JPH0248873B2 - MAIKUROPUROSETSUSAOMOCHIITATOKETSUYOCHISOCHINIOKERUHOSEIHOHO - Google Patents

Description

BACKGROUND OF THE INVENTION The present invention relates to a correction method in a freeze prediction device using a micro processor.

Predicting road surface freezing is beneficial because it allows appropriate measures to be taken in terms of vehicle operation management. Generally, prediction of road surface freezing is performed by measuring the temperature of the road surface, calculating a predicted road surface temperature after a certain period of time from the measured road surface temperature data, and comparing the predicted road surface temperature with a predetermined freezing temperature. To this end, it is necessary to understand the trends in road surface temperature decline (strictly speaking, the thermal radiation balance between the road surface and the atmosphere) and the presence or absence of road surface moisture.
However, road surface temperature changes in a complex manner due to people passing on the road, passing vehicles, other obstacles, the natural environment such as sunlight, rain, snow, hail, etc., and combinations of these. Since it fluctuates in short cycles, it is not possible to accurately predict road surface temperature without taking these disturbances into account. However, it is almost impossible to accurately detect all of the above-mentioned disturbances and thereby accurately correct the predicted road surface temperature.

Summary of the Invention The present invention aims to reduce prediction errors as much as possible in a device that predicts freezing by calculating a predicted road surface temperature after a certain period of time from a measured road surface temperature and comparing this predicted road surface temperature with a predetermined freezing temperature. Therefore, it is an object of the present invention to provide a method that can easily correct road surface temperature prediction errors caused by the above-mentioned disturbances.

In this invention, in the process of decreasing road surface temperature,
The road surface temperature during each measurement period is measured at regular intervals and the measured road surface temperature is stored in memory, and the measured road surface temperature at the time of measurement and the measured road surface temperature before the required sample time stored in memory are stored. By applying the data to a predetermined arithmetic formula, the predicted road surface temperature at the prediction point after the required prediction time from the measurement point is calculated, and in the freezing prediction device that predicts freezing based on this predicted road surface temperature, the calculated The predicted road surface temperature is stored in the above memory corresponding to each prediction point, and when each prediction point is reached, the measured road surface temperature measured at that point is compared with the predicted road surface temperature at the same point stored in the memory. The difference is calculated as a prediction error, the slope of the road surface temperature drop is measured during a predetermined period of time after the measured road surface temperature reaches a predetermined correction start temperature, and based on this slope and the calculated prediction error, The present invention is characterized in that a correction value is calculated and the predicted road surface temperature is corrected using this correction value.

According to this invention, a correction value is calculated based on the prediction error and the slope of the road surface temperature drop during a predetermined period of time, and the predicted road surface temperature is corrected using this correction value. Therefore, even if the pattern of temperature drop varies widely depending on weather conditions and other disturbance factors, freezing can be predicted almost accurately and appropriate measures can be taken for vehicle operation management. can. Further, since the correction is performed by measuring the gradient of the road surface temperature drop during a predetermined period of time, the correction process is relatively simple. Furthermore, as the road surface temperature decreases, its downward change becomes stable. Therefore, in this invention, the slope is measured and the correction value is calculated from the point when the measured road surface temperature drops to a predetermined correction start temperature, thereby reducing prediction errors as much as possible. Corrections can be made to make it smaller.

DESCRIPTION OF EMBODIMENTS Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings.

(1) Decreasing pattern of road surface temperature In general, assuming that the underground temperature is uniform and the heat radiation flow is constant, the road surface temperature T (t) after t hours on a road (a semi-infinite solid with a constant thermal constant) is , is said to be given by the following equation.

T (t) = TO−2f _O / (ρCK) ^1/2 (t/π) ^1/2 (℃)
...(1) Here, TO: Initial road surface temperature at t=0 (℃) f _O : Heat flow as boundary condition (constant value) ρ: Density C: Specific heat K: Thermal conductivity.

2f _O / (ρCK) ^1/2・π ^1/2 ≡α …(2), then equation (1) becomes T(t)=TO−α・t ^1/2 …(3), Draw a curve as shown in Figure 1. t=0
Road surface temperature TO (℃) at (minutes) and t=t1 (minutes)
If the road surface temperature T1 (°C) after (sampling time) is measured with a road surface thermometer, the coefficient α can be obtained from the following equation.

α=TO−T1/t1 ^1/2 …(4) If the road surface temperature gradually decreases as time (t) passes, the road surface temperature after t=t2 (minutes) (t2>t1) T2 (°C) can be predicted as T2=T1−α(t2 ^1/2 −t1 ^1/2 ) (5) using α determined by equation (4). t3=t2−t1 (minutes)
is the predicted time. Therefore, the predicted temperature T2 is set to a predetermined freezing temperature TF (for example, 0 to
-3°C) or lower, it is possible to issue a freezing prediction that "the road surface will freeze after the predicted time t3."

However, the trend of the decrease in road surface temperature is
It is not as simple as expressed by a formula, but varies greatly depending on weather conditions and other factors. As a result of detailed observation by the inventors, it was found that the pattern of decrease in road surface temperature, which is rich in variation, can be roughly divided into three types.

The first type, as shown in FIG. 2, is one in which the road surface temperature decreases almost linearly, and often appears when the weather is clear or sunny. Maximum road surface temperature is +10℃
The following are common. This is called a linear descent type.
In the second type, as shown in FIG. 3, the road surface temperature decreases almost exponentially, which often occurs when it is cloudy or rainy. The maximum road surface temperature is often over +10℃. This is called an exponential descent type. The third type, as shown in FIG. 4, is one in which the road surface temperature drops rapidly, and often occurs during light snow. The maximum road surface temperature is often below +10℃. This is called a steep descent type. The solid line graphs shown in FIGS. 2 to 4 show an ideal transition with no fluctuations, but in reality, as shown by the chain lines, fluctuations occur at short intervals.

The approximation equation shown in equation (1) or equation (3) fits the second exponential descent type relatively well, but
If applied to other types of transitions, the error may become large. The present invention provides a method for correcting the predicted temperature calculated using the above-mentioned equation (1) and equation (5) derived therefrom, according to a temperature drop pattern.

(2) Configuration of the freezing prediction device Figure 5 shows how each measuring device is installed on the road. A pillar 2 is erected on one side of the road, and a horizontal support arm 3 extending above the road surface 1 is attached to the upper end of the pillar 2. A light projector 4, a light receiver 6, and a road surface thermometer 7 are attached to the support column 2 in this order from the top to the bottom. Further, a light receiver 5 is attached to the support arm 3. From floodlight 4 to road surface 1
Infrared light is projected toward the road surface 1, and the specularly reflected light from the road surface 1 is received by the light receiver 5, and the diffusely reflected light is received by the light receiver 6. A radiation thermometer is used as the road surface thermometer 7, and the road surface temperature is measured in a non-contact manner.

FIG. 6 shows the configuration of the main parts of the freeze prediction device. The output signals of the light receivers 5, 6 and the road surface thermometer 7 are converted into digital data and read by a microprocessor (hereinafter referred to as CPU) 10 via an input/output interface 11. Further, the projector 4 is controlled by the CPU 10 via the input/output interface 11. The clock pulses output from the clock pulse generator 13 are counted by a counter 14, and the counter 14 sends a 1-minute pulse and a 5-minute pulse to the CPU 10 every 1 minute and every 5 minutes, respectively.

The CPU 10 includes a register 12 and a data memory (hereinafter referred to as RAM) 15. The register 12 temporarily stores the road surface temperature measured every minute obtained from the road surface thermometer 7. RAM
15, an area M1 used as a flag for storing the road surface condition and a flag indicating that the measured temperature has reached the correction start temperature TS, an area M2 for storing the freezing temperature TF, an area M3 for storing the correction start temperature TS, Area M4 that stores measured road surface temperatures every minute, area M5 that stores average measured road surface temperatures every 5 minutes, predicted road surface temperature RT after 30 minutes, actual measured road surface temperature T at the same time, and measurement error. An area M6 for storing ΔT is provided. The 1-minute data storage area M4 has five storage locations, and the 5-minute data storage area M5 has four storage locations. Also memory area M6
, a large number of storage locations are provided for each temperature RT, T and measurement error ΔT.

The output level of the light receivers 5 and 6 changes depending on the road surface condition, that is, whether the road surface 1 is dry, wet, or covered with snow. The output level of the light receiver 5 increases in the order of dryness, snowfall, and wetness. Further, the output level of the light receiver 6 increases in the order of wetness, dryness, and snowfall. The CPU 10 discriminates the output signal levels of both of these light receivers 5 and 6, and determines whether the road surface is in a state where there is a possibility of freezing based on the combination of the discrimination results, that is, whether there is moisture on the road surface (whether it is wet or snowy). Determine.
Then, if there is a possibility of freezing, area M1
A flag is placed in the memory location of the road surface condition.

(3) Prediction of road surface temperature As shown in FIGS. 2 to 4 and enlarged in FIG. 7, road surface temperature fluctuates in short cycles as it changes from a macroscopic perspective. Therefore, we measure the road surface temperature every 1 minute, and every 5 minutes, we remove the maximum and minimum of the 5 1-minute data and calculate the arithmetic average of the remaining 3 1-minute data. Get data. Counter 14 outputs 1 minute pulse and 5
A minute pulse is being sent to the CPU 10. CPU10
reads the output of the road surface thermometer 7 every time a one-minute pulse is input, temporarily stores it in the register 12, then transfers it to the storage area M4 and stores it in each storage location in sequence. When 5 minutes have passed, the 1-minute data Ta to Te (see Figure 7) for the previous 5 minutes are stored in area M4.
It will be stored inside. When a 5-minute pulse is input, the CPU 10 outputs the maximum value Tc and the minimum value Te of these data.
The 5-minute data TO is calculated by taking the arithmetic average of the three 1-minute data Ta, Tb, and Td excluding . Then, this 5-minute data TO is stored in a predetermined storage location within the storage area M5. For the next 5 minutes, 5-minute data T5 is calculated in the same way and stored in the storage area M5. Similarly, 5-minute data T10 and T15 are sequentially calculated every 5 minutes and stored in the storage area M5.
Store within 5.

In this example, the sample time t1 is set to 15 minutes, and the prediction time t3 is set to 30 minutes. When 15 minutes have passed from the time when the first 5-minute data TO was measured (this time is the starting point of measurement),
The CPU 10 calculates the above coefficient α using the 5-minute data T15 at this point and the first 5-minute data TO, and calculates the predicted road surface temperature RT45 30 minutes after this point (45 minutes from the measurement start point). . Coefficient α
is calculated using the following formula.

α=TO−T15/15 ^1/2 …(6) Also, the predicted road surface temperature RT45 is calculated from RT45=T15−α(45 ^1/2 −15 ^1/2 ) …(7) using the above α. . The predicted road surface temperature RT45 calculated in this way is stored in a predetermined storage location in the storage area M6.

The predicted road surface temperature RT is calculated every 5 minutes.
The calculated predicted road surface temperatures (RT50, etc.) are sequentially stored in area M6.

As time passes, the contents of storage area M4 are completely rewritten every 5 minutes, and the contents of storage area M4 are completely rewritten every 5 minutes.
The contents will be updated sequentially starting from the oldest. It is preferable that the coefficient α is determined based on the latest 5-minute data each time the predicted road surface temperature is calculated.

When 45 minutes have passed from the starting point of measurement, CPU1
0 calculates the 5-minute data T45 at this point, so this data T45 is stored in the storage location corresponding to the predicted road surface temperature RT45, and the predicted road surface temperature
The difference between RT45 and this 5-minute data T45, ΔT45=RT45−T45 (8), is calculated and similarly stored in the corresponding storage location in area M6. As shown in FIG. 8, this difference ΔT45 is the difference between the predicted road surface temperature and the actual measured road surface temperature, and is a so-called prediction error. This prediction error calculation is also executed every 5 minutes, and the prediction error (ΔT50, etc.) calculated each time is stored in a predetermined storage location in area M6.

(4) Correction of predicted road surface temperature and freezing prediction As mentioned above, there are three types of patterns of decrease in road surface temperature. However, since the predicted road surface temperature is calculated based on Equation (1), if the temperature drop pattern is of the second type (Figure 3), it may be approximated relatively well. ,
A prediction error occurs when the temperature transition is other than the second type.

In this invention, this prediction error is corrected when the measured road surface temperature (5-minute data) is a predetermined correction start temperature TS.
Start from the moment you become This correction start temperature
TS is variable in the range of approximately 0 to +6°C, is set to +2.5°C, for example, and is stored in advance in area M3. Correction start temperature TS +
The temperature is selected to be around 2.5℃ because the drop in road surface temperature is stable at temperatures lower than this.

Furthermore, the freezing temperature TF is stored in advance in the area M2. This freezing temperature TF is also variable and is set in the range of approximately 0 to -3°C.

The CPU 10 calculates the predicted road surface temperature RT at each point in time.
A freezing prediction signal is output when satisfies RT−TA≦TF (9). Here, TA is a correction value, which is given by the following equation, where ΔTA is the arithmetic mean value of the above prediction errors and β is the coefficient.

TA=β・ΔTA (10) The coefficient β is selected to be 2, 1, or 1/2 depending on the gradient of the measured temperature. In other words, as shown in Figure 9, when the measured road surface temperature becomes lower than the correction start temperature TS, it drops by 0.5℃ or more in the first 5 minutes (corresponding to type 3 above (Figure 4)). ), β=2. If the temperature drop in the first 10 minutes is in the range of 0.5 to 1.0℃ (type 1 (type 2)
If the temperature is less than 0.5°C (corresponding to the second type (Fig. 3)), β = 1/2.

When the 5-minute data T falls below the correction start temperature TS, the CPU 10 sets a correction start flag for area M1, calculates the predicted road surface temperature RT and the prediction error ΔT as described above, and adds the prediction error ΔT. Calculate the average ΔTA and find the correction value TA.
Since the gradient is unknown for the first 5 minutes, β=
Set it to 1/2. Then, if the 5-minute data T of the road surface temperature drops by 0.5° C. or more after 5 minutes, β is set to 2. If the temperature drop is less than 0.5℃, maintain β = 1/2. 10 minutes after the correction start flag is set, the 5-minute data T becomes 1.0 based on the start temperature TS.
If it has fallen by more than ℃, β = 2, and 0.5~
If the temperature drops within the range of 1.0℃, β = 1, and 0.5
If the temperature is less than ℃, β = 1/2. Even after the correction start flag is set, the predicted road surface temperature is displayed every 5 minutes.
It goes without saying that RT and prediction error ΔT are calculated. And the coefficient β determined as above
When the predicted road surface temperature RT satisfies Equation (9) using , and the road surface condition flag is set, a freezing prediction signal is output.

In the above example, the coefficient β is 3 of 1/2, 1, and 2.
The type is set and the value of β is determined according to the temperature drop for 5 minutes and 10 minutes, respectively, but the coefficient β
There are various ways to determine the value of and how to determine it (for example, the value of the temperature drop). In short, the value of β may be determined according to the gradient of the road surface temperature drop.
Further, the sample time t1 and the prediction time t3 can also be arbitrarily determined.

[Brief explanation of drawings]

Figure 1 is a graph to explain the road surface temperature prediction principle, Figures 2 to 4 are graphs showing the change pattern of road surface temperature, Figure 5 is a configuration diagram showing the installation status of various detectors, Figure 6 is a block diagram showing the electrical configuration of the freezing prediction device, Fig. 7 is a graph showing the method of obtaining measured temperature data, Fig. 8 is a graph showing prediction error, and Fig. 9 is a graph showing correction by temperature drop gradient. It is a graph showing how the coefficient of the value is obtained. 7...Road surface temperature meter, 10...Microprocessor,
15...Data memory.

Claims (1)

[Claims] 1. In the process of the road surface temperature decreasing, the road surface temperature in each measurement period is measured at regular intervals, the measured road surface temperature is stored in a memory, and the measured road surface temperature at the time of measurement is stored. By applying the measured road surface temperature stored in the memory before the required sample time as data to a predetermined calculation formula, the predicted road surface temperature at the predicted point in time after the required predicted time from the measured point in time is calculated. In a freezing prediction device that predicts freezing based on predicted road surface temperature, the calculated predicted road surface temperature is stored in the above memory corresponding to each prediction time, and when each prediction time is reached, the measurements taken at that time are stored. The difference between the road surface temperature and the predicted road surface temperature at the same point in time stored in memory is calculated as the prediction error, and the road surface temperature decreases within a predetermined period of time from when the measured road surface temperature reaches a predetermined correction start temperature. correction value in a freezing prediction device using a microprocessor, characterized in that a correction value is calculated based on the slope and the calculated prediction error, and the predicted road surface temperature is corrected using the correction value. Method.