JP2010241342A - Road surface condition determining device and road surface condition determination method - Google Patents

Abstract

<P>PROBLEM TO BE SOLVED: To suppress the load of processing required for determination, by reducing erroneous determination due to the irregularity of the road surface of a partial road, and determining the road surface conditions of a spot where the road surface conditions are switched, with accuracy. <P>SOLUTION: Similarity in the power spectrum of vehicle acceleration detected by an acceleration sensor and a reference power spectrum expressing the features of the road surface conditions is calculated, and the highest similarity road conditions whose similarity is the highest are selected from among the similarity; and the road surface situations are determined, based on the highest similarity road conditions and the previously determined road surface conditions (determination result). <P>COPYRIGHT: (C)2011,JPO&INPIT

Description

  The present invention relates to a road surface state determination device and a road surface state determination method for determining a road surface state.

Conventionally, as a method of determining the road surface condition, there is a method of determining the road surface condition from an image of a camera installed on the road. In Patent Document 1, an image feature amount is calculated from a road surface image photographed by a TV camera installed for existing road monitoring, and a road surface state is determined.
In addition, a method for determining a road surface condition using an acceleration sensor or a microphone mounted on a vehicle has been proposed. Patent Document 2 describes an apparatus and a road surface determination system that determine a road surface state using an acceleration sensor and a sound pressure sensor arranged inside a tire. In this invention, the respective waveforms and power spectra are acquired from the acceleration sensor and the sound pressure sensor, the comparison result of the acceleration waveform and the acceleration reference waveform, the comparison result of the acceleration power spectrum and the acceleration reference power spectrum, the sound pressure The road surface condition is determined by the majority of the four comparison results of the comparison result of the waveform and the reference waveform of the sound pressure and the comparison result of the power spectrum of the sound pressure and the reference power spectrum of the sound pressure.

JP 2004-274431 A JP 2002-340863 A

In Patent Document 1, although road surface conditions are determined using an existing road monitoring camera, no camera is currently installed on all roads. For this reason, it is necessary to newly install a camera in order to obtain the road surface conditions of all roads. On the other hand, the method of determining the road surface condition using the sensor mounted on the vehicle of Patent Document 2 does not require maintenance of infrastructure such as a camera.
According to the comparison between the acceleration and sound pressure waveforms and the power spectrum and the respective reference waveforms in Patent Document 2, the dry road, bad road, snow road, and ice road depending on the magnitude of the variance between the observed waveform and the reference waveform. Judging hydroplaning. In order to prevent misjudgment that occurs when driving on a partially rough spot on the road surface, the accumulated dispersion value so far is used for judgment. However, since the determination is based on the accumulation so far, there is a problem that even if the road surface condition is suddenly switched, it is not immediately determined to be the correct road surface condition. For example, when the running road surface condition changes from a dry road to a bad road, the determination immediately after the change takes into account the road surface condition (dry road) so far, and thus is not immediately determined as a bad road.

As described above, the conventional road surface condition determination has not been considered for accurately determining the point where the road surface condition is switched.
The present invention has been made in order to solve the above-described problems, and reduces misjudgment due to unevenness on the road surface of a partial road, etc., and determines the road surface condition at a point where the road surface condition is switched. The purpose is to perform processing with high accuracy and to reduce the processing load required for determination.

The road surface condition determination device according to the present invention includes a spectrum calculation means for calculating a power spectrum of the vehicle acceleration based on the vehicle acceleration detected by the acceleration sensor,
Reference spectrum storage means for storing a reference power spectrum representing the characteristics of each road surface condition;
Spectrum similarity calculation means for calculating the similarity between the power spectrum calculated by the spectrum calculation means and the reference spectrum stored by the reference spectrum storage means;
Spectral similarity comparison means for selecting a high similarity road surface condition having the highest similarity from the similarities calculated by the spectral similarity calculation means;
A comparison result history storage means in which a plurality of high similarity road surface conditions selected so far by the spectrum similarity comparison means are stored;
A determination result history storage means for storing the road surface condition determined immediately before;
Road surface condition determining means for determining a road surface condition based on the plurality of high similarity road surface conditions stored in the comparison result history storage means and the road surface condition stored in the determination result history storage means. Is.

  The present invention obtains the similarity between the power spectrum of the vehicle acceleration detected by the acceleration sensor and the reference power spectrum representing the characteristics of the road surface condition, and the road surface condition having the highest similarity among the similarities is obtained. By selecting and determining the road surface condition based on this high similarity road surface condition and the road surface condition (determination result) determined immediately before, the road surface condition can be determined with high accuracy even if the road surface condition changes. Can do.

1 is a configuration diagram of a road surface condition determination device according to a first embodiment. 3 is a flowchart showing an operation procedure from acquisition of acceleration to road surface condition determination in the first embodiment. 5 is a flowchart showing an operation procedure for calculating and comparing a similarity with a reference spectrum according to the first embodiment. Explanatory drawing when the similarity corresponding to the road surface condition is memorize | stored in the comparison result log | history memory means 10 in Embodiment 1. FIG. 6 is a flowchart showing an operation procedure when the road surface condition determination unit 9 determines the road surface condition in the first embodiment. 5 is a flowchart showing an operation procedure when the determination interval adjustment unit 12 sets a value of k in the first embodiment. FIG. 4 is an explanatory diagram of a determination example that is determined without executing step S6 of FIG. 2 in the first embodiment. Explanatory drawing of the example of determination determined by performing step S6 of FIG. 2 in Embodiment 1. FIG.

Embodiment 1 FIG.
FIG. 1 is a configuration diagram of a road surface condition determination apparatus according to Embodiment 1 of the present invention.
Broadly speaking, the road surface condition determining apparatus is composed of an acceleration sensor 1 that detects the acceleration of the vehicle and a road surface condition determining apparatus 2 that determines the road surface condition based on the acceleration acquired from the acceleration sensor 1.
The acceleration acquisition means 3 in the road surface condition determination device 2 acquires acceleration data detected by the acceleration sensor 1 and stores it in the acceleration storage means 4.
The spectrum calculation means 5 calculates the power spectrum of the acquired acceleration based on the acceleration data stored in the acceleration storage means 4.
The reference spectrum storage means 7 stores in advance a power spectrum (reference spectrum) representing the characteristics of each road surface condition.

The spectrum similarity calculation means 6 calculates the similarity between the power spectrum calculated by the spectrum calculation means 5 and the reference spectrum of each road surface condition stored in the reference spectrum storage means 7.
The spectrum similarity comparison unit 8 selects the road surface condition (high similarity road surface condition) of the reference spectrum having the highest similarity from among the similarities calculated by the spectrum similarity calculation unit 6, and has the highest similarity. The road surface condition of the reference spectrum and the similarity value at that time are stored in the comparison result history storage means 10.
The comparison result history storage means 10 stores the road surface condition selected by the spectrum similarity comparison means 8 and its similarity as a comparison result history of the similarity to the reference spectrum.
The determination result history storage unit 11 stores the road surface state determined immediately before by the road surface state determination unit 9.

In the road surface condition judging means 9, the comparison result history of the similarity degree stored in the comparison result history storage means 10, that is, the road surface situation and the similarity degree so far selected by the spectrum similarity degree comparing means 8, and the judgment result history The road surface condition is determined based on the determination result of the road surface condition determined immediately before stored in the storage unit 11.
In the determination by the road surface condition determination means 9, since the comparison result history of the similarity with the reference spectrum so far is also used, even if the road surface condition is switched, a delay occurs until the road surface condition is determined to be correct. In order to shorten this delay, the determination interval adjustment unit 12 adjusts the interval until the next determination based on the comparison result of the similarity stored in the comparison result history storage unit 10. Specifically, the acceleration acquisition number of the acceleration acquisition means 3 is adjusted so that the determination interval is shortened when the road surface condition is estimated to be changed, and the determination interval is increased when it is estimated that there is no change in the road surface condition. Do

Next, the operation will be described.
FIG. 2 is a flowchart showing an operation procedure from acquiring acceleration from the acceleration sensor 1 to determining the road surface state by the road surface state determination device 2.
First, the acceleration acquisition means 3 acquires vehicle vibration (acceleration) from the acceleration sensor 1 at regular intervals and stores it in the acceleration storage means 4 (step S1). The operation of step S1 is repeated until the acceleration acquisition means 3 acquires k accelerations (step S2).
When the acceleration acquisition means 3 acquires k accelerations and stores them in the acceleration storage means 4, the spectrum calculation means 5 calculates a power spectrum using the k accelerations stored in the acceleration storage means 4 ( Step S3). For the calculation of the power spectrum, Fourier transform or the like is used.

  The spectrum similarity calculation means 6 calculates the similarity between the power spectrum calculated in step S3 and the reference spectrum of each road surface condition stored in advance in the reference spectrum storage means 7. Then, the spectrum similarity comparison means 8 compares the values of the similarities calculated by the spectrum similarity calculation means 6, selects the road surface condition having the largest value (the highest similarity), and compares the result history storage means. 10 stores the road surface condition and the similarity (step S4). For calculating the similarity to the reference spectrum, a correlation coefficient, a variance value, or the like is used.

Subsequently, based on the comparison result history of the similarity to the reference spectrum of each road surface condition stored in the comparison result history storage unit 10 and the previous determination result stored in the determination result history storage unit 11. The road surface condition determining means 9 determines the current road surface condition (step S5).
Finally, the determination interval adjustment unit 12 sets the number of accelerations acquired (k value) in the next determination (step S6), and returns to the next determination procedure (step S1).
The above is a large flow of road surface condition determination in the road surface condition determination device 2.

Next, detailed operations of step S4, step S5, and step S6 of FIG. 2 will be described.
FIG. 3 is a flowchart showing an operation procedure for calculating and comparing the similarity with the reference spectrum in step S4.
First, the spectrum similarity calculation means 6 calculates the similarity between the acceleration power spectrum calculated by the spectrum calculation means 5 in step S3 and the reference spectrum of each road surface condition stored in the reference spectrum storage means 7 (step S3). S41).
Subsequently, the spectrum similarity comparison means 8 compares the value of similarity with the reference spectrum of each road surface condition calculated in step S41, and selects the road surface condition with the highest similarity (step S42). Further, the road surface condition selected by the spectrum similarity comparison unit 8 in step S42 and the similarity value at that time are recorded in the comparison result history storage unit 10 (step S43).

Here, the operation of step S4 in FIG. 2 will be specifically described.
FIG. 4 is an explanatory diagram when the spectrum similarity comparison unit 8 stores the similarity corresponding to the road surface condition in the comparison result history storage unit 10 in step S4.
The reference spectrum storage means 7 in FIG. 4 is the same as the reference spectrum storage means 7 shown in FIG. 1, and the reference spectrum storage means 7 has reference spectra 71, 72, 73 for road surface conditions A, B, C. Stored.
Thus, it is assumed that the reference spectrum of the road surface condition to be determined is stored in the reference spectrum storage means 7 in advance. A plurality of reference spectra may be stored for one road surface condition.

In step S41, the spectrum similarity calculation means 6 calculates the similarity 71 ′, 72 ′, 73 ′ between the power spectrum calculated in step S3 and the reference spectrum of each road surface situation A, B, C. Here, it is assumed that the similarity of road surface conditions A, B, and C is similarity A = 0.7, similarity B = 0.3, and similarity C = 0.5, respectively.
The spectrum similarity comparing means 8 compares the similarity values in step S42, selects the road surface condition A having the highest similarity, and in step S43, compares the road surface condition A and the similarity value in the comparison result history storage means 10. Remember 0.7.

Next, the operation of step S5 in FIG. 2 will be specifically described.
FIG. 5 is a flowchart showing an operation procedure when the road surface condition determining means 9 determines the road surface condition in step S5.
The determination of the road surface condition is stored in the N road surface conditions selected when the road surface condition determination unit 9 has executed the past N times stored in the comparison result history storage unit 10 and the determination result history storage unit 11. This is performed using the road surface condition determined immediately before.
For explanation, it is assumed that the road surface condition selected in step S4 in FIG. 2 at the current determination is X, and the road surface condition determined in step S5 in FIG. 2 at the previous determination is Y (step S51).

First, the road surface condition determining means 9 checks whether or not the road surface conditions of the comparison results of the past N spectrum similarities stored in the comparison result history storage means 10 are all the same road surface condition X (step S52). . If all the road surface conditions of the past N comparison results of the spectrum similarity are X (step S52 is YES), the road surface state determination means 9 determines the current road surface state to be X (step S53). Further, the road surface condition determination unit 9 stores the road surface condition X in the determination result history storage unit 11 (step S54).
In step S52, when the road surface condition that is different from X is included in the road surface conditions of the past N comparisons of the degree of similarity of the spectrum (when step S52 is NO), the road surface condition determination means 9 determines the previous road surface condition. The road surface condition determined by the determination result at the determination step S5 is determined as Y (step S55), and the road surface condition Y is stored in the determination result history storage unit 11 (step S56).

  As described above, by determining the road surface condition using the history of the comparison results of the spectral similarity and the immediately preceding determination result, it is possible to eliminate erroneous determination that a different road surface condition is determined only at a certain determination. Can do. For example, the spectrum similarity comparison unit 8 selects the road surface condition A as a high similarity road surface condition in step S4 in a certain determination, but selects the road surface condition B in the determination step S4 performed before and after that. In such a case, there is a situation in which a different road surface condition A is determined only at a certain time while it is continuously determined as the road surface condition B in the determination that the history of past comparison results is not used. However, in the present invention, even if the spectrum similarity comparison means 8 selects a road surface situation A different from the road surface situation B selected so far only in a certain step S4, the road surface situation judgment means 9 in step S5. Is determined to be road surface condition B from the comparison result of the previous degrees of similarity and the previous determination result. By making such a determination, it is possible to prevent a change in road surface condition in a short time due to partial road surface unevenness or the like.

However, even when there is a partial change in the road surface condition, a delay occurs until the determination is switched to the next road surface condition. When it is estimated that the road surface condition has changed in order to reduce this delay, the number of accelerations (value of k) acquired by the acceleration acquisition unit 3 in the determination operation of the next road surface state by the determination interval adjustment unit 12. Is reduced to shorten the determination interval (step S6). As a specific example, the acceleration acquisition means 3 halves the number of accelerations acquired in steps S1 and S2 in FIG. 2, thereby reducing the time from the acceleration acquisition in step S1 in FIG. 2 to the power spectrum calculation in step S3. Cut in half. In addition to changing the number of accelerations, the determination interval can also be shortened by shortening the time interval from when the acceleration acquisition means 3 acquires acceleration from the acceleration sensor 1 at step S1 until the acceleration is acquired again at step S1. can do.
Thus, by shortening the determination interval, the time until the next road surface condition is determined is shortened, and the time until the determination is switched can be shortened.

Next, the operation of step S6 in FIG. 2 will be specifically described.
FIG. 6 is a flowchart showing an operation procedure when the determination interval adjusting means 12 sets the value of k in step S6.
The value of k is set using a history of comparison results of similarity with the reference spectrum so far stored in the comparison result history storage means 10. The road surface condition selected in step S4 in this determination operation and the similarity at that time are X and X ', respectively, and the road surface condition selected most frequently in the history of the previous N comparison results is Z. (Step S61).

  The value of k is (1) the road surface condition Z that has been selected most frequently from the history of comparison results, and the road surface condition X that is most similar in step S4 is the same (step S62), (2) road surface condition When any of the conditions that the similarity value X ′ to X is equal to or greater than a threshold value α (step S63) is not satisfied, the value of k is decreased, that is, the determination interval until the road surface condition is determined. Decrease (step S65). Further, in order to increase the determination interval when the road surface condition is stable, when both conditions (1) and (2) are satisfied, the value of k is increased, that is, until the road surface condition is determined. The interval is increased (step S64).

The condition for reducing the value of k is not limited to (1) and (2), and may be any case where it is estimated that the currently determined road surface condition is not stable. In addition, an upper limit value and a lower limit value can be set for the value of k so that the determination interval is not too long (too short). For example, if the value that can be set to k is set in two stages, and the condition (1) or (2) is not satisfied and the road surface condition is estimated to be unstable, the value of k is set to the smaller value, If both (1) and (2) are established, the value of k may be set larger.
In this embodiment, the determination interval is adjusted by adjusting the value of the number of accelerations k necessary for calculating the spectrum. However, instead of the value of k, the time for acquiring the acceleration from the acceleration sensor 1 is adjusted. The determination interval may be adjusted by adjusting.

7 and 8 are explanatory diagrams of determination examples determined according to the present embodiment.
FIG. 7 is a determination example when step S6 of FIG. 2 is not executed and the determination is always performed at the same interval with the value of k being fixed (determination A). FIG. 8 is the same road surface condition as FIG. A determination example when the determination is performed in step S6 of FIG. 2 while adjusting the determination interval (determination B) is shown. The same number is given to the same thing in both figures.
In FIG. 7 and FIG. 8, 100 represents the actual road surface condition, and the vehicle travels from left to right. 101A and 101B are points where the road surface state is switched from A to B and from B to C, respectively. Reference numeral 102 denotes a timing at which the acceleration acquisition unit 3 acquires acceleration from the acceleration sensor 1 in step S1 of FIG.

  Reference numeral 103 in FIG. 7 represents a section in which the road surface condition is determined in the determination A, and the acceleration obtained in each section is used in the calculation of the spectrum in step S3 in FIG. 104 is the similarity between the acquired acceleration spectrum and the reference spectra of road surface conditions A, B, and C calculated in step S41 of FIG. Reference numeral 105 denotes a determination result in the determination A of each section. 106A and 106B indicate delays until the road surface condition is actually determined after the road surface state is switched at 101A and 101B.

In FIG. 8, 203 is a section in which the road surface condition is determined in the determination B, 204 is a similarity between the acceleration spectrum in each section and the reference spectrum of the road surface conditions A, B, and C, and 205 is a determination in the determination B in each section. As a result, 206A and 206B represent delays until the road surface state is determined to be correct after the road surface state is switched.
In the determinations of FIGS. 7 and 8, if N = 3 and the determination that the similarity of a certain road surface state X is the highest is continued three times, the road surface state of that section is determined to be X.
By using the history of comparison results, even if the road surface condition is switched, it is not immediately determined to be the next road surface condition. Therefore, until the road surface condition is actually determined to be the next road surface condition after switching at 101A and 101B. Delay (106A, 106B, 206A, 206B) occurs.

Therefore, when it is detected based on the similarity value with the reference spectrum of each road surface state 204 that the road surface state has been switched between 101A and 101B as in the case of determination B in FIG. 8, the determination interval is shortened. The delay until the next road surface state is determined is shortened. In determination B, if the value of α in step S63 in FIG. 6 is 0.8 and the similarity value of the road surface condition having the highest similarity selected in step S4 in FIG. Is shortened.
By shortening the determination interval, a delay until the road surface condition is determined after the road surface state is switched is shortened. Accordingly, the delays 206A and 206B in the determination B in FIG. 8 are shorter than the delays 106A and 106B in the determination A in FIG. 7, and the road surface state switching points (101A and 101B) can be determined with high accuracy. it can.

As described above, according to the present embodiment, the similarity between the power spectrum of the vehicle acceleration detected by the acceleration sensor and the reference power spectrum representing the characteristics of the road surface condition is obtained, and the similarity is determined from the similarities. Even if the road surface condition changes, by selecting the highest high similarity road surface condition and determining the road surface condition based on this high similarity road surface condition and the road surface condition (determination result) determined immediately before, The road surface condition can be determined with high accuracy.
In addition, the judgment interval can be shortened at the point where the road surface condition is changed, and the change of the road surface condition can be detected quickly, so even if the road surface condition changes, the determination to immediately determine the correct road surface condition Can be switched, and the delay until switching of determination can be shortened.
Furthermore, since the determination interval becomes longer at a point where the road surface condition is stable, the number of determinations performed by the apparatus can be reduced, so that the processing load such as the CPU load can be reduced.

  DESCRIPTION OF SYMBOLS 1 Acceleration sensor, 2 Road surface condition determination apparatus, 3 Acceleration acquisition means, 4 Acceleration memory means, 5 Spectrum calculation means, 6 Spectrum similarity calculation means, 7 Reference spectrum storage means, 8 Spectrum similarity comparison means, 9 Road surface condition determination means 10 comparison result history storage means, 11 determination result history storage means, 12 determination interval adjustment means.

Claims (8)

Spectrum calculation means for calculating a power spectrum of the vehicle acceleration based on the vehicle acceleration detected by the acceleration sensor;
Reference spectrum storage means for storing a reference power spectrum representing the characteristics of each road surface condition;
Spectrum similarity calculation means for calculating the similarity between the power spectrum calculated by the spectrum calculation means and the reference spectrum stored by the reference spectrum storage means;
Spectral similarity comparison means for selecting a high similarity road surface condition having the highest similarity from the similarities calculated by the spectral similarity calculation means;
A comparison result history storage means in which a plurality of high similarity road surface conditions selected so far by the spectrum similarity comparison means are stored;
A determination result history storage means for storing the road surface condition determined immediately before;
Road surface condition determining means for determining a road surface condition based on the plurality of high similarity road surface conditions stored in the comparison result history storage means and the road surface condition stored by the determination result history storage means. A road surface condition judging device characterized by that. Determination interval adjustment for adjusting an interval until the road surface condition determination unit next determines the road surface condition based on the plurality of high similarity road surface conditions and the plurality of similarities stored in the comparison result history storage unit With means,
2. The road surface condition determining apparatus according to claim 1, wherein the road surface condition determining unit determines a road surface condition based on an adjustment result by the determination interval adjusting unit. An acceleration acquisition means for acquiring a predetermined number of vehicle accelerations detected by the acceleration sensor;
The road surface condition determination apparatus according to claim 1 or 2, wherein the spectrum calculation means calculates a power spectrum of a predetermined number of vehicle accelerations acquired by the acceleration acquisition means. The determination interval adjusting means sets the number of acquisitions of vehicle acceleration when the road surface condition determining means determines the road surface condition next time,
The road surface condition determination apparatus according to claim 3, wherein the acceleration acquisition means acquires the acquired number of vehicle accelerations set by the determination interval adjustment means. The road surface condition judging means is selected by the spectrum similarity comparing means when the plurality of high similarity road surface conditions stored in the comparison result history storage means are all the same as the road surface condition at the time of the current judgment. When it is determined that the similarity road surface condition is a road surface condition, and the road surface condition different from the road surface condition at the time of the current determination is included in the plurality of high similarity road surface conditions, the road surface condition determined this time is the road surface condition The road surface condition determination apparatus according to claim 1 or 2, wherein the road surface condition determination apparatus according to claim 1 or 2 is determined. The determination interval adjustment means, when the high similarity road surface condition selected by the spectrum similarity comparison means is the same as the road surface condition most frequently selected immediately before from the comparison result history storage means, or The determination interval until the road surface condition determining unit determines the road surface condition when the similarity value of the high similarity road surface condition selected by the spectrum similarity comparing unit is not any of the predetermined threshold value or more. The road surface condition judging device according to claim 2, wherein the road surface condition judging device is adjusted to be small. The determination interval adjustment means has the same high-similarity road surface condition selected by the spectrum similarity comparison means and the road surface condition most frequently selected immediately before from the comparison result history storage means, and When the similarity value of the high similarity road surface state selected by the spectrum similarity comparison unit is equal to or greater than a predetermined threshold, the determination interval until the road surface state determination unit determines the road surface state is greatly adjusted. The road surface condition determination device according to claim 2, wherein A power spectrum calculating step for calculating a power spectrum of the vehicle acceleration based on the vehicle acceleration detected by the acceleration sensor;
A spectrum similarity calculating step for calculating a similarity between the power spectrum calculated by the power spectrum calculating step and a reference power spectrum representing characteristics of each road surface condition stored in advance;
A high similarity road surface condition selecting step for selecting a high similarity road surface condition having the highest similarity from the similarities calculated by the spectrum similarity calculation step;
A comparison result history storage step for storing a plurality of high similarity road surface conditions selected so far by the high similarity road surface condition selection step;
A determination result history storage step for storing the road surface condition determined immediately before;
A road surface condition determination step for determining a road surface condition based on the plurality of high similarity road surface conditions stored in the comparison result history storage step and the road surface condition stored in the determination result history storage step. A road surface condition determination method characterized by that.

Images