JP2581639B2 - Road surface freeze detection device self-diagnosis method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention improves the accuracy of road surface freezing prediction by automatically discriminating and removing a failed sensor by comparing the appearance pattern characteristics of observation data stored in advance for each point with the observation data. The present invention relates to a road surface freezing detection device self-diagnosis method.

[0002]

2. Description of the Related Art Road surface freeze detection detects and predicts road surface freeze based on output signals from sensors measured by a road surface temperature sensor, a temperature sensor, and a road surface moisture sensor. The management of general roads is a route to be managed, for example, 10 to 20 k.
Road surface freezing detectors are installed at several places where it is easy to freeze over the section of m, and the freezing is started from around the day to the next day by about 16:00 based on the information of the road surface condition, the road surface temperature, the temperature, and the moisture. It is necessary to predict whether or not it will occur, and if there is a possibility that freezing will occur, it is necessary to spray an antifreezing agent in the area where the occurrence is expected to prevent the occurrence of freezing and to ensure the safety of road traffic.

However, if each sensor of the road surface freezing detection device does not operate normally, an erroneous output will be output. In such a case, if the road surface is frozen but the output of the freezing detection is not output. Also, there has been a problem that a serious obstacle occurs in road management, such as an output of a freezing prediction signal that outputs an incorrect prediction result.

FIG. 1 shows a configuration of a conventionally used road surface freezing detecting device. The road surface freezing detection device 1 includes an air temperature sensor 4 fixedly mounted on a support arm 3 protruding from a support 2 standing on a road shoulder so as to extend to a position several meters above a road detection position (near the center of the lane) a, It is composed of a road surface temperature sensor 5 and a road surface moisture sensor 6.

The temperature sensor 4 is a ventilation type thermometer which is installed in the air several meters from the ground and measures the atmospheric temperature (hereinafter referred to as "temperature") by forced ventilation. The road surface temperature sensor 5 is an infrared non-contact thermometer that is installed so as to face the detection position a of the road surface in the air several meters from the ground, and measures the temperature of the detection position a in a non-contact manner. Further, the road surface moisture sensor 6 is installed in the air several meters from the ground so as to face the detection position a of the road surface, and the reflected light characteristic of the infrared ray projected toward the road surface at the detection position a depends on the presence of moisture. This is a road surface moisture meter that detects a difference and detects road surface moisture. The road area snow detects that the road surface becomes white due to the snow and the amount of reflected light increases.

The above-mentioned road surface freezing detection device 1 is designed to operate when the air temperature is lower than a set value (for example, 2 ° C.), the road surface temperature is lower than 1 ° C., and the road surface moisture or snow is detected.
"Freeze detection" is output. However, when the road surface temperature sensor 5 fails, the road surface temperature cannot be detected. This is true for the temperature sensor 4 or the road surface moisture sensor 6.
The same applies to.

[0007] In general, a large number of such road surface freezing detection devices are scattered around a bridge in a mountainous area or an urban area, and an accident such as a vehicle coming into contact with a sensor occurs. It may be. In this case, when the cause of the failure is simple and a significantly different output is output, the failure can be easily found.
If the output is close to the normal state, the failure may be overlooked, a long monitoring may be required to find the failure, or it may take a long time to find the abnormality. Sometimes it was judged normal.

[0008]

Therefore, conventionally, as a solution to such a problem, the frequency of maintenance in the winter season has been increased. However, in the case of poor contact, a normal value is output when the maintenance is performed. In many cases, it was difficult to find a failure to make a quick decision in a short survey. Further, it is almost impossible to constantly monitor all of the road surface freezing detection devices installed on the entire route due to the difficulty of road management in winter.

The present invention has been made in view of the above-mentioned problems, and detects a fault condition reliably even if the output is close to a normal condition so that the erroneous prediction as described above does not occur. A self-diagnosis method for a freezing detection device, a fault condition is output to a road surface freezing prediction device, and a self-diagnosis method for a road surface freezing detection device for predicting a fault excluding the faulty sensor. To propose a self-diagnosis method for road surface freezing detection device that will improve the prediction accuracy by detecting the sensor, transmitting the failure status to the centralized monitoring device through the communication line, and performing the prediction without the data of the bad sensor. It is the purpose.

[0010]

In order to achieve the above-mentioned object, a self-diagnosis method for a road surface freezing detection device according to the present invention determines an operation state of a sensor based on output data obtained from the sensor, and the device operates normally. The freeze prediction is performed using the output signal of only the sensor performing the freezing. 2 to 4
Is a graph showing, over time, the relationship between the road surface temperature and changes in temperature and moisture with respect to the weather, and uses a basic pattern due to the weather, such as sunny, rain, or snowfall, to determine whether the sensor is abnormal, and The gist of the invention is to exclude data from the sensor from data of the road surface freezing detection device.

It is known that the winter weather and the road surface environment have the following characteristics except for a severely cold area. (1) When it is sunny, the road surface is warmed by sunlight during the day, and the road surface becomes highest at around 14:00 in the afternoon of that day, and then gradually decreases. In this case, the road surface temperature becomes higher than the air temperature, and the road surface temperature becomes lower than the air temperature around 18:00. Road surface temperatures are warmed by the sun during the day, and radiative cooling lowers the ambient temperature during the night, with the lowest just before sunrise at dawn. That is, on a sunny day, 1
The road surface temperature and the air temperature are reversed around 8 o'clock, and thereafter the road surface temperature becomes lower than the air temperature. Eventually, the next day, when the sun starts to rise, the road surface will be heated by the sun, and the temperature will gradually rise and become higher than the temperature. Such a phenomenon shows almost the same pattern on a sunny day, and there is no road surface moisture at this time (see FIG. 2). (2) In the case of rain or snow, the air temperature and the road surface temperature change at substantially the same temperature. That is, it has been found that the rainfall becomes almost the same temperature as the air temperature and wets the road surface, and the road surface temperature becomes the same as the rainfall temperature. However, in the case of rain in winter, it is about 2 to 4 ° C, but in the case of snowfall, it is 0 to 1 ° C,
It tends to stay constant at the same temperature as long as rain or snow continues.

As described above, in a place where the sun hits the road surface on a sunny day, the road surface temperature rises sharply above the air temperature during the day, and then becomes lower than the air temperature after sunset. Furthermore, on rainy days and snowy days, there is almost no difference between the air temperature and the road surface temperature, and there is a unique pattern that the temperature becomes the same.

The failure of the sensor is diagnosed by utilizing such characteristics. If the road surface temperature rises rapidly when the temperature is rising slowly, the weather at that time will be "clear"
Is known empirically. Further, at this time, if an output appears on the moisture meter, the moisture meter is abnormal. Next, when the road surface temperature and the air temperature are substantially the same,
If the output of “snow” or “wet” is not obtained at ゜ C or less, it can be similarly determined that the moisture meter has an abnormality.

Next, if the road surface temperature rises rapidly, drops sharply around 14:00, and the air temperature and the road surface temperature do not reverse even after around 18:00, some sort of abnormalities in the thermometer There seems to be. Especially, when the road surface temperature is -3 ゜ C
If the temperature is 2 ° C. or more below, it is an abnormality of the thermometer.

When the road surface temperature changes between 0 ° C. and 1 ° C. and “snow” or “wet” is detected at the same time, if the difference between the air temperature and the road surface temperature is large, a thermometer is used. It turns out that there is something wrong. Similarly, if the thermometer changes to a constant value of 1 ° C. or less and the moisture meter outputs at the same time and the road surface temperature at this time fluctuates, it is understood that the road surface thermometer is abnormal. As for the road surface moisture, if the road surface temperature and the air temperature are kept constant, and if it is between 0 and 2 ° C, “snowfall”
Or, it appears as an output of "wet", and even if the weather changes to sunny and the road surface temperature rises, the feature is that wetting is continued for a while.

[0016] Such features may exhibit slightly different features depending on the detection location, but on a sunny day, if the road surface is illuminated by the sun during the day, almost the same changes will occur. It is known that the same change occurs even in a place that is not exposed to the sun.

According to the present invention, a characteristic of the change pattern is measured in advance for each point, a representative pattern is artificially created, and the sensor output is compared with the representative pattern to automatically determine a failed sensor. The gist of the present invention is to make predictions with an improvement in diagnostic accuracy except for a failed sensor.

[0018]

According to the above configuration, on a general road, the temperature, the road surface temperature, and the moisture condition are each generated depending on the weather of the day, and the abnormality of the sensor is diagnosed by using the characteristics. The weather can be determined at the same time.

Next, such a failure signal and the result of weather determination are transmitted to the central unit through a communication line. The central unit manually inputs the weather for the day, or collects data from an external organization such as the Meteorological Association using a communication line if no weather expert is available. The sensor failure signal is used only when the weather matches. In other words, the failure diagnosis automatically performs the same operation as the method of diagnosing whether or not there is an abnormality by manually checking the recorded data by the central control unit. As a result, reliable diagnosis is possible.

[0020]

FIG. 5 shows an example of the configuration of a road surface freezing detection device 1 according to the present invention, which will be described in conjunction with FIG. The temperature sensor 4 is a ventilation type thermometer which is installed in the air several meters from the ground and measures the atmospheric temperature by forced ventilation. The road surface temperature sensor 5 is an infrared non-contact thermometer that is installed so as to face the detection position a of the road surface in the air several meters from the ground, and measures the temperature of the detection position a in a non-contact manner. Further, the road surface moisture sensor 6 is installed in the air several meters from the ground so as to face the detection position a of the road surface, and the reflected light characteristic of the infrared ray projected toward the road surface at the detection position a depends on the presence of moisture. This is a road surface moisture meter that detects a difference and detects road surface moisture, and the road surface snow detects that the road surface becomes white due to snow and the amount of reflected light increases.

The outputs of the road surface temperature sensor 5, the air temperature sensor 4 and the road surface moisture sensor 6 are used to apply measured data to the self-diagnosis device 7 via the road surface freezing detection device 1, and these data are, for example, personal data. Diagnosis is performed by a self-diagnosis device 7 such as a computer. That is, the change patterns of the road surface temperature, the temperature, and the signal of the moisture are input to the self-diagnosis device 7 in advance, and the relation between the change of the weather and the change of the output of each sensor is stored for the representative pattern.

Next, the output detected by each sensor is received by the self-diagnosis device (personal computer) 7 from the road surface freezing detection device 1 and the change in this signal is compared with a representative pattern previously input in advance.

FIG. 6 shows a flowchart of a specific diagnosis. In a concrete implementation, a difference between a basic pattern and an input signal is obtained, and a case where the difference is small is regarded as normal, and a large difference is determined. The generated signal is recorded as an abnormal signal.

Incidentally, the basic pattern for comparison with the input signals of the air temperature and the road surface temperature is translated in parallel until the average value in the comparison target time zone coincides, and the deviation of the average temperature due to the season is corrected. Since the detection of road surface moisture is “present” or “absent”, the time zone is limited (12:00 to 16:00).

Next, if the difference from the representative pattern is small for any of the sensors, it is determined to be "normal". If the difference from the representative pattern is significantly different, the number of times the sensor is shifted is recorded with 24 hours as one time.
In addition, since the deviation width of the abnormality determination differs depending on the location, data obtained by the installation location is compared with a representative pattern classified in advance, and as a result of comparison on the same day, two or more sensors have the same tendency. If it corresponds to any of the representative patterns, it is determined that the weather is in the matched pattern.

At this time, if only one of the three sensors does not match, it is determined that the sensor is "possible a failure" and recorded as "one mismatch". Then, if such a mismatch occurs two or more times, it is determined to be “failure”. When only one of the three sensors matches, the next measurement is performed again as "indeterminable". If all three sensors match, all sensors are determined to be normal and the weather for that day is output at this time.

Generally, the pressure distribution is stable during the winter season from December to February, so that many outputs close to the representative pattern appear. However, since the pressure distribution is not stable from the end of winter to the spring, In some cases, it is difficult to make the weather like a representative pattern, and the conditions for determination cannot be established. Therefore, when a signal deviating from the representative pattern is obtained, if two of the three sensors are in accordance with the pattern, the signal is adopted as a representative pattern, and a signal that does not match is regarded as abnormal. However, if two or more patterns do not match, the weather on that day may be different from the weather on the representative pattern,
It is not used for judgment of abnormality as "undefined".

Further, when the number of deviations from the representative pattern is equal to or greater than a set number of times (for example, twice in seven days) during a set period, a diagnostic signal is output as "sensor abnormality" and a failure occurs through the transmission device. The weather determined by the sensor is sent to the center. The reason for this is that a sudden change in the weather, such as "cloudy after cloudy and occasionally rainy", may occur, and the present invention cannot cope with the change.

Next, the central unit CPU compares the past weather input from the manual input or the weather association with the weather from the road surface freezing detection device of the terminal. A prediction of "freeze" is issued except for the determined signal. At this time, if there is a condition (for example, a failure of the road surface moisture sensor) that makes freezing prediction impossible, the freezing prediction is performed according to the freezing order pattern of the entire route.

[0030]

The self-diagnosis method of the road surface freezing detection device according to the present invention is constructed as described above. Therefore, even if there is a defect in a part of the sensor in the past, the data was used as it is, so that the prediction result was not obtained. Mistakes can occur,
According to the present invention, there is a case where it is impossible to predict, but according to the present invention, it is possible to automatically self-diagnose an abnormality of a road surface freezing detection device installed in a large number of mountainous areas or places where maintenance is difficult, and perform prediction with a normal sensor. Therefore, the accuracy of the prediction can be greatly improved.

Further, according to the present invention, since the judgment result can be evaluated in comparison with the weather by an external organization or the like, the weather of the data to be compared becomes accurate and the judgment accuracy is further improved.

As described above, according to the self-diagnosis method of the road surface freezing detecting device of the present invention, (1) the freezing can be predicted except for the signal from the abnormal sensor, so that the accuracy of the prediction result is more accurate than before. . (2) Since the freezing can be predicted in consideration of the freezing order pattern of the route section, it can be easily determined whether or not the suddenly occurring freezing phenomenon is normal. The effects obtained by implementing the present invention are extremely large.

[Brief description of the drawings]

FIG. 1 is an external view showing an embodiment of a road surface freezing detection device installed on a road.

FIG. 2 is an output graph showing a typical pattern at the time of sunny due to an input from each sensor.

FIG. 3 is an output graph showing a typical pattern at the time of rain due to an input from each sensor.

FIG. 4 is an output graph showing a typical pattern at the time of snow according to an input from each sensor.

FIG. 5 is a block diagram showing one embodiment of a self-diagnosis device according to the present invention.

FIG. 6 is a flowchart illustrating a self-diagnosis method for a road surface freezing detection device according to the present invention.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Road surface freezing detection apparatus 2 Prop 3 Support arm 4 Air temperature sensor 5 Road surface temperature sensor 6 Road surface moisture sensor 7 Self-diagnosis device (personal computer) CPU Central unit

Claims (1)

(57) [Claims] 1. A thermometer installed on a pole near a road,
A road surface thermometer and a road surface moisture meter are provided.
Modeling the change pattern of the temperature and road surface temperature and the characteristics of the appearance of moisture and storing them in a storage device. Self-diagnosis of a sensor abnormality is performed by comparing the stored pattern with an input signal, and road surface freezing for determining the weather. A self-diagnosis method for a road surface freezing detection device, which is used as accurate data for road surface freezing prediction only when the weather prediction determined from the outside based on the data from the detection device coincides with the determined weather prediction.