JP6132304B2 - Road surface evaluation apparatus and method - Google Patents

Description

The present invention relates to a road surface evaluation apparatus and method.

In Japan today, we are entering an era of maintenance and management of road infrastructure that has been developed in large quantities during the period of high economic growth. Road maintenance that is damaged due to aging causes vibration, noise, traffic accidents, etc., so road maintenance is important to prevent such damage. However, considering the maintenance of a large amount of road infrastructure in the future, severe cost constraints are expected. Therefore, it is required to evaluate the road surface condition cheaply and efficiently. The following two methods are mainly used in the current road surface inspection. The first is a visual inspection method by the investigator, but this is a qualitative judgment and not an objective inspection. The second method is to measure road surface irregularities with an accelerometer or laser using a high-performance inspection vehicle. With this method, high-precision and quantitative measurement can be performed, but high-performance inspection vehicles are very expensive and data analysis takes time, so high-frequency measurement cannot be performed. These techniques are not efficient techniques.

Our laboratory is working on the development of a simple road surface diagnosis system called Vehicle Intelligent Monitoring System (VIMS). VIMS is a system that runs on the road with an accelerometer and GPS installed in the vehicle, and evaluates the flatness of the road from its vertical acceleration response. As an index for road surface evaluation, the International Roughness Index (IRI) proposed by the World Bank in 1986 is used (Non-Patent Document 1).

Furukawa's research has devised a vehicle response measurement system that combines an accelerometer and GPS, and has developed a basic VIMS system, albeit with the restriction of constant speed running at 60km / h by a land cruiser (Non-patent Document 2). After that, Asakawa et al. Inter-vehicle calibration that can estimate the same IRI value even with different measurement vehicles and speeds by running on the road surface with known road profile data with multiple measurement vehicles and speeds and comparing the actual measurement data A method and speed calibration method were proposed to enable IRI estimation for various vehicle types and speeds (Non-Patent Documents 3 and 4).

However, in vehicle type / speed calibration, it is necessary to measure the road surface profile on a known road surface, or to travel on the same road surface with a plurality of specific vehicles, so there is no means for measuring the road surface profile and the vehicle response is known. In developing countries where it is difficult to bring in IRI, there was a problem that IRI could not be measured accurately. Therefore, Yano builds a vehicle response model (full car model) that shows the same behavior from the vertical behavior of the vehicle when crossing the hump by a hump test that can be performed anywhere, and estimates each parameter of the model with a genetic algorithm Proposed a new method for vehicle and speed calibration (Non-Patent Document 5). By this method, a model that can reproduce the frequency spectrum of the acceleration response of the actual vehicle was constructed.

Takahashi modeled the behavior of an actual vehicle during a hump test with a QC model that takes nonlinearity into account, and combined this with the speed calibration method proposed by Asakawa (Non-Patent Document 5), hybrid calibration that does not require a road surface profile. A method was proposed (Non-Patent Document 6). With this method, although there are exceptions for some models, we succeeded in simplifying the full car model and showed that IRI can be accurately estimated even in the high speed range of 80 km / h.

Based on Shimada's research, IRI estimation, which had previously been a condition of constant speed driving, was not necessary for constant speed driving if the speed fluctuation range was below a certain level using the inter-vehicle response ratio corresponding to the speed for each evaluation section. Highly accurate IRI estimation has become possible (Non-Patent Document 7). According to Shimada's research, response amplitude ratios are prepared in increments of 20 km / h in the speed range of 70 km / h to 110 km / h, and response amplitude ratios at arbitrary speeds can be created by linear interpolation.

The IRI estimated by VIMS is an index of road surface flatness, and it cannot discriminate visibility such as cracks, potholes, and road surface expansion / contraction devices (joints). Road maintenance is important not only for IRI but also for other indicators, so it is necessary to identify these visibility conditions. Another problem is that IRI cannot be estimated at locations where GPS cannot be used, such as tunnels, because the vehicle's running speed is measured using GPS. Therefore, Miwa aimed at identifying the vehicle speed and discriminating the visibility by performing image analysis (Non-patent Document 8). As a result, we succeeded in identifying the speed at a practical level for both constant speed and free running, and succeeded in detecting the road surface expansion device with high accuracy and detecting cracks on the road surface only in fine weather.

As described above, VIMS has been studied for many years, and IRI estimation under various vehicle and running speed conditions has become possible, and it has become possible to discriminate visibility by image analysis. However, VIMS requires an accelerometer, GPS, and a personal computer for recording, and includes devices that are not generally available. In order to increase the versatility of VIMS, improvements to a system that can be easily obtained and can be used on a daily basis are required. Considering that many VIMS will be operated for many years in the future, there may be cases where accurate measurement cannot be performed due to poor contact of acceleration wiring.

It is expected that the above-described problems can be solved if a device with a built-in sensor that is generally distributed and has few wiring defects, such as smartphones, can be used. Furthermore, in conventional VIMS, the synchronization between accelerometer recording and imaging equipment was complicated, but using a smartphone, it is easy to achieve synchronization, and IRI information obtained from acceleration measurement and image information can be combined and analyzed. It is done. At the end of 2011, an iPhone application that can measure and record 3-axis acceleration and angular velocity / position / velocity information was developed, and if it can be used, VIMS operation is expected to be much easier. In addition, it is considered that problems caused by equipment such as wiring mistakes and poor contact are greatly reduced. However, when replacing conventional accelerometers and GPS with smartphones, it is essential to consider measurement accuracy and installation location. In the measurement using a smartphone with an integrated operation panel and sensor, it will be necessary to install a smartphone near the driver or operator.

In the measurement using an accelerometer, the acceleration response varies depending on the location of the sensor in the vehicle. Therefore, it was necessary to strictly define the location and verify the validity of the location every time the location was changed. . For example, when the installation location is defined as “in-vehicle position on the rear wheel axle”, there is no place suitable for sensor installation depending on the vehicle type, and it is necessary to examine the installation location for each vehicle type. This is not a topic limited to the problem of using a smartphone for IRI estimation, but is a problem common to measurements using an accelerometer.

Although there are some patent documents related to a road surface flatness measuring device, all use an acceleration detector (Patent Documents 1 to 3).
JP 2005-315675 A JP 2010-66040 A JP2013-79889A Takuya Ikeda, Naoko Higashijima: Research on the method of measuring the international roughness index, Pavement Engineering Papers Vol. 3 December 1998 Kiyoshi Furukawa: Development of a real-time simple diagnosis system for pavement and expansion / contraction devices using dynamic response of patrol cars, Master's thesis, Department of Social Infrastructure, University of Tokyo, 2006 Yasuyuki Asakawa: Generalization and improvement of the simple road surface diagnosis system (VIMS) aimed at the development in Asia, The University of Tokyo Faculty of Engineering, Bachelor of Science in Social Infrastructure, 2007 Tomoyuki Asakawa: Improving versatility and performance analysis for practical application of Vehicle Intelligent Monitoring System-Based on application examples overseas-Master's thesis, Department of Social Infrastructure, University of Tokyo, 2009 Shinjiro Yano: Identification of vehicle response model considering driving posture and application to VIMS calibration, Thesis, Graduate School of Engineering, The University of Tokyo, 2011 Kosuke Takahashi: Improvement of VIMS simple calibration accuracy considering the required performance of IRI estimation and application in Kyrgyz, Master's thesis, Department of Social Infrastructure, University of Tokyo, 2011 Yuki Shimada: Proposal of IRI estimation method by VIMS using free running response of vehicle, Bachelor of Science in Social Infrastructure, University of Tokyo, 2012 Yoshihiko Miwa: Multipurpose image processing for road surface diagnosis with a view to integration with VIMS, The University of Tokyo, Faculty of Engineering

As a method to easily and quantitatively evaluate the riding comfort on the road surface, install an accelerometer, GPS, distance meter, etc. on the vehicle, record the acceleration response of the traveling vehicle, and estimate the riding comfort index for each evaluation route section (VIMS) is used, but the acceleration response varies depending on the location of the sensor in the vehicle, which causes the above-mentioned problems.
An object of the present invention is to eliminate dependence on a sensor installation location in a road surface diagnosis method using a vehicle response.
One of the more specific purposes of the present invention is to consider the IRI estimation method using a smartphone to simplify VIMS and facilitate daily inspection. Because it is limited, it provides an IRI estimation method that does not depend on the installation location using angular velocity response measurement.

The first technical means adopted by the present invention to solve this problem is as follows:
An angular velocity sensor installed in the measurement vehicle to obtain the pitching angular velocity of the vehicle;
GPS installed in the measurement vehicle so as to acquire GPS information including at least position information, time information, travel speed, and
Data recording means;
Data analysis means;
With
The data recording means records the pitching angular velocity of the vehicle acquired by the angular velocity sensor in synchronization with the GPS information acquired by the GPS,
The data analysis means includes a transfer function from an angular velocity response of a measurement vehicle to an acceleration response of a quarter car (hereinafter referred to as “QC”) which is a reference virtual vehicle, and an acceleration response of the QC and an international roughness index (IRI). And a correlation function
The data analysis means estimates a QC acceleration response using the acquired vehicle pitching angular velocity and the transfer function, and uses the estimated QC acceleration response and the correlation function to obtain an international roughness index (IRI). Estimate
Road surface evaluation device.
In one aspect, the transfer function and the correlation function are acquired in advance and stored as part of software constituting the data analysis means.

In one aspect, the transfer function is an inter-vehicle response amplitude ratio defined as the ratio of the angular velocity response of the measurement vehicle to the acceleration response of the QC.
In one aspect, the inter-vehicle response amplitude ratio is the acceleration of the QC obtained by running the same road surface by simulation and the angular velocity response of the measured vehicle acquired by running a route with a known road surface profile by the measuring vehicle. It is the power spectrum ratio of the response.
In the embodiment described later, the square root of the power spectrum ratio is used as the vehicle-to-vehicle response amplitude ratio, the QC acceleration RMS (Root Mean Square) value is estimated from the pitching response angular velocity, and the QC acceleration is calculated from the estimated acceleration RMS value. IRI is calculated using the correlation between RMS and IRI.
The angular velocity response of the measurement vehicle is also affected by the type and travel speed of the vehicle. Therefore, if the angular velocity response of the measurement vehicle is not obtained in advance, it is measured by running the measurement vehicle on a road with a known road profile. The vehicle angular velocity response is acquired, and the inter-vehicle response amplitude ratio is calculated using the QC acceleration response obtained by running the same road surface by simulation.
Further, it will be understood by those skilled in the art that existing calibration methods such as those described in Non-Patent Documents 3 to 7 can be applied to the present invention. For example, the response amplitude ratio of an arbitrary speed can be created by linear interpolation at the angular velocity as well as the response amplitude ratio of the acceleration. Like the conventional VIMS, if the speed fluctuation is below a certain level, the IRI is accurate even when the vehicle is not running at a constant speed Can be estimated well. When it is difficult to measure the road surface profile, it is considered that the versatility is further enhanced by combining with the hybrid calibration proposed by Takahashi.
Further, IRI estimation may be performed by applying a low-pass filter to the measured angular velocity response to limit the frequency band.

The data recording means and the data analysis means can be configured from a computer (including an input unit, an output unit, a storage unit such as a RAM and a ROM, a processing unit mainly including a CPU, and the like).
In one aspect, the angular velocity sensor includes a gyroscope, but the type of the angular velocity sensor is not limited.
In one mode, the above-mentioned road surface evaluation device comprises a smart phone provided with the angular velocity sensor, the GPS, the data recording means, and the data analysis means. In fact, the proposed method can be applied to the angular velocity and GPS information measured using a smartphone, and IRI estimation is possible, it does not depend on the installation location, and the IRI estimation method using a smartphone is the same as the conventional VIMS. It was confirmed that the accuracy was almost the same. Moreover, the image of a road surface can also be acquired using the camera function of a smart phone.

The second technical means adopted by the present invention is:
An angular velocity sensor installed in the measurement vehicle to obtain the pitching angular velocity of the vehicle;
GPS installed in the measurement vehicle so as to acquire GPS information including at least position information, time information, travel speed, and
A transfer function from the angular velocity response of the measurement vehicle to the acceleration response of the quarter car (hereinafter referred to as “QC”), which is the reference virtual vehicle,
Correlation function between acceleration response of QC and International Roughness Index (IRI),
Use
Recording the pitching angular velocity of the vehicle acquired by the angular velocity sensor in synchronization with the GPS information acquired by the GPS;
Using the acquired vehicle pitching angular velocity and the transfer function to estimate the acceleration response of the QC;
Estimating an International Roughness Index (IRI) using the estimated QC acceleration response and the correlation function;
The road surface evaluation method which consists of.

In one aspect, the transfer function is an inter-vehicle response amplitude ratio defined as the ratio of the angular velocity response of the measurement vehicle to the acceleration response of the QC.
In one aspect, the inter-vehicle response amplitude ratio is the acceleration of the QC obtained by running the same road surface by simulation and the angular velocity response of the measured vehicle acquired by running a route with a known road surface profile by the measuring vehicle. It is the power spectrum ratio of the response.

  In one aspect, the angular velocity sensor and the GPS are mounted on a smartphone, and the steps are executed by the smartphone.

The third technical means adopted by the present invention is a computer program for causing a computer to execute the road surface evaluation method.
In one aspect, the computer program is stored on a computer readable medium.

The present invention solves the problems caused by the installation location of the road surface property investigation method using acceleration response measurement by using the angular velocity response instead of the conventional acceleration response.
According to the present invention, the ride comfort index, that is, the IRI can be estimated regardless of where the measurement device is installed in the vehicle.
According to the present invention, a simple measurement device such as a smartphone is installed at the driver's hand and traveled, so that a simple and low-cost riding comfort diagnosis can be performed.

It is a conceptual diagram of a Quarter Car model. It is a conceptual diagram of the conventional VIMS. It is a figure which shows the correlation of QC acceleration RMS and IRI. It is a figure explaining the position dependence of acceleration and angular velocity. It is a figure which shows the relationship between the acceleration measured at two positions (position A, position B) shown in FIG. 4, and a frequency. It is a figure which shows the relationship between the angular velocity measured at two positions (position A, position B) shown in FIG. 4, and a frequency. It is a conceptual diagram of VIMS concerning the present invention. (A) is Tomei Expressway road surface profile (B) is Tomei Expressway road surface profile power spectrum, and (C) is Tomei Expressway IRI _true . The upper figure shows the response amplitude ratio (downline) between the measurement data and QC acceleration, where (a) angular velocity and (b) acceleration. The figure below shows the response amplitude ratio (upline) of measured data and QC acceleration, where (a) angular velocity and (b) acceleration. The upper figure shows IRI estimated values (downline), (a) angular velocity and (b) acceleration. The figure below shows the IRI estimated value (upline): (a) angular velocity, (b) acceleration. It is the figure which contrasted IRI estimation using a road surface profile, IRI estimation using vertical acceleration, and IRI estimation using pitching angular velocity. It is a figure explaining angular velocity measurement in a different position of vehicles by a smart phone. It is a figure which shows the power spectrum comparison in each measurement position.

[A] Background Art of the Present Invention A general concept and technique as the background of the present invention will be described. The concepts and techniques described below are not only conventional techniques for the present invention, but also necessary techniques for understanding the present invention. In addition, some techniques described below are techniques that can be used to implement the embodiments according to the present invention.

[A-1] International Roughness Index (IRI)
IRI is an index for evaluating road flatness proposed by the World Bank in 1986. IRI is obtained by the motion of a Quarter Car (QC) model, which is a virtual vehicle model that takes out only one wheel of a two-axle and four-wheel car (Fig. 2).
The equation of motion of the QC model is expressed by equation (2.1).
here,
z _s : absolute displacement of sprung mass [mm],
z _u : absolute displacement of unsprung mass [mm],
m _s : sprung mass,
m _u : unsprung mass,
c _s : damping constant of shock absorber,
k _s : elastic modulus of suspension,
k _t is the elastic modulus of the tire.
In addition, the value of each parameter is defined as follows.

IRI is obtained by dividing the sum of the relative displacement of the sprung mass and unsprung mass when the QC travels on the road at 80 km / h by the evaluation section length L, and is defined by equation (2.3) Is done.
Here, L is the evaluation section length, and V is the traveling speed (80 km / h).
IRI's evaluation section length L is generally 200m or 0.1 miles (160m). In the IRI calculation of this embodiment, the evaluation interval is set to 200 m in principle, and the IRI is calculated while shifting the interval by 10 m.

[A-2] Basic configuration of VIMS
Several methods are known for calculating IRI, and VIMS is one of them. VIMS measures the roughness index of an arbitrary scale with a response type road roughness measuring device and converts it to IRI using a correlation equation (Fig. 1). The conventional VIMS is composed of an accelerometer, GPS, notebook computer and data aggregation software.

The accelerometer is used to acquire the vertical acceleration of the vehicle. As the accelerometer, for example, a + 5V power supply analog acceleration sensor SD2422-002 manufactured by Silicon Design can be used. In one example, the recording sampling frequency is 1000 Hz. However, the recording sampling frequency is not limited to this. For example, sampling may be performed at 200 Hz. Although this accelerometer is a three-axis accelerometer, a one-axis acceleration sensor may be used. The accelerometer shall be installed in the interior of the vehicle, and since the interior structure differs depending on the interior of the vehicle, select an appropriate location for each vehicle type.

By using GPS, it is possible to estimate the latitude, longitude, and travel speed of the measurement location. As will be described later, the transfer function according to the present invention is speed-dependent, and it is important to estimate the traveling speed. GPS information is also used for IRI mapping after IRI estimation. As the GPS, for example, 747pro manufactured by Transystem Co., Ltd. can be used. As the aggregation software, for example, LabVIEW2012 of National Instrument can be used. LabVIEW is software that can record data using a graphical programming language. The VIMS measurement program is described by this software. The information finally output is an acceleration waveform (voltage) and GPS information (latitude, longitude, speed, date and time). The inventor's laboratory uses the Land Cruiser (TOYOTA) as a measurement vehicle for VIMS.

[A-3] IRI estimation process
VIMS converts vertical acceleration into QC acceleration RMS using a transfer function (inter-vehicle response amplitude ratio), and calculates IRI using a regression equation between RMS and IRI. Here, the most basic process of VIMS based on the 60km / h constant speed driving by Land Cruiser proposed and verified by Furukawa and Asakawa is described.

[A-3-1] An acceleration response obtained by running a vehicle with a known vehicle-to-vehicle response amplitude ratio road profile with a measurement vehicle, and an acceleration response of a QC obtained by running the same road surface by simulation. The square root of the power spectrum ratio is defined as the response amplitude ratio between the measurement vehicle and the QC vehicle, and is expressed by equation (2.4).
Where ω: spatial angular frequency [cycle / m], | TF (ω) |: vehicle-to-vehicle response amplitude ratio from measurement vehicle to QC, PSD _QC (ω): vertical acceleration power spectrum of QC, PSD _car (ω): acceleration power spectrum in the vertical direction of the measurement vehicle.

[A-3-2] Conversion of QC to acceleration RMS Parseval's formula is used as a method for converting the acceleration response of a measurement vehicle to QC acceleration RMS. This makes use of the fact that the integral of the QC sprung mass acceleration, which is a time function, is equal to the integral of the Fourier transform of the QC sprung mass acceleration. Equation (2.5) shows the equation for estimating the RMS RMS of the sprung mass of the QC from the vertical acceleration response of the measurement vehicle.
here,
RMS: QC sprung mass vertical acceleration RMS,
| TF (ω) |: Inter-vehicle response amplitude ratio from the measurement vehicle to the QC.

[A-3-3] Use of correlation between acceleration RMS and IRI of QC As shown in FIG. 3, it is shown that there is a strong correlation between the vertical acceleration RMS and IRI of the sprung mass of QC. Using this relationship, a relational expression is estimated based on a simulation in a virtual test course made from an actual road surface. The IRI evaluation section length is generally 200m, but in actual road inspections, there are cases where the evaluation section length is shortened, so three types of correlation equations are required for each evaluation section length, both of which are very high. Indicates the correlation coefficient. The correlation equation is shown in Table 1. In this embodiment, the evaluation section length is 200 m.
In some cases, the VIMS measurement uses a low-pass filter applied to the acceleration. Even in this case, the correlation equation is obtained as shown in Table 2, and these also have a high correlation coefficient. However, when the cut-off frequency is low, local road surface characteristics are also cut. When traveling at 60km / h, 1cycle / m = about 16.7Hz.

[A-4] Principle of speed calibration The basic VIMS measurement conditions were constant speed travel of 60 km / h, but speed calibration allows measurement at various speeds. The velocity calibration described below is a technique that can be applied to this embodiment using angular velocity data.

[A-4-1] Speed calibration based on response amplitude ratio measurement A speed calibration method based on response amplitude ratio measurement proposed by Asakawa will be described. The speed response amplitude ratio is obtained by traveling at the constant speed at various speeds on the same road surface and taking the ratio of the power spectrum of the acceleration response. Then, the inter-vehicle response amplitude ratio, which is the ratio of the power spectrum of the acceleration response between the measurement vehicle that traveled at 60 km / h and the QC that traveled at 80 km / h, is multiplied by this inter-speed response amplitude ratio. Obtain the response amplitude ratio.
This process is expressed as follows.
Where TF _{dif_speed} : Inter-vehicle response amplitude ratio for different speeds, TF _{60 km / h} : Inter-vehicle response amplitude ratio for 60 km / h, TF _{bet_speed} : Inter-speed response amplitude ratio,
It is.

[A-4-2] Calibration method based on vehicle model identification This method is proposed by Yano. Create a vehicle model that exhibits a vertical acceleration response equivalent to the vehicle response when getting over a hump. The vehicle model used in this method is a two-shaft, four-wheel vehicle model (full car model) that also takes into account the rotational movement of the vehicle body. By changing the simulation speed of this model, a frequency response function when the speed is different is calculated, and based on this, an inter-vehicle response amplitude ratio when the speed is different is obtained, and speed calibration is performed.

[A-4-3] Hybrid Calibration Method This method was proposed by Takahashi and is a combination of the above two calibration methods. A numerical model is created based on the acceleration response of the actual vehicle obtained by the hump test, and the inter-vehicle response amplitude ratio between the model and QC is obtained. Furthermore, by changing the speed on the same road surface, the response ratio between speeds is obtained, and speed calibration is performed. In the hybrid calibration method, the QC model was used instead of the full car model to reduce the time required for parameter identification.

[A-4-4] Speed calibration when speed changes
As mentioned above, VIMS has been developed into a highly versatile system, as various calibration methods have been studied, but all of these studies were subject to constant speed driving. Shimada calculated the inter-vehicle response ratio for typical speeds, and showed that inter-vehicle response ratios of arbitrary speeds can be obtained by linear interpolation. In the 40-70 km / h speed range, 10 km / h increments, 70-110 km In the speed range of / h, the vehicle-to-vehicle response ratio is required in increments of 20km / h. This linear interpolation is expressed as follows.
The inter-vehicle response ratio obtained by linear interpolation using TF _V1 and TF _V2 when the speed vkm / h satisfies v ₁ <v <v ₂ is defined as TF _{in, v} .
Here, TF _{in, v is} the vehicle-to-vehicle response ratio when the speed is vkm / h.

[A-5] IRI estimation method during free running Shimada uses the inter-vehicle response ratio of the above arbitrary speed, so that if the speed fluctuation range of the IRI evaluation section is below a certain level, the inter-vehicle according to the speed for each evaluation section It was shown that IRI can be estimated even in free running that is not constant speed using response ratio. When the section average velocity of IRI evaluation interval was V _m, is available free running type estimation method satisfies the speed v _{V m -0.1 V m <v <} V m + 0.1V m in evaluation interval. The average speed of the IRI evaluation section is obtained, and the IRI is estimated using the inter-vehicle response ratio of that speed.

[B] IRI estimation using angular velocity [B-1] Angular velocity response and acceleration response While the acceleration response of the vehicle depends on the measurement position, assuming that the vehicle is a rigid body, the angular velocity response is independent of the measurement position. It is thought that there is. In order to confirm that the angular velocity response is independent of the sensor position, a highly accurate gyroscope (Crossbow's inertial measurement unit VG400CA-200-1) is used as the acceleration and angular velocity measurement device. It was installed at a position directly above the wheel) and at position B (vehicle front center), and the measured values of acceleration and angular velocity were compared (FIG. 4). Vertical acceleration and angular velocity in the pitching direction are used as values for verifying the difference between the measurement values. The measurement was performed while driving at a constant speed of 80 km / h on the expressway. As shown in FIG. 5A, in the acceleration response, the position A and the position B were clearly different values in the vicinity of 2 Hz. On the other hand, as shown in FIG. 5B, in the angular velocity signal, the values acquired at the position A and the position B at 8 Hz or less are common. The angular velocity response of the vehicle has a wider frequency band in which the measurement values coincide with each other than the acceleration response, and is less affected by the change of the measurement position. The same tendency was observed at other travel speeds other than 80 km / h.

As described above, it was confirmed that when the measurement position is changed, the angular velocity can be regarded as the same in a wider frequency band than the acceleration. Conventional VIMS measures acceleration, but in this embodiment, it is possible to estimate effective IRI even at different measurement positions by adopting an IRI estimation method using angular velocity. In the following, IRI estimation method based on angular velocity is proposed, and its accuracy is compared with IRI estimation accuracy based on acceleration.

[B-1] Proposal of IRI estimation method using angular velocity
Then, the power spectrum PSD _ang of the angular velocity response of the actual vehicle is expressed by the following equation (4.1).

If the transfer function from the road surface P to the acceleration response of _QC is TF _QC , the power spectrum PSD _QC of the acceleration response of _QC is expressed by the following equation (4.2).

Is represented by the following equation (4.3).

Therefore, the acceleration RMS of the QC sprung mass can be estimated from the pitching angular velocity as in the following equation (4.4) as in the equation (2.5).
Thus, if the pitching angular velocity is measured, the RMS value of the QC can be estimated, and the IRI can be obtained from the correlation between the QC acceleration RMS and the IRI described above.

[B-2] Setting of reference IRI
Set IRI as a reference for verification of IRI estimation accuracy. The following two IRIs are considered as standards.
IRI (Formula 2.3) → IRI _true calculated according to the definition by performing QC simulation using road surface profile measured by DAM or high-performance inspector
Profile with the inter-vehicle response amplitude ratio TF _X is an acceleration ratio of the power in the spectrum when the acceleration power spectrum and the measurement vehicle when the known road QC is traveling at 80 km / h was constant speed running in X km / h Calculated IRI (see [A-3]) → IRI _VIMS
In this embodiment, in order to propose a new IRI estimation method, basically, IRI _true is used as a reference IRI for comparison and examination.
The C value defined by the following equation (4.5) is used as an index for verifying IRI estimation accuracy.
Here, IRI (bar): IRI for verification, IRI _st : reference IRI, N: number of data.

[B-3] Inter-vehicle response amplitude ratio characteristics
Measurements were taken in a predetermined section of the Tomei Expressway up and down line with a known road profile. IRI was calculated by performing QC simulation on the road surface profile of the measurement lane in this section. FIG. 7A shows the Tomei Expressway road surface profile, FIG. 7B shows the Tomei Expressway road surface profile power spectrum, and FIG. 7C shows the Tomei Expressway IRI _true .

In order to obtain the vehicle-to-vehicle response amplitude ratio at each speed, a constant speed running test is performed at a speed from 70 km / h to 110 km / h.
Two inertial measurement units are installed, and the difference in measured values depending on the installation position can be examined. In addition, in order to compare with IRI estimation using VIMS accelerometer, an accelerometer is also installed and measured directly above the rear wheel of the vehicle.

70km / h to 90km / h, 90km / h to 110km / h were obtained from this measurement.
Is obtained as shown in the upper diagram (a) of FIG. 8 and the lower diagram (a) of FIG. In this study, the inter-vehicle response ratio is calculated by equation (4.3) in the 800m section running at a constant speed. In addition, a low-pass filter with a cut-off frequency of 1.0 cycle / m is applied during calculation. For comparison, the inter-vehicle response amplitude ratio TF (ω) obtained by VIMS is also shown in FIG. 8 (b) and FIG. 8 (b).

[B-4] IRI Estimation IRI is estimated using the obtained QC acceleration and the response amplitude ratio of the actual vehicle angular velocity response. Like the conventional VIMS, the measured angular velocity of the 200m section is converted to the acceleration RMS of the QC using the inter-vehicle response amplitude ratio, and the IRI is estimated using the regression equation established between the RMS and the IRI. The IRI estimated value at Vkm / h constant speed is defined as IRI _V. The upper diagram in FIG. 9 shows the IRI estimated value (downlink), and the lower diagram in FIG. 9 shows the IRI estimated value (uplink). As shown in FIG. 9, it was confirmed that even if the IRI was estimated from the angular velocity, the IRI could be estimated with almost the same accuracy as the acceleration. Table 3 below summarizes the estimation error (C) from IRI _true for each velocity estimated from the angular velocity and acceleration.
Further, as shown in FIG. 10, the values of IRI estimation using a road surface profile, IRI estimation using vertical acceleration, and IRI estimation using pitching angular velocity are substantially the same.

The accuracy of the IRI estimated using the measured values with different measurement positions was verified. Even if the measurement position was changed, if the measurement position at the time of response amplitude ratio calculation and IRI estimation was the same, the estimation accuracy was found to be equivalent to that directly above the left rear wheel, which is the conventional acceleration measurement position. In addition, even when the measurement position at the time of response amplitude ratio calculation and IRI estimation is different, the response ratio created at the measurement value with different measurement position matches up to around 0.3 cycle / m, so the response amplitude created at a specific location It was found that IRI can be accurately estimated by using the ratio, that is, the component of angular velocity response up to 0.3 cycle / m at the time of IRI estimation. The numerical value of 0.3 cycle / m is between the frequency of the primary mode (0.06-0.09 cycle / m) and the frequency of the secondary mode (0.5-0.8 cycle / m).

[C] IRI can be estimated with the same accuracy even if angular velocity is used instead of IRI estimated acceleration using a smartphone, and in order to enable IRI estimation regardless of the change of measurement position, angular velocity of 0.3cycle / m It was confirmed that the above components may be used. In order to improve VIMS into a simpler and cheaper system, an IRI estimation method using a smartphone is proposed. Smartphones with integrated sensor and operation units have a high need for installation near the driver and measurement vehicle, and the installation location is expected to vary depending on the vehicle. Instead, focus on the pitching angular velocity as a response independent of the installation location.

[C-1] Measuring equipment The equipment required for IRI estimation using angular velocity is GPS, angular velocity meter, and data recording software, but these can all be bundled or installed in Apple's smartphone iPhone4S. There is no need to prepare additional measuring equipment. An iPhone app that can measure and record 3-axis acceleration and angular velocity / position / velocity information has been developed, and this app is a system that records GPS position time information, acceleration, angular velocity, and video in synchronization, with a measurement time of up to 60 minutes is there. At the time of measurement, the smartphone is fixed to the front center of the vehicle using a double-sided tape. The performance specifications of the angular velocity meter are shown below.

[C-2] IRI estimation accuracy verification When the accuracy of the IRI estimation method using the smartphone was verified, it was confirmed that the IRI was estimated with the same accuracy as the conventional VIMS by the IRI estimation using the smartphone. In addition, the creation of inter-vehicle response amplitude ratios at arbitrary speeds by linear interpolation proposed by Shimada, and whether the free-running estimation method can be used even with a smartphone by using the IRI estimation method during free-running using them As a result, it was confirmed that IRI estimation using smartphones has the same accuracy as conventional VIMS, including free running.

[C-3] Verification of measurement position In the IRI estimation method using a smartphone, the installation of the measurement device only installs the smartphone. For this reason, it is considered that the IRI can be estimated wherever it is installed in a level place within the reach of the operator. The measured values are compared with the inertial measurement unit (gyroscope) installed at multiple locations on the left rear wheel. The measurement position (fixed position) is the front center of the vehicle (A), near the shift lever (B), near the passenger door knob (C), on the dashboard (D), the video shooting holder (E), and the inertial measurement unit true Above (F). The video shooting holder is used to fix the smartphone during video measurement. Each power spectrum is compared and the power spectrum in each measurement position is shown in FIG.

As shown in FIG. 12, the measured values up to around 3 Hz almost coincided with each other, and the values after 3 Hz were close except for the position (E). From this, it is considered that the measurement position can be freely changed if the measurement with the smartphone is installed so that the rotation axis of the pitching angular velocity is parallel to the wheel axis. Moreover, when the power spectrum was compared by the spatial frequency, the measured values were almost the same up to around 0.3cycle / m. If the smartphone is installed so that the rotational axis of the pitching angular velocity is parallel to the wheel axis, the measurement position can be changed by using a low-pass filter with a response ratio created at a specific location and a cutoff frequency of 0.3 cycle / m. However, IRI estimation using a smartphone is considered possible with the same accuracy as conventional VIMS.

The present invention solves the problem of the installation location of the road surface property investigation method using acceleration response measurement, and can be easily mounted on a smartphone that is becoming widespread.

Claims (10)

An angular velocity sensor installed in the measurement vehicle to obtain the pitching angular velocity of the vehicle;
GPS installed in the measurement vehicle so as to acquire GPS information including at least position information, time information, travel speed, and
Data recording means;
Data analysis means;
With
The data recording means records the pitching angular velocity of the vehicle acquired by the angular velocity sensor in synchronization with the GPS information acquired by the GPS,
The data analysis means includes a transfer function from an angular velocity response of a measurement vehicle to an acceleration response of a quarter car (hereinafter referred to as “QC”) which is a reference virtual vehicle, and an acceleration response of the QC and an international roughness index (IRI). And a correlation function
The data analysis means estimates a QC acceleration response using the acquired vehicle pitching angular velocity and the transfer function, and uses the estimated QC acceleration response and the correlation function to obtain an international roughness index (IRI). Estimate
Road surface evaluation device.   The road surface evaluation device according to claim 1, wherein the transfer function is an inter-vehicle response amplitude ratio defined as a ratio between an angular velocity response of a measurement vehicle and an acceleration response of QC.   The vehicle-to-vehicle response amplitude ratio is the angular velocity response of the measurement vehicle obtained by running the route with a known road surface profile with the measurement vehicle, and the power spectrum ratio of the acceleration response of QC obtained by running the same road surface by simulation. The road surface evaluation apparatus according to claim 2, wherein   The road surface evaluation apparatus according to any one of claims 1 to 3, wherein the road surface evaluation apparatus includes a smartphone including the angular velocity sensor, the GPS, the data recording unit, and the data analysis unit. An angular velocity sensor installed in the measurement vehicle to obtain the pitching angular velocity of the vehicle;
GPS installed in the measurement vehicle so as to acquire GPS information including at least position information, time information, travel speed, and
A transfer function from the angular velocity response of the measurement vehicle to the acceleration response of the quarter car (hereinafter referred to as “QC”), which is the reference virtual vehicle,
Correlation function between acceleration response of QC and International Roughness Index (IRI),
Use
Recording the pitching angular velocity of the vehicle acquired by the angular velocity sensor in synchronization with the GPS information acquired by the GPS;
Using the acquired vehicle pitching angular velocity and the transfer function to estimate the acceleration response of the QC;
Estimating an International Roughness Index (IRI) using the estimated QC acceleration response and the correlation function;
Road surface evaluation method consisting of   The road surface evaluation method according to claim 5, wherein the transfer function is an inter-vehicle response amplitude ratio defined as a ratio between an angular velocity response of a measurement vehicle and an acceleration response of QC.   The vehicle-to-vehicle response amplitude ratio is the angular velocity response of the measurement vehicle obtained by running the route with a known road surface profile with the measurement vehicle, and the power spectrum ratio of the acceleration response of QC obtained by running the same road surface by simulation. The road surface evaluation method according to claim 6, wherein   The road surface evaluation method according to claim 5, wherein the angular velocity sensor and the GPS are mounted on a smartphone, and the steps are executed by the smartphone.   The computer program for making a computer perform the road surface evaluation method of any one of Claims 5-8. A computer-readable medium in which the computer program according to claim 9 is stored.