JPH0290092A - Apparatus for recognizing state of road surface - Google Patents

Abstract

PURPOSE:To obtain data in which state a dry state, a wet state and the like the road surface is by operating the membership function values for the output levels of sensors for the positive irregular reflected light of visible light, infrared-ray irregular reflected light, road surface temperature and the like from the road surface. CONSTITUTION:An irregular reflected light receiver 11 includes photodetectors 21 and 22 for infrared rays and visible light. The outputs of the photodetectors are amplified. Then, noise components are removed. A positive reflected light receiver 14 includes a photodetector 24 for visible light. The output signal from the photodetector 24 is also amplified, and noise component is removed. The detected signal from a thermometer 23 is amplified and outputted. The output signals from the elements 21, 22, 23 and 24 are normalized in normalizing circuits 15. Then, the signals are converted in A/D converter 16. Thus, the signals representing the levels of the visible irregular reflected light, the infrared-ray irregular reflected light, the road surface temperature and the visible regular reflected light are obtained. The apparatus for recognizing the state of the road surface is composed of a dry inference part 30, a wet inference part 40, a lamination inference part 50, freeze inference part 60, a MAX circuit 70 as a generalizing circuit and an output device 71. In the inference parts 30-60, membership functions such as output level distribution of visible-light irregular reflection of each road surface state are preset.

Description

DETAILED DESCRIPTION OF THE INVENTION Summary of the Invention The present invention detects visible light, diffusely reflected light, infrared diffusely reflected light, road surface temperature, etc. from the road surface. By setting membership functions that represent the output level distribution of these sensors, obtaining membership function values for the measured sensor outputs, and calculating these membership function values, you can determine whether the road surface is dry, wet, snowy, or Obtain information on which state of freezing is likely to occur.

BACKGROUND OF THE INVENTION This invention relates to an apparatus for identifying road surface conditions such as dry, wet, snowy or frozen.

Ensuring the safety of drivers by taking appropriate measures to deal with traffic disturbances caused by frozen roads, snow accumulation, etc. is a very important issue in transportation system management. It is necessary to accurately grasp the road surface condition.

SUMMARY OF THE INVENTION An object of the present invention is to provide a device that can identify a desired road surface condition among dry, wet, snowy, and frozen road conditions.

The road surface condition detection device according to the present invention includes a means for detecting the level of visible diffusely reflected light and a means for detecting the level of infrared diffusely reflected light depending on the road surface condition selected from dry, wet, snowy, and frozen. at least two detection means selected from means for detecting road surface temperature, means for detecting the level of visible specular reflection light, and the level of visible diffusely reflected light and the level of infrared diffusely reflected light in the road surface condition to be identified.

Membership functions representing distributions of at least two items selected from road surface temperature and visible specular reflection light level are respectively set, and membership function values corresponding to the output signals are determined in response to the output of the selected detection means. At least two membership function output means output, respectively, and predetermined calculations are performed on the membership function values output from the membership function output means to obtain information regarding the possibility of one selected road surface condition to be identified. It is characterized by comprising a calculation means for generating.

According to the present invention, the degree of conformity of the output of the detection means to the membership function regarding the item selected according to the road surface condition to be identified is determined, and by performing a predetermined calculation on this degree of conformity, it is possible to identify the road surface condition to be identified. Get accurate information about sexuality. Therefore, it becomes possible to accurately judge the road surface condition based on this information.

DESCRIPTION OF THE EMBODIMENTS In FIG. 1, a pillar 17 is erected on one side of a road 19, and a support arm 18 extends upward from the vicinity of the upper end of the pillar 17 toward the upper side of the road 19. A projector 15. is attached to this support arm 18. Diffuse reflection receiver 11. A specular reflection light receiver 14 and a light projection level measuring device 16 are respectively attached, and a road surface thermometer 23 is provided on the support column 17. Floodlight 1
5 uses a water pot lamp that includes not only visible light but also near-infrared light in one emission spectrum. The projector 15 is arranged so as to be directed toward the road surface at an appropriate angle of incidence. The specular reflection receiver 14 receives specularly reflected visible light of the road surface reflected light projected from the light projector 15 and converts it into an electrical signal. The diffused reflection receiver 11 separately detects visible light and near-infrared light reflected from the road surface. Near-infrared light has a wavelength of about 1.5 to 2.
A thickness in the range of 5 (μm) is preferable. The light emitting level measuring device 16 directly receives the light from the projector 15, and
In this example, the amount of visible light is measured. The light projector 15 is driven and controlled based on the output signal of the level measuring device 16 so that the amount of light projected is always constant. A radiation thermometer is used as the road surface temperature gauge 23.

As shown in FIG. 7, the double-reflection light receiver 11 includes light receiving elements 21 and 22 for infrared light and visible light.
After the output of 22 is amplified by an amplifier, it is passed through a bandpass filter (
Noise components are removed in each case (both paths shown in the figure). The specular reflection light receiver 14 includes a visible light light receiving element 24, and the output signal of this light receiving element 24 is also outputted through an amplifier and a bandpass filter (both paths shown). A signal representing the temperature detected by the thermometer 23 is also amplified and output.

These light receiving elements 21.22. Thermometer 23. Light receiving element 2
4 (hereinafter referred to as sensors 21 to 24) are each normalized by a normalization circuit 15 and then sent to an A/D conversion circuit 1.
6, it is converted into a digital signal. In this way, signals representing the levels of visible diffusely reflected light, infrared diffusely reflected light, road surface temperature, and visible specularly reflected light are obtained.

Figures 2 to 5 show each of the above sensors 2 depending on the road surface condition.
1 to 24 respectively show distributions of output signal levels. The horizontal axis represents the normalized level of each sensor, and the vertical axis represents the grade of each level. The grade is a value corresponding to the probability of occurrence of each level obtained by statistical processing of a large amount of measurement data. This can be said to be the degree to which various sensor output levels belong to each road surface condition.

It can be said to be a kind of membership function.

FIG. 2 shows the relationship between visible light diffuse reflection output level distribution and road surface condition. In wet conditions, the road surface becomes mirror-like, so the visible light diffuse reflection output level decreases.

The distribution is skewed toward lower levels. On the other hand, since snow that appears white is a diffusing surface with a light reflectance of nearly 100%, the visible light diffuse reflection output level increases and its distribution shifts toward higher levels. In snow jam conditions or snow soiled with mud, etc., the surface becomes a diffusing surface with low reflectance, and the visible light diffuse reflection output concentrates around the intermediate level. This is almost the same as the dry state.

FIG. 3 shows the infrared light diffuse reflection output level distribution and the road surface condition. Since infrared light is easily absorbed by moisture, the level of diffusely reflected light tends to be low in snowy conditions. Although illustration is omitted, the infrared light diffuse reflection output level distribution shows the same tendency in wet and frozen conditions as in snowy conditions. On the other hand, in a dry state, infrared light is not absorbed, so the diffused reflection output level increases relatively, and the level is distributed mostly at high levels.

FIG. 4 shows the relationship between road surface temperature distribution and road surface condition. Naturally, in freezing and snowy conditions, the temperature is low and the distribution is biased towards lower temperatures. Conversely, in wet conditions, the road surface temperature will be distributed more toward the higher side. In a dry state, any temperature can be approximately equal, so the distribution will be approximately flat.

FIG. 5 shows the visible light specular reflection output level distribution and the road surface condition. As mentioned above, in a wet state and a frozen state, the mirror surface becomes a mirror surface, so the specular reflection output increases and its level distribution concentrates on the higher side, but the level is higher in the frozen state. In snowy conditions, the surface becomes a diffusing surface, so the specular reflection output of visible light decreases and its level distribution concentrates on the lower side. Dry conditions also show the same tendency.

FIG. 6 summarizes the above sensor output level distribution for each road surface condition. The level distribution shown in this figure is hereinafter referred to as a membership function. Membership functions not shown in FIGS. 2-5 are shown in dashed lines. Furthermore, the membership function of the visible light diffused reflection output for the snow jam and dirty snow in FIG. 2 is omitted.

The membership function of the infrared diffuse reflection output level in the dry state is completely opposite to that of the other three road surface conditions, so the dry state can be distinguished from other road surface conditions by using the infrared diffuse reflection output. It is.

For wet conditions, the membership function of visible light specular reflection output level is different from that of other road surface conditions, so it is thought that identification can be done using this membership function. Since the ship function is relatively similar to that for snow and ice conditions, it is preferable to consider other sensor outputs as well. The membership function that shows a marked difference between the snowy state and the wet state is the visible light diffuse reflection output level distribution,
The contrast between frozen and wet conditions is the temperature distribution. Therefore, it is preferable to use membership functions of the visible light diffuse reflection output level distribution, the temperature distribution, and the visible light specular reflection output level distribution to determine the wet state.

Similarly, membership functions of the visible light diffuse reflection output level distribution and the infrared light diffuse reflection output level distribution are preferably used to judge the snow cover state. It is preferable to use membership functions of temperature distribution and visible light specular reflection output level distribution to determine the frozen state.

Figure 7 shows the overall configuration of the road surface condition identification device.
An example of the operation is shown in FIG. This device includes a drying inference section 30. Moisture inference section 40. Snowfall inference section 50. Frozen reasoning section 60. MAX circuit 70. and an output device 7, for example, a display device.

Each inference unit 30 to 60 has four memories or memory φ areas 31 to 34, 41 to 44, 51 to 54, respectively.
61 to 84, and the visible light diffuse reflection output level distribution for each road surface condition shown in FIG. 5, respectively.

Membership functions of the infrared light diffuse reflection output level distribution, temperature distribution, and visible light specular reflection output level distribution are set in advance. Memories 32, 41, 43, . . . are indicated by solid lines.
As mentioned above, 44.51.52.63.84 stores membership functions necessary or preferable for determining each road surface condition, and other memories indicated by chain lines indicate those that are not necessarily necessary. . In each memory, the variable (level) of the membership function on the horizontal axis is specified by an address.

The signal (M) representing the visible light level detected by the visible light diffuse reflection sensor 21 is converted into a digital signal and then stored in the memory 3.
1.41.51.81, and this signal specifies the address of these memories.

As a result, the grades of the membership functions stored therein are read from these memories, and M
are applied to IN circuits 35, 45, 55, and 65, respectively. The read grade indicates the degree to which the output of the sensor 21 conforms to each membership function.

Similarly, the address of the memory 32.42.52.82 is specified by the output (R) of the infrared diffuse reflection sensor 22, and the corresponding grade is read out and sent to the MIN circuits 35, 45.
．． 55.65. Temperature sensor 23. The output of the visible light specular reflection sensor (-2°C, C) is similarly given to the corresponding memory, and the grade of the corresponding membership function is input to the corresponding MEN circuit.

The MIN circuit 35 is connected to the memory 31. 32. 33. 34
The lowest grade output from the MIN circuit 37 is selected and applied to the MIN circuit 37 at the next stage. Similarly, the MIN circuit 45 inputs the minimum output of the memories 41 to 44 to the MIN circuit 47. Other MIN circuits 55. [The same applies to 15.

Each of the inference units 30 to 60 is provided with a memory 3[1, 48, 58, 86 for storing grade 1 bar graph data in order to represent the inference results in bar graphs for each road surface condition. The bar graph data stored in these memories 36 to 66 represents bar graphs having different positions on the horizontal axis. These bar graphs, from left to right,
Dry, wet, 8! F snow.

They are arranged in order of frozen state.

MIN circuit 37.47.57.8 of each inference unit 30 to BO
7, these bar graph data are input to MIN circuit 3.
5. It is cut according to the calculation result of 45.55.65.

Then, the output data of the MIN circuits 37 to 67 are added in a MAX circuit 70 and provided to one display device 71. As a result, an inference result as shown in FIG. 7 or 8 is obtained. In the illustrated inference results, the bar graph indicating the snow-covered state has the highest value, indicating that the possibility that the road surface condition is snow-covered is extremely high.

If necessary, the state of the bar graph showing the maximum value can be determined to be the road surface state, and the result of this determination can be output.

In the above example, the snow condition refers to snow that appears white, but it is also possible to create a membership function for snow jams or dirty snow and use similar reasoning to identify the condition of dirty snow. can.

[Brief explanation of drawings]

Figure 1 is a diagram showing the settings of various sensors, Figures 2 to 5 are graphs showing the relationship between the output level distribution of each sensor and road surface conditions, and Figure 6 is a graph of Figures 2 to 5. FIG. 7 is a block diagram showing the configuration of the road surface condition identification device, and FIG. 8 is a graph showing its operation. 11... Diffuse reflection receiver. 14... Specular reflection receiver. 21, 22.24... Light receiving element. 23...Thermometer. 30, 40.50.80... Reasoning part. 31~34.41~44.51~54.61~64...
·memory. 35, 37.45. 47. 55. 57. 85
．． 87...MIN circuit. 3B, 4L 5B, 8B...Memory. that's all

Claims (1)

[Claims] Means for detecting the level of visible diffusely reflected light, means for detecting the level of infrared diffusely reflected light, and road surface according to the road surface condition to be identified selected from dry, wet, snowy, and frozen. At least two detection means selected from a means for detecting temperature and a means for detecting the level of visible specular reflection light, the level of visible diffuse reflection light, the level of diffuse infrared reflection light, the road surface temperature, and the visible specular reflection light level in the road surface condition to be identified. Membership functions representing the distribution of at least two items selected from among the reflected light levels are set, and each of the at least two members outputs a membership function value corresponding to the output signal in response to the output of the selected detection means. a membership function output means, and a calculation means for performing a predetermined calculation on the membership function value output from the membership function output means and generating information regarding the possibility of the selected road surface condition to be identified. A road surface condition identification device equipped with .