration component when the impulse excitation is performed with the average wheel cylinder oil pressure command being 2, 4, 6, 8, 10, 12 (MPa).

The calculation of the resonance frequency of the wheel speed signal by the frequency calculation circuit 30 will be described. The resonance frequency calculates a correlation between a plurality of sine waves and cosine waves having different predetermined frequencies and the time-series data of the wheel speed signal, and a sine wave or a sine wave having the highest correlation with the time-series data of the wheel speed signal. Select and find the cosine wave.

First, the correlation between the time-series data having an excitation cycle of 0.512 (s) and the sine and cosine waves having wave numbers of 6, 7, 8, 9, and 10 is calculated. Sinusoidal C _j ^m and the cosine wave S _j ^m of wave number m is expressed as follows. Note that the sine wave and the cosine wave may be represented by using the frequency instead of the wave number.

[0021]

C _j ^m = (1/256) cos (2πm (2j + 1) / 1024) (1) S _j ^m = (1/256) sin (2πm (2j + 1) / 1024) (2) where j = 0, 1, 2,..., 511, the time series data at the time point i is represented by x _n (n = i−511, i−
510, ..., when i), the real part R _i ^m and an imaginary part I _i ^m of the correlation coefficient at the i point is given by the following equation.

[0022]

(Equation 2) Here, n ′ = n mod 512.

From the above, the sum of squares of the correlation coefficient (square of the amplitude) is obtained by the following equation.

[0024] _{^{P i m = (R i m}} ) 2 + (I i m) 2 ··· (5) above (3), (4) can be modified as follows.

[0025]

(Equation 3)

As understood from the above equations (6) and (7), the sum of squares of the correlation coefficient is obtained by multiplying the second term by the first
Since it can be obtained by adding the term and the second term,
The number of operations can be greatly reduced.

The sum of squares P _{i of} the correlation coefficients detected at the time point _i
^m (m = 6, 7, 8, 9, 10) the largest sum of squares P
Select _i ^m, determines the resonance frequency f _m of the wheel speed signal according to the following equation from the wave number of the largest square sum P _i ^m.

[0028] _{f m = m / 0.512 ··· (} 8) For example, if the P _i ^m is the maximum in the case of m = _8, f 8 =
Assuming that 8 / 0.512 = 15.625 (Hz), the resonance frequency can be determined.

As described above, 512 for one wave number
To calculate the correlation coefficient of the time series data at the sample points,
The resonance frequency is calculated in a long cycle.

[0030] The resonance frequency ^{_{is, P i m (m = 6}} ,
7, 8, 9, 10) can be determined by the following equation.

[0031]

(Equation 4) Next, at the resonance frequency of the wheel speed signal detected at a long cycle as described above, the amplitude calculation circuit 32 calculates the amplitude of the wheel speed signal at a short cycle. This amplitude is calculated from the correlation coefficient between the sine wave and cosine wave of the frequency calculated by the frequency calculation circuit.

With respect to the time-series data having an excitation cycle of 0.512 (s), the frequencies having wave numbers of 6, 7, 8, 9, and 10 correspond to the sample points having one cycle of 85, 73, 64, 57, and 51, respectively. Corresponding to

[0033] Similar to the frequency of detection of a long period as described above, it represents the sine wave c _j ^m and cosine wave s _j ^m of wave number m as follows.

[0034]

C _j ^m = (2 / N _m ) cos (π (2j + 1) / N _m ) (9) s _j ^m = (2 / N _m ) sin (π (2j + 1) / N _m ) (10) where j = 0, 1, 2,..., N _m −1, N _m = 8
5, 73, 64, 57, 51.

As described above, since the correlation coefficient of N _m sample points is calculated for one wave number, the amplitude is calculated in a short cycle.

Then, the real part R _i of the correlation coefficient at the time point _i
calculating by the following formula ^m and an imaginary part I _i ^m.

[0037]

(Equation 6) Here, n ′ = n mod N _m .

From the above, the amplitude is obtained by the following equation.

[0039] ^{_{a i m = {(r i}} m) 2 + (i i m) 2} 2 ··· (13) (11), (12) can be modified as follows.

[0040]

(Equation 7) Here, i ′ = (iN _m ) mod N _m = i mod N _m .

By modifying the equation as described above, it is possible to greatly reduce the number of calculations as described in the case where the frequency is calculated in a long cycle.

In the attenuation factor calculation circuit 34, the amplitude one cycle before
By comparing with the current amplitude, the amplitude increases
Judge whether the amplitude is increasing or not.
Chia _i ^m> A_i-Nm ^mIn the case of wheels-suspension
Since the vibration continues for a relatively long time,
Judging that vibration was applied by the rake hydraulic control,
Ignore decay.

On the other hand, when the amplitude is decreasing,
Since the suspension vibration converges relatively quickly, the damping rate k
Calculate _i ^m as follows.

The damping properties including amplitude ^{_{a i m, a i-1}} m, the ··· a _i-Nm ^m, as the curve shown in FIG. 6 (A), can be approximated by an exponential function of the following .

[0045] ^{_{y = exp (-k i m t}} ) sin (ωt + φ) envelope Y of the exponential function represents the amplitude variation is expressed by the following equation.

[0046] Y = exp Taking (-k _i ^m t) logarithm of the above expression, logY = -k _i ^m t, and the
As shown in FIG. 6 (B), by determining the slope for the t of logy, it is possible to obtain an attenuation factor k _i ^m.

The road surface μ estimating device 14 determines the road surface condition based on the attenuation factor calculated by the attenuation factor calculation circuit 34. When the damping rate is large, that is, when the vibration is rapidly damped, the road surface is slippery, that is, it is determined that the road surface μ is approaching the limit value because the road surface μ gradient is small or the road surface μ gradient is small. .

Conversely, when the damping rate is small, that is, when the vibration continues for a long time, the road surface is hard to slip, that is, the road surface μ gradient is large, or the road surface μ gradient is large, so Judge that there is enough time.

Further, if the pressure sensor 20 can be mounted, the average wheel cylinder oil pressure is calculated from the measured wheel cylinder oil pressure, and the relationship between the average wheel cylinder oil pressure and the damping rate calculated as described above is shown in FIG. A table showing the damping characteristics of suspension vibration can be created. By comparing this table with the current wheel cylinder oil pressure and damping rate, it is possible to determine whether the traveling road surface is a low μ road or a high μ road. Can be determined.

In the above description, the example in which the wheel cylinder oil pressure is excited by the ABS control valve has been described.
Alternatively, an impulse-like fluctuation may be generated in the wheel cylinder oil pressure by using a VSC actuator to cause wheel-suspension excitation.

Further, only the vibration component is extracted by removing the average deceleration from the wheel speed signal, and the waveform is cut out at each cycle of the excitation, and the waveform is averaged in a speed range (for example, every 10 (rad / s)). The amplitude may be extracted by calculating the correlation with the sine wave and cosine wave described above.

As described above, according to the present embodiment, since the brake hydraulic pressure is not always excited but is excited only when the amplitude of the vibration is reduced, the noise of the actuator can be reduced. At the same time, the durability of the actuator can be improved. Further, since the target frequency is detected from a predetermined band, it is possible to cope with a case where the vibration characteristics of the wheel-suspension system are unknown or the vibration characteristics change due to tire replacement or the like. The road μ gradient can be estimated regardless of whether or not the brake is operated.

It is to be noted that, other than on a smooth road, a wheel-suspension excitation can be caused by disturbance.
The wheel cylinder oil pressure exciting means described in the above embodiment may be omitted.

Further, in the present embodiment, the description has been made of the wheel-suspension vibration (resonance frequency of about 20 Hz) as an example.
(near z).

[0055]

As described above, according to the present invention,
Since a plurality of detected frequencies are used when estimating the road surface state, for example, the μ gradient of the road surface, an effect of providing a road surface state estimation device that is robust against vibration frequency changes due to tire changes or the like can be provided. Is obtained.

[Brief description of the drawings]

FIG. 1 is a block diagram of an embodiment of the present invention.

FIG. 2 shows an average wheel cylinder oil pressure command of 4 (MPa).
FIG. 7 is a diagram showing changes in wheel cylinder oil pressure and wheel speed when impulse excitation is performed.

FIG. 3 shows an average wheel cylinder oil pressure command of 2, 4, 6,
FIG. 8 is a diagram illustrating changes in a wheel speed vibration component and an average vibration component when an impulse excitation is performed at 8 (MPa).

FIG. 4 shows an average wheel cylinder oil pressure command of 10, 12
FIG. 9 is a diagram illustrating changes in a wheel speed vibration component and an average vibration component when an impulse excitation is performed as (MPa).

FIG. 5 is a diagram showing a table showing a damping characteristic between a damping rate and suspension vibration.

FIG. 6 is a diagram for explaining a method of calculating an amplitude using an exponential function.

[Explanation of symbols]

 Reference Signs List 10 Wheel speed sensor 14 Road surface μ gradient estimating device 18 Hydraulic unit 34 Damping rate calculation circuit

 ──────────────────────────────────────────────────続 き Continuation of the front page (71) Applicant 000004260 DENSO Corporation 1-1, Showa-cho, Kariya-shi, Aichi Prefecture (72) Inventor Ken Sugai 41-Cho, Yukumichi, Okumachi, Nagakute-cho, Aichi-gun, Aichi Prefecture, Ltd. Inside Toyota Central Research Laboratory (72) Inventor Eiichi Ono 41-cho, Chuchu-Yokomichi, Nagakute-cho, Aichi-gun, Aichi Prefecture 1 in Toyota Central Research Laboratory Co., Ltd. 41 No. 1 Inside Toyota Central Research Laboratory Co., Ltd. (72) Inventor Masanori Miyashita 41-Chome Toyoda Central Research Laboratories Co., Ltd. (72) Inventor Makoto Yamamoto Toyota City, Aichi Prefecture 1st Toyota Town Inside Toyota Motor Co., Ltd. (72) Inventor Shoji Ito 1st Toyota Town Toyota City, Aichi Prefecture Toyota Motor (72) Inventor Yoshiyuki Yasui 2-1-1 Asahi-cho, Kariya-shi, Aichi Aisin Seiki Co., Ltd. (72) Inventor Satoshi 2-1-1 Asahi-cho, Kariya-shi, Aichi Aisin Seiki Co., Ltd. (72) Inventor Shusaku Fujimoto 1-1-1 Showa-cho, Kariya-shi, Aichi F-term in DENSO Corporation (reference) 3D046 BB21 BB23 BB28 BB29 HH16 HH36 HH46 KK06

Claims (2)

[Claims] A wheel speed detecting means for detecting a wheel speed and outputting a wheel speed signal; and time-series data of the wheel speed signal in a state where the wheel speed signal includes a vibration component in a first cycle. A resonance frequency calculating means for calculating a correlation with waveform data of a plurality of predetermined frequencies having different frequencies and calculating a resonance frequency of a wheel speed signal from the waveform data having a high correlation, At the resonance frequency, the time series data of the wheel speed signal and the vibration obtained by the resonance frequency calculation in a state where the wheel speed signal includes the vibration component in the second cycle shorter than the first cycle. Amplitude calculation means for calculating the correlation coefficient with the number of waveform data, calculating the amplitude of the wheel speed signal from the obtained correlation coefficient, and the degree of attenuation of the amplitude of the wheel speed signal calculated by the amplitude calculation means Road based Road surface condition estimating apparatus comprising a road surface state estimating means for estimating a state, the. 2. A road surface state estimating apparatus further comprising a wheel cylinder oil pressure exciting means for exciting wheel cylinder oil pressure when the amplitude of the wheel speed signal is small.