JP2002234323A - Tire air pressure detector - Google Patents

Abstract

PROBLEM TO BE SOLVED: To detect a tire air pressure drop accurately by eliminating an influence of a slip of a driving wheel and eliminating an influence of a tread lift. SOLUTION: A wheel speed deviation value D computed according to a vehicle speed is sequentially corrected to produce a wheel speed deviation value D' presumed if an influence of a tread lift is eliminated. On the basis of the wheel speed deviation value D' for every short time from which the tread lift influence is eliminated, a regression line representing a relation between the wheel speed deviation value D' and a front-to-rear wheel speed ratio β is computed, and from the regression line, a corrected wheel speed deviation value D"AVE from which an influence of slips of the driving wheels 1c and 1d is eliminated is computed. On the basis of the corrected wheel speed deviation value D"AVE thus computed, a tire air pressure drop is detected.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a tire pressure detecting device for detecting a tire pressure condition in a vehicle.

[0002]

2. Description of the Related Art As a conventional tire air pressure detecting device, there is one disclosed in Japanese Patent Application Laid-Open No. Hei 10-100624. The tire pressure detecting device disclosed in this conventional publication will be described.

The tire pressure detecting device disclosed in the prior art detects a decrease in tire pressure based on the relationship between the wheel speed deviation value D and the front and rear wheel speed ratio β shown as follows.

[0004]

(Equation 1)

[0005]

(Equation 2)

Here, V _FR is the right front wheel speed, V _FL is the left front wheel speed, V _RR is the right rear wheel speed, and V _RL is the left rear wheel speed.

The wheel speed deviation value D is a rotation state value obtained from the wheel speeds of the four wheels, and is a variable given as a difference between the wheel speed ratios of the front and rear wheels in a diagonal relationship, for example. When the air pressure of such a tire decreases, it increases or decreases accordingly. The front and rear wheel speed ratio β is 4
This is a slip state value obtained from the wheel speed of the wheel, and represents a degree of a slip state generated in the drive wheel by the action of the driving force transmitted to the drive wheel. A smaller value indicates that the drive wheel is slipping.

When the tire air pressure of each wheel is a specified value, the wheel speed deviation value D becomes 0. However, when the tire air pressure of any of the four wheels falls below the specified value, the wheel speed deviation value D becomes zero. Since D increases or decreases, it is possible to detect a decrease in tire air pressure based on this.

However, for example, in a rear-wheel drive vehicle, when the tire pressure of the right rear wheel, which is the driving wheel, drops below a specified value, the turning radius of the right rear wheel becomes small due to the decrease of the air pressure. Since the contact area becomes larger than the wheels and the force for suppressing the slip is increased, the other drive wheels are more likely to slip, and the wheel speed deviation value D varies depending on the degree of the slip state.

Therefore, as shown in FIG. 10, the front-rear wheel speed ratio β representing the degree of the slip state using the least squares method.
A regression line is obtained by regressing the relationship between the wheel speed deviation value D and a linear function, and the wheel speed deviation value D is a value in an ideal running state in which no slip occurs (that is, a front-rear wheel speed ratio β = 1). In this way, the influence of the slip of the drive wheels is eliminated, and a decrease in tire air pressure can be accurately detected.

[0011]

However, the wheel speed deviation value D is a tread lift (a phenomenon in which the dynamic load radius of the tire increases due to centrifugal force) in addition to the slip of the driving wheel.
Therefore, when the calculation result is plotted, the result becomes a white circle in FIG. 11, and a deviation occurs from the plot of the black circle in the figure assumed when the influence of the tread lift is not given. For this reason, there is a problem that the regression accuracy in deriving the regression line is reduced, the pressure to be alerted when the tire air pressure is reduced varies, and it is impossible to accurately determine the tire air pressure reduction.

SUMMARY OF THE INVENTION In view of the foregoing, it is an object of the present invention to remove the influence of a slip of a driving wheel and the influence of a tread lift, and to accurately detect a decrease in tire air pressure.

[0013]

In order to achieve the above object, according to the first aspect of the present invention, a wheel speed detecting means (2a to 2d) for detecting a wheel speed of each wheel of a front wheel drive or rear wheel drive vehicle. 3a), and the deviation of the wheel speed between the left and right wheels caused by the turning of the vehicle is canceled.
A rotation state value calculation means (3b) for calculating a rotation state value (D) obtained by relating each wheel speed detected by the wheel speed detection means, and a rotation state value calculation means using the vehicle speed as a parameter. First rotation state value correcting means (3j) for correcting the detected rotation state value, and the degree of slip state between the driving wheel and the driven wheel based on the wheel speed detected by the wheel speed detecting means. A slip state value calculating means (3e) for calculating a slip state value (β), a rotation state value (D ') corrected by the first rotation state value correcting means, and a slip state calculated by the slip state value calculating means Regression calculating means (3h) for deriving a regression line by regressing the value to a linear function, and a first regression line based on the regression line determined by the regression calculation means.
Second rotation state value correction means (3k) for further correcting the rotation state value corrected by the rotation state value correction means;
Air pressure decrease determining means (3n) for determining a decrease in tire air pressure of each wheel based on the corrected rotational state value (D '' _AVE ) obtained by the rotational state value correcting means. And

In such a configuration, the rotation state value obtained every short time, which is determined according to the vehicle speed, is corrected to remove the influence of the tread lift. Then, a regression line indicating the relationship between the rotation state value and the slip state value is obtained based on the rotation state value for each short time after removing the influence of the tread lift, and the influence of the slip of the drive wheels is determined from the regression line. The corrected rotational state value after removal is obtained.

Therefore, the tire pressure drop can be detected based on the corrected rotation state value from which the influence of the drive wheel slip has been removed and the influence of the tread lift has been removed, and the tire pressure drop can be accurately detected. it can.

More specifically, as set forth in claim 2, the first
The rotation state value correction means obtains a variation amount of the rotation state value with respect to the vehicle speed based on a correlation between the rotation state value and the vehicle speed obtained by the rotation state value calculation means, and uses the change amount as a correction amount to determine the rotation state value. By adding the correction amount to the value, the rotation state value obtained by the rotation state value calculation means is corrected.

The correlation between the rotation state value and the vehicle speed is determined, for example, by the first rotation state value correction means, as described in claim 3, by the rotation state value data stored in the rotation state value storage means (3c). Is determined based on In such a case, before the correlation is obtained, the rotation state value calculated by the rotation state value calculation means is corrected based on a preset default value indicating the correlation between the rotation state value and the vehicle speed, and the correlation is calculated. After is determined, based on the determined correlation,
The rotation state value obtained by the rotation state value calculation means is corrected.

According to the present invention, the ideal running state value calculating means (3i) calculates an ideal state value (βid = F (A)) corresponding to a slip state value in an ideal running state without slip. The rotation state value correction means calculates the rotation state value in the ideal running state from the regression line calculated by the regression calculation means and the ideal state value calculated by the ideal running state calculation means. It is characterized by having.

As described above, the rotation state value in the ideal traveling state can be obtained from the regression line and the ideal state value, and based on the rotation state value in the ideal traveling state,
It is possible to accurately determine a decrease in tire air pressure.

According to the fifth aspect of the present invention, the vehicle speed is divided into a plurality of speed regions, and for each speed region,
After dividing the rotation state value obtained by the rotation state value calculation means and the slip state value obtained by the slip state value calculation means, for each of the divided speed regions, the rotation state value and the slip state value are converted into a linear function. Regression calculating means (3f) for performing regression to derive a regression line;
Rotation state value correction means (3h) for correcting the rotation state value obtained by the rotation state value calculation means for each vehicle speed region based on the regression line obtained by the regression calculation means, and rotation state value correction means A regression curve is obtained from the corrected rotation state value (D '' _AVE ) for each vehicle speed region, and a determination value (D '') is obtained from the corrected rotation state value for each vehicle speed region based on the regression curve. ' _AVE ), and a pneumatic pressure drop judging means (3j) for judging a drop in the tire air pressure of each wheel based on this judgment value.

As described above, even if the effect of the tread lift is removed after the correction for removing the effect of the slip, the same effect as the first aspect can be obtained. In this case, as set forth in claim 6, storage means is provided for storing the data of the rotation state value and the slip state value in a plurality of speed regions, and the time when the data is stored in the storage means is provided. A counter that counts each data is provided, and if the count value of the counter reaches a predetermined value, if it is deleted from the storage means, when the tire pressure starts to decrease, data that is too far apart Can be used to prevent the regression calculation from being performed.

Note that the reference numerals in parentheses of the above means indicate the correspondence with specific means described in the embodiments described later.

[0023]

(First Embodiment) FIG. 1 shows a schematic configuration of a tire pressure detecting device according to an embodiment of the present invention. The tire pressure detecting device will be described with reference to FIG.

The tire air pressure detecting device detects that the air pressure of one of the tires of each wheel has decreased, and when the decrease of the tire air pressure is detected, the driver is warned to that effect. . The tire pressure detection device is mounted on a front-wheel drive or rear-wheel drive vehicle. In the present embodiment, a case where the tire pressure detection device is mounted on a rear-wheel drive vehicle will be described as an example.

The tire pressure detecting device is provided for each wheel 1 of the vehicle.
Wheel speed sensors 2a, 2b, 2c, 2d as wheel speed detecting means provided corresponding to a, 1b, 1c, 1d
And an arithmetic processing unit 3 to which detection signals from the wheel speed sensors 2a to 2d are input, and an alarm unit 4 for warning a driver of a decrease in tire air pressure based on a warning signal from the arithmetic processing unit 3. It is configured.

Of the wheel speed sensors 2a to 2d, two wheel speed sensors 2a and 2b are driven wheels (front right wheel and front left wheel).
1a and 1b are detected, and the remaining 2
The two wheel speed sensors 2c and 2d detect wheel speed signals in driving wheels (right rear wheel and left rear wheel) 1c and 1d.

The arithmetic processing unit 3 is composed of a microcomputer or the like, and performs various arithmetic operations based on detection signals input from the wheel speed sensors 2a to 2d. This arithmetic processing device 3 is configured as follows.

The arithmetic processing unit 3 includes a wheel speed calculating unit 3a as wheel speed calculating means, and a wheel speed deviation value processing unit 3b.
Is provided. The wheel speed calculator 3a calculates the wheel speed of each of the wheels 1a to 1d based on detection signals (for example, pulse signals) from the wheel speed sensors 2a to 2d. The wheel speed deviation value processing unit 3b includes a wheel speed deviation value calculation unit as a rotation state value calculation unit, a first wheel speed deviation value storage unit, and a wheel speed deviation value average processing unit. Various processes relating to the wheel speed deviation value D are performed based on the calculation result in the wheel speed calculation unit 3a.

In these configurations, first, the wheel speed
Detection from the wheel speed sensors 2a to 2d by the calculation unit 3a
Calculation of the wheel speed of each wheel 1a to 1d is performed based on the signal.
Is For example, each wheel speed sensor input within a few ms
From the number of detection signals from 2a to 2d,
Wheel speed V_FL, V_FR, V_RL, V_RRIs calculated. Next, the car
When the wheel speed is calculated, the data on the wheel speed
Based on the above equation,
Is calculated. And this operation
Data relating to the result is provided in the first wheel speed deviation storage unit.
Is stored in the specified memory. Also, wheel speed deviation average
In the processing unit, the average value D of the wheel speed deviation values D _AVEIs required
You. However, this average value D_AVEIs the first wheel speed deviation value
It is not directly obtained from the data stored in the memory,
The tread lift correction processing unit described later
3e (wheel speed deviation value D ')
Desired.

The processing unit 3 is provided with a front and rear wheel speed ratio processing unit 3c. The front and rear wheel speed ratio processing unit 3c includes a front and rear wheel speed ratio calculation unit as a slip state value calculation unit, a front and rear wheel speed ratio storage unit, and a front and rear wheel speed ratio average processing unit. In the front / rear wheel speed ratio processing unit 3c, the front / rear wheel speed ratio β shown in Equation 2 is calculated by the front / rear wheel speed ratio calculation unit based on the calculation result in the wheel speed calculation unit 3a. Is stored in the front and rear wheel speed ratio storage unit, and based on the stored contents, the front and rear wheel speed ratio average processing unit obtains the average value β _AVE of the front and rear wheel speed ratio β. The average value β _AVE of the front and rear wheel speed ratio β is expressed by the following equation, and corresponds to an average of n ₀ front and rear wheel speed ratios β.

[0031]

(Equation 3)

The arithmetic processing unit 3 is provided with an average wheel speed processing unit 3d. The average wheel speed processing unit 3d includes an average wheel speed calculation unit and an average wheel speed storage unit. In the average wheel speed processing unit 3d, based on the calculation result in the wheel speed calculation unit 3a, the average wheel speed _VFL , _VFR , _VRL , and the average value of the wheel speed _VRR are calculated by the average wheel speed calculation unit. _{After the AVE} is calculated, the calculation result is stored in a memory provided in the average wheel speed storage unit. The average wheel speed V _AVE
Is, for example, each wheel 1a~1d all of the wheel speed V _FL, V
The sum of _FR , V _RL , and V _RR may be divided by 4 (the number of wheels), or the wheel speeds V _FL , V _FR , A value obtained by dividing the sum of three wheels from the smaller of V _RL and V _RR by three may be used.

The processing unit 3 is provided with a tread lift correction processing unit 3e. The tread lift correction processing unit 3e includes a correction value characteristic learning unit, a tread lift correction unit, and a second wheel speed deviation value storage unit. In the correction value characteristic learning unit, the wheel speed deviation values D stored in the first wheel speed deviation value storage unit in the wheel speed deviation value processing unit 3b are classified according to vehicle speed, and then the wheel speed deviations are calculated for each of the divided regions. The average value of the deviation values D is obtained, and a regression line or a regression line of the wheel speed deviation values D is calculated based on the locus of the average value of the wheel speed deviation values D for each region, thereby obtaining the wheel speed deviation value D with respect to the vehicle speed. Learn the variation of

In the tread lift correction section, the calculated wheel speed deviation value D is calculated based on the fluctuation of the wheel speed deviation value D with respect to the vehicle speed, and the calculated wheel speed deviation value D corresponds to a value assumed when there is no influence of the tread lift. '. This tread lift correction unit basically corrects the wheel speed deviation value D according to the learning result in the correction value characteristic learning unit.
Before the correction value characteristic learning unit learns the fluctuation of the wheel speed deviation value D with respect to the vehicle speed, the correction of the wheel speed deviation value D is performed using a preset default characteristic. This tread lift correction unit corresponds to a first rotation state value correction unit.

Then, in the second wheel speed deviation value storage unit,
The wheel speed deviation value D 'after correction by the tread lift correction unit is stored. In this way, the correction of the wheel speed deviation value D in the tread lift correction processing unit 3e is performed. The wheel speed deviation value averaging unit in the wheel speed deviation value processing unit 3b averages the wheel speed deviation values D 'thus corrected to obtain an average value D' _AVE . ing. The average value D ' _AVE of the wheel speed deviation values D' is represented by the following equation, and corresponds to an average of n ₀ wheel speed deviation values D '.

[0036]

(Equation 4)

The arithmetic processing unit 3 is provided with a slip deviation value calculation unit 3f, an ideal running state value calculation unit 3g, and a wheel speed deviation value correction processing unit 3h.

In the slip deviation value calculating section 3f, the wheel speed deviation value D 'stored in the second wheel speed deviation value storage section in the tread lift correction processing section 3e and the front and rear wheel in the front and rear wheel speed ratio processing section 3c. The slip deviation value A is calculated based on the front and rear wheel speed ratio β calculated by the speed ratio calculation unit. The slip deviation value A corresponds to a change amount (ΔD ′ / Δβ) of the wheel speed deviation value D ′ with respect to the front-rear wheel speed ratio β, and n ₀ wheel speed deviation values D ′ and the front-rear wheel speed ratio β Is calculated based on the least squares method. The slip deviation value calculator 3f corresponds to a regression calculator.

The ideal running state value calculator 3g calculates the ideal running state value βid based on the data on the calculation result in the slip deviation value calculator 3f. The ideal running state value βid corresponds to the front-rear wheel speed ratio β in an ideal running state without slip as a correction reference, and is calculated as a function of the first order or higher of the slip deviation value A. That is, the ideal running state value βid is represented by βid = F (A). For example, when it is a linear function of the slip deviation value A, it is represented by βid = 1−Coef × | A |. Here, Coef is a constant.

The wheel speed deviation value correction processing unit 3h includes a wheel speed deviation value correction unit corresponding to the second rotation state value correction means and a third wheel speed deviation value storage unit. In the wheel speed deviation value correction unit, the wheel speed deviation average value D ′ _AVE calculated by the wheel speed deviation value average processing unit in the wheel speed deviation value processing unit 3b and the front and rear wheels in the front and rear wheel speed ratio processing unit 3c. The front-rear wheel speed ratio average value β _AVE calculated by the speed ratio averaging section, the slip deviation value A calculated by the slip deviation value calculation section 3f, and the ideal running state calculated by the ideal running state value calculation section 3g. A corrected wheel speed deviation value D ″ _AVE is calculated based on the value βid. The corrected wheel speed deviation value D ″ _AVE corresponds to the wheel speed deviation value D in an ideal running state. Specifically, the corrected wheel speed deviation value D '' _AVE is obtained as in the following equation.

[0041]

(Equation 5)

Then, in the third wheel speed deviation value storage section,
The reference value D '' _AVE std of the corrected wheel speed deviation value D '' _AVE calculated by the wheel speed deviation value correction unit is stored in the memory. The reference value D ″ _AVE std is a corrected wheel speed deviation value D ″ _AVE at the time of equal pressure of the four wheels, which is a reference of the air pressure determination. Calculated from the wheel speed deviation average value D ' _AVE obtained from the speed deviation value D' and the front and rear wheel speed ratio β, the front and rear wheel speed ratio average value β _AVE , the slip deviation value A, and the ideal running state value βid. Is equivalent to

Further, the arithmetic processing unit 3 is provided with a differential pressure determination value calculation section 3i and an air pressure drop determination section 3j. In the differential pressure determination value calculation unit 3i, the reference value D ″ _AVE std stored in the third wheel speed deviation value storage unit in the wheel speed deviation value correction processing unit 3h and the value obtained by the wheel speed deviation value correction unit are obtained. A differential pressure determination value ΔD ″ _AVE is obtained based on the corrected wheel speed deviation value D ″ _AVE . This differential pressure determination value Δ
D '' _AVE, the reference value D '' _{AV E} std and the wheel speed deviation value D '' difference between the _{AVE (ΔD '' AVE = D} '' AVE std-
D ″ _AVE ), and this differential pressure determination value ΔD ″ _AVE is used for evaluating the amount of decrease in air pressure.

In the air pressure drop determining section 3j, the differential pressure determining value Δ
By comparing the absolute value | ΔD ″ _AVE | of D ″ _AVE with a preset threshold value D ″ sh,
Perform air pressure judgment. Specifically, the absolute value | ΔD ″ _AVE
If | is greater than the threshold value D ″ sh, a warning signal indicating that the tire air pressure is decreasing is sent to the warning device 4.

When the warning signal indicating that the tire air pressure is low is input, the warning device 4 turns on the warning lamp provided in the vehicle interior, for example. It warns that the air pressure has dropped.

Next, FIGS. 2 and 3 show a flowchart of the tire pressure judging process by the tire pressure detecting device having the above configuration, and FIG. 4 shows a flowchart showing details of the tread lift correction process in FIG. The details of the tire pressure determination process will be described based on these drawings.

First, in step S100, the operation count N is reset to N = 0. In step S101, as wheel speed calculation processing based on the detection signals from the wheel speed sensors 2a to 2d, the wheel speed calculation unit 3a calculates the wheel speeds V _FL , V _FR , V _RL , and V _RR of each wheel. Thereafter, the number N of times of calculating the wheel speed is incremented. This processing is performed by calculating an average value of the wheel speed every few seconds for each wheel based on a wheel speed pulse of several seconds, for example.

In step S102, as a wheel speed deviation value calculation process, a wheel speed deviation value D is calculated by a wheel speed deviation value calculation unit in the wheel speed deviation value processing unit 3b. The wheel speed deviation value D is obtained by substituting each wheel speed obtained in step S101 into the above equation (1).
Then, in step S103, the wheel speed deviation value D stored so far in the memory of the first wheel speed deviation value storage unit.
(N) is one of the wheel speed deviation values D calculated this time.
Is stored. Incidentally, D (N) is an array of the wheel speed deviation value D n ₀ pieces of, so that the wheel speed deviation value D n ₀ or stored, stores the wheel speed deviation D in locations that match the number of calculations N It has become. And n ₀ wheel speed deviation values D
Is stored, for example, when the number of calculations N is reset to 0 in a counter reset process (step S100) described later, the wheel speed deviation value D stored at a location corresponding to the number of calculations N is newly calculated. The wheel speed deviation value D is appropriately updated.

In the following step S104, the front / rear wheel speed ratio β is calculated by the front / rear wheel speed ratio calculation section in the front / rear wheel speed ratio processing section 3c as front / rear wheel speed ratio calculation processing. The front and rear wheel speed ratio β is also obtained by substituting each wheel speed obtained in step S101 into the above equation (2).
Then, in step S105, the front-rear wheel speed ratio β stored so far in the memory of the front-rear wheel speed ratio storage unit is used.
As one of (N), the front-rear wheel speed ratio β calculated this time
Is stored. Incidentally, beta (N) is an array of longitudinal wheel speed ratio beta of the n ₀ pieces of the front and rear wheel speed ratio beta n ₀ or stored, to store the longitudinal wheel speed ratio beta in locations that match the number of calculations N It has become. And n ₀ front-rear wheel speed ratios β
Is stored, as in the case of D (N) described above, the data is updated to the newly calculated front and rear wheel speed ratio β as appropriate.

In the following step S106, the average wheel speed
As the degree calculation processing, the average vehicle in the average wheel speed processing unit 3d
Each wheel speed V is calculated by the above-described method in the wheel speed calculation unit._FL, V
_FR, V_RL, V_RRAverage wheel speed V_AVEAct
Calculate. Then, in step S107, the average wheel speed is recorded.
The average wheel speed V that has been stored in the memory
_AVEAs one of (N), the average wheel speed calculated this time
V_AVE(N) is stored. Note that V_AVE(N) is n₀Pieces
The average wheel speed V _AVETo
n₀And store the average wheel speed in a location that matches the number of operations N.
Degree V_AVEIs to be remembered. And n₀Pieces
Average wheel speed V_AVEAfter is stored,
V_AVEAs in (N), the newly calculated average
Wheel speed V_AVEIt is to be updated to.

Then, in step S108, a tread lift correction process is performed. This processing is performed by a correction characteristic learning unit, a tread lift correction unit, and a second wheel speed deviation value storage unit in the tread lift correction processing unit 3e. FIG.
Shows a flowchart of the tread lift correction process,
The details of the tread lift correction process will be described with reference to FIG.

In step S201, a correction characteristic learning process is performed based on the obtained wheel speed deviation value D and the average wheel speed _VAVE . This processing is performed, for example, in Japanese Patent Laid-Open No. 10-67.
This is performed by the method described in Japanese Patent Publication No. Hereinafter, the correction characteristic learning process will be described.

First, the vehicle is divided into a plurality of speed regions in accordance with the permissible speed of the vehicle, and the average wheel speed V _AVE obtained this time is divided into a plurality of speed regions. For example, from 85 km / h to 155 km / h, 5 km / h
Each is divided into 1 to 14 speed regions, and if the current average wheel speed V _AVE is 100 km / h, it belongs to the fourth region. Thus, the obtained average wheel speed V _AVE is sequentially divided into regions.

Next, the current wheel speed deviation value D is entered into the area, and the current wheel speed deviation value D is added to the sum of the wheel speed deviation values D added to the value added before that. Then, the counter is incremented so that the number of data in the speed area to which the current wheel speed deviation value D is added can be confirmed.

Then, for all the regions, it is checked how many speed regions have stored a predetermined number (for example, 15) or more of data, and if there are four or more speed regions that satisfy the condition, for example, The average value of the wheel speed deviation values D that have been added up to that point in the region and the median value of the speed region (87.5 km in the region of 85 to 90 km / h)
/ H). Further, the maximum value and the minimum value of the average values of the wheel speed deviation values D in the respective speed regions are compared in magnitude, and when the difference is less than 0.04, for example, the wheel speed deviation values in all the speed regions are determined. The average value of D is set to 0.

After completing these operations, the regression curve (or regression line) obtained by plotting the average value of the wheel speed deviation values D in the speed range satisfying the above conditions and the intermediate value of each speed range is determined by the least squares method. Ask by. The regression equation at this time indicates the variation of the wheel speed deviation value D with respect to the vehicle speed, and the correction characteristics are obtained based on this. That is, based on the fact that the wheel speed deviation value D fluctuates with respect to the speed as in the above-described regression equation, the fluctuation of the wheel speed deviation value D due to the influence of the tread lift can be assumed from the regression equation. It is possible to learn a correction amount when removing the influence of the tread lift from the wheel speed deviation value D. In this way, the correction characteristic learning process is performed, and the regression curve (or regression line) is derived by the least squares method.
When the derivation of the relational expression between each vehicle speed and the correction amount based on the regression line is completed, processing such as setting a flag is performed.

In the following step S202, it is determined whether or not the correction characteristic learning has been completed. Specifically, the determination is made based on whether the derivation of the regression curve (or regression line) by the least square method has been completed, and the derivation of the relationship between each vehicle speed and the correction amount has been completed based on the regression line. The determination is made based on whether or not the above-mentioned flag is set.

Here, if a negative determination is made, the correction characteristic learning has not been completed, so the flow proceeds to step S203, and a default characteristic set in advance through experiments or the like is set as the characteristic used for the tread lift correction. . The default characteristic is obtained by, for example, experimentally obtaining a change in the wheel speed deviation value D with respect to the vehicle speed when the tire air pressure is reduced by one atmospheric pressure from a specified value, and using the experimental value. On the other hand, if an affirmative determination is made, the correction characteristic learning has been completed, so the process proceeds to step S204, and the correction characteristic obtained by the learning is set as a characteristic used for tread lift correction.

Thereafter, the process proceeds to step S205, where the wheel speed deviation value D is calculated based on the relational expression between the vehicle speed and the correction amount determined by the default characteristic or the learned correction characteristic.
Is calculated, and a wheel speed deviation value D ′ is calculated in which the influence of the tread lift according to the vehicle speed is removed. Specifically, a wheel speed deviation value D 'is obtained by adding a correction amount to the calculated wheel speed deviation value D. Thereby, the wheel speed deviation value D 'assumed when there is no influence of the tread lift.
Is required.

FIG. 5 shows the relationship between the wheel speed deviation values D and D 'and the front and rear wheel speed ratio β before and after the tread lift correction process is performed. In this figure, the white circles indicate the wheel speed deviation values D before processing, and the black circles indicate the wheel speed deviation values D 'after processing. As shown in this figure, even if the wheel speed deviation value D varies due to the effect of the tread lift, the wheel speed deviation value D is corrected according to the vehicle speed to remove the influence of the tread lift. The value D 'can be obtained.

After the above-described tread lift correction processing is completed, the process proceeds to step S109. In this step, it is determined whether the number of operations N is equal to or greater than n ₀ . If an affirmative determination is made, the wheel speed deviation value D 'for n ₀ wheels is obtained.
And step S assuming that the front and rear wheel speed ratio β is stored.
The process proceeds to 110, and if a negative determination is made, the process returns to step S101.

In step S110, the slip deviation value calculation section 3f calculates a slip deviation value A as a slip deviation value calculation process. That is, a regression line is derived by regressing the relationship between the front and rear wheel speed ratios β for n ₀ and the wheel speed deviation value D ′ to a linear function using the least squares method, and the slip deviation value A is calculated from the slope of the regression line. Ask for. This slip deviation value A is the front and rear wheel speed ratio β of the wheel speed deviation value D ′.
Represents a dependency on

Subsequently, in step S111, as a wheel speed deviation value averaging process, an average value D' _AVE of the wheel speed deviation values D 'is calculated by a wheel speed deviation value averaging processing unit in the wheel speed deviation value processing unit 3b. . The average value D ′ _AVE is obtained by substituting the wheel speed deviation values D for n ₀ stored in step S103 into the above equation (3).

[0064] Subsequently, in step S112, the front-rear wheel speed ratio averaging process to calculate the average value beta _AVE around wheel speed ratio beta before and after the wheel velocity ratio averaging processor 3e. The average value β _{A VE} is obtained by substituting the front-rear wheel speed ratio β for n ₀ stored in step S105 into the above _equation (4).

In step S113, as an ideal traveling state value calculation process, the ideal traveling state value βid is computed by the ideal traveling state value computing unit 3g. This ideal running state value βid is obtained from a first-order or higher-order function relating to the slip deviation value A calculated in step S110.

In the following step S114, as a wheel speed deviation value correction process, the wheel speed deviation value correction unit in the wheel speed deviation value correction processing unit 3h performs steps S110 to S11.
The corrected wheel speed deviation is obtained by substituting the slip deviation value A, the wheel speed deviation average value D ' _AVE , the front-rear wheel speed ratio average value β _AVE , and the ideal running state value βid obtained in step 3 into the above equation (5). _{Find the} value D '' _AVE .

FIG. 6 shows the corrected wheel speed deviation value D '' _AVE
When the correlation of the corrected wheel speed deviation D '' _AVE slip deviation A used in the derivation of the ideal running state values Betaid, wheel speed deviation mean D _'AVE, and the front-rear wheel speed ratio average value beta _AVE The relationships are shown, and these relationships will be specifically described.

FIG. 6 shows the correlation between the wheel speed deviation value D 'and the front-rear wheel speed ratio β when the tire pressure of one of the drive wheels 1c and 1d, here the left rear wheel, is reduced. . In this figure, open circles indicate the relationship between the calculated n ₀ wheel speed deviation values D ′ and the front and rear wheel speed ratio β, and solid circles indicate the wheel speed deviation average values D ′ _AVE and front and rear wheel speed ratio averages. The relationship with the value β _AVE is shown.

When the tire pressure of the left rear wheel, one of the drive wheels 1c and 1d, decreases, the wheel speed V _RL of the left rear wheel is reduced.
Increases, the front-rear wheel speed ratio β drops below 1 with a decrease in tire air pressure. Then, in the ideal running state, the ideal running state value βid is βid = F
(A). Therefore, as shown in step S113 of the present embodiment, based on the slip deviation value A,
An ideal running state value βid is obtained from the relationship βid = F (A).

On the other hand, as shown in step S110,
Using the least squares method, it is possible to derive a regression line obtained by regressing the relationship between the front and rear wheel speed ratios β for n ₀ and the wheel speed deviation value D ′ into a linear function.

Therefore, as shown in step S114, the derived regression line and the ideal running state value βid = F
By determining the intersection with (A), the driving wheels 1c, 1d
The wheel speed deviation value D 'in an ideal running state when the air pressure drops, that is, the corrected wheel speed deviation value D'' _AVE can be obtained.

In this manner, the corrected wheel speed deviation value D '' _AVE which is the wheel speed deviation value D 'in an ideal running state when the air pressure of the drive wheels 1c and 1d is reduced can be accurately obtained. .

Subsequently, at step S115, the reference value D ''
It is determined whether _AVE std has already been detected.
This is determined by whether or not the reference value D ″ _AVE std is stored in the memory of the third wheel speed deviation value storage unit in the wheel speed deviation value correction processing unit 3h. If the corrected wheel speed deviation value D ″ _AVE calculated this time is the first value obtained after the operation of the arithmetic processing unit 3, the reference value D ″ _AVE std is not stored. Proceeding to S116, the D ″ _AVE calculated this time is stored in the memory as the reference value D ″ _AVE std, and the process proceeds to step S1.
Return to 01. On the other hand, if the corrected wheel speed deviation value D '' _AVE calculated this time is not the one obtained first, the process proceeds to step S117.

In step S117, as a differential pressure determination value calculation process, the reference value D ″ is calculated by the differential pressure determination value calculation unit 3i.
_A differential pressure determination value ΔD ″ _AVE which is a difference between the _AVE std and the corrected wheel speed deviation value D ″ _AVE is obtained.

Then, in step S118, the differential pressure determination value
ΔD ''_AVEAbsolute value of | ΔD ″_AVE| Is preset
The magnitude is compared with the threshold value Dsh, and the absolute value | Δ
D '' _AVEIs greater than threshold value Dsh
Determine whether or not.

Thus, if an affirmative determination is made, step S
The program proceeds to 119, in which it is determined that the tire air pressure has decreased, a warning signal to that effect is sent to the alarm device 4, and if a negative determination is made, the process is terminated and the process returns to step S101. Through the above processing, it can be determined whether or not the tire air pressure at each of the wheels 1a to 1d has decreased.

In the above-described processing, the wheel speed deviation value D obtained according to the vehicle speed is successively corrected, and the wheel speed deviation value D 'assumed when the influence of the tread lift is removed is obtained. I have. That is, the wheel speed deviation value D 'is calculated every short time. Then, based on the wheel speed deviation value D 'for each short time after removing the influence of the tread lift, a regression line indicating the relationship between the wheel speed deviation value D' and the front-rear wheel speed ratio β is obtained. To obtain the corrected wheel speed deviation value D'' _{AVE from} which the influence of the slip on the drive wheels 1c and 1d is removed.

In order to eliminate the influence of the slip on the drive wheels 1c and 1d, a plurality of relatively short (for example, 2 s) wheel speed deviation values D are required, while the influence of the tread lift has a correlation with the vehicle speed. Since it is high, it is effective to correct data for a relatively short time in order to remove this.

Therefore, by correcting the data of the wheel speed deviation value D obtained every short time in advance using the vehicle speed as a parameter as in the above processing, the wheel speed deviation value D ′ from which the influence of the tread lift is removed can be obtained. In addition, the influence of the slip on the drive wheels 1c and 1d can be corrected based on the wheel speed deviation value D '. Then, the corrected wheel speed deviation value D "" which eliminates the influence of the slip of the drive wheel and the influence of the tread lift is removed.
_Since the tire pressure drop is detected based on the _AVE , the tire pressure drop can be accurately detected.

In the prior art, the wheel speed deviation value D is corrected with the front and rear wheel speed ratio β = 1 as a reference value for determining an ideal running state. Since the front and rear wheel speed ratio β does not become 1, overcorrection occurs. In such a case, the amount of change in the wheel speed deviation value D with respect to the decrease in tire air pressure differs depending on the driving wheel and the rolling wheel (follower wheel), and the pressure alerted when the tire air pressure decreases varies.

However, in the tire pressure detecting device according to the present embodiment, the slip deviation value A is obtained based on the wheel speed deviation value D and the front and rear wheel speed ratio β, and the ideal running state value βid = F is obtained from the slip deviation value A. The wheel speed deviation value D is corrected based on (A). For this reason, the corrected wheel speed deviation value D ″ _AVE that is the wheel speed deviation value D in an ideal running state when the air pressure of the drive wheels 1c and 1d is reduced can be accurately obtained without overcorrection. .

Therefore, the variation of the corrected wheel speed deviation value D ″ _AVE coincides with that of the driving wheel and the driven wheel, and it is possible to prevent variations in the alarmed pressure when the tire air pressure drops.

(Second Embodiment) In the present embodiment, the tire pressure is detected by a procedure different from that of the first embodiment. FIGS. 7 and 8 show flowcharts of the tire pressure determination process in the present embodiment, and the details of the tire pressure determination process will be described based on these drawings. However, in the tire pressure determination processing, the same parts as in the first embodiment will not be described with reference to the first embodiment. Further, in the present embodiment, the configuration of the arithmetic processing device 3 is the same as that of the first embodiment, and the description is omitted.

First, in steps S300 to S302,
Steps S100 to S102 of the first embodiment shown in FIG.
By the same processing as described above, the number of operations count N is reset,
The number of wheel speed calculations N is counted and the wheel speed deviation value D is calculated. Subsequently, the process proceeds to step S303, and the front-rear wheel speed ratio β is calculated by the same processing as step S104 of the first embodiment.

Next, the process proceeds to step S 304, where the average wheel speed is calculated as step S 106 of the first embodiment.
The average wheel speed V _AVE is calculated by the same processing as described above.
Then, in step S305, with an average wheel speed V _AVE obtained this time and stores the memory of the wheel speed deviation D of the corresponding speed range, in step S306, obtains the average wheel speed V _AVE is the corresponding velocity region currently The front and rear wheel speed ratio β is stored in the memory. The speed region corresponding to the average wheel speed V _AVE obtained this time has the same concept as the correction characteristic learning process (see step S201 in FIG. 4) in the tread lift correction process of the first embodiment. A memory is provided for each of the plurality of speed regions, and when the average wheel speed V _AVE is divided, the wheel speed deviation value D and the front-rear wheel speed ratio are stored in the memory (storage means) of each corresponding speed region as described above. Be sure to remember β.

Further, the flow advances to step S307 to perform processing for deleting old data in the memory of each speed area. Specifically, it is performed as follows. First, a counter is provided for each data stored in the memory. Then, when a certain data is stored, even if the data is not stored in the memory of the speed area thereafter, every time the data is stored in the memory of any speed area, a counter corresponding to the data is obtained. Is incremented. Thus, the time since the data was stored in the memory can be known. When the time measurement by the counter elapses a predetermined time (for example, 10 minutes), the data that has passed the predetermined time is sequentially deleted. In this way, when the tire pressure starts to decrease, regression calculation is not performed using data that is too far away.

[0087] Then, the process proceeds to step S308, the speed region in which a certain number of data N ₀ is accumulated it is determined whether the K there. If the number is equal to or more than K, the process proceeds to step S309. If the number is less than K, the process returns to step S301.

[0088] Thereafter, in step S309～S313, respectively, in each speed for each region a certain number of data N ₀ is accumulated, the slip deviation A, the average value D _AVE, longitudinal wheel speed ratio of the wheel speed deviation D beta The average value β _AVE , the ideal running state value βid, and the corrected wheel speed deviation value D ″ _AVE are calculated. These correspond to step S110 in the first embodiment.
The processing is performed in the same manner as in S114. However, in this embodiment, since the tread lift correction as in the first embodiment is not performed, each process is performed using the wheel speed deviation value D obtained in step S302, and each process is performed. The difference from the first embodiment is that the processing is performed in parallel for each speed region in which a fixed number of data N ₀ is accumulated.

After completing each of these processes, step S
Proceeding to 314, in each speed region, a determination value D ''' _AVE std at the vehicle speed Vbase which is not affected by the tread lift is obtained from the K data. FIG. 9 shows the slip deviation value A and the corrected wheel speed deviation value D "" in each speed region.
The relationship with _AVE is shown, and based on this figure, step S31
The processing in 4 will be specifically described.

For example, the speed range in which a fixed number of data N ₀ is stored is 110 km, 140 km, 160 km, 180 km.
km, the slip deviation value A in each speed region is obtained from the relationship between the wheel speed deviation value D and the front-rear wheel speed ratio β shown in FIG. Then, based on the determined slip deviation value A, the corrected wheel speed deviation value D ″ _{AV E}
Is found. At this time, the tread lift correction is not performed on each wheel speed deviation value D, but each wheel speed deviation value D
Since the difference in the effect of the tread lift on the vehicle is substantially the same for each speed region, the corrected wheel speed deviation value D ″ is independent of the effect of the tread lift for each speed region.
_AVE can be calculated.

FIG. 9B shows the relationship between the corrected wheel speed deviation value D ″ _AVE obtained for each speed band and the speed region, as indicated by a white circle in FIG. 9B. A regression curve can be derived by the least squares method based on the speed deviation value D '' _AVE . The black circles in the figure are the corrected wheel speed deviation values D ″ _AVE calculated at a speed Vbase not affected by the tread lift, for example, about 80 km.

Further, when the regression curve is obtained, the first actual
Correction method similar to step S201 of FIG. 4 shown in the embodiment
Wheel speed deviation after correction in each speed range
D '' _AVETo compensate for tread lift
Judgment value D "" at no speed Vbase_AVECalculate
You.

When the determination value D ′ ″ _AVE is obtained in this way, the process proceeds to step S315. Then, step S
In steps 315 to S317, step S of the first embodiment is performed.
117 to S119, the differential pressure determination value Δ
The calculation of D ″ ″ _AVE , the magnitude comparison between the differential pressure determination value ΔD ″ ″ _AVE and the threshold value ΔD ″ ″ sh, and the transmission of a warning signal to the alarm device 4 are performed.

As described above, after the slip deviation value A is obtained for each speed range, the determination value D ''' _AVE at a speed which is not affected by the tread lift is obtained. The order of the slip correction and the removal of the influence of the tread lift may be reversed.

(Other Embodiments) In the above-described embodiment, an example in which one embodiment of the present invention is applied to a rear-wheel drive vehicle has been described. However, the present invention may be applied to a front-wheel drive vehicle. Good. In this case, the relationship is such that the ideal running state value βid becomes greater than 1 as the tire pressure of the drive wheel decreases.

In the above description, the wheel speed deviation value D shown in Expression 1 is used as the rotation state value.
Others may be used. That is, the rotation state value is
Any value may be used as long as the wheel speed of each of the wheels 1a to 1d is related so that the deviation of the wheel speed between the left and right wheels that may occur due to the vehicle turning is canceled out. For example, there are the following.

[0097]

(Equation 6)

[0098]

(Equation 7)

[0099]

(Equation 8)

All of these relational expressions are generated due to the turning of the vehicle by taking the respective differences between the left and right front wheels and between the right and the front right wheels, where similar wheel speed deviations may occur during turning of the vehicle. The wheel speeds of the wheels 1a to 1d are associated with each other so that the deviation of the wheel speed between the left and right front wheels and between the left and right rear wheels is canceled.

As described in the above embodiment, when the system warns of a decrease in tire pressure when the wheel speed deviation value D exceeds a predetermined threshold value, the slip deviation value (slope) A Is small, the wheel speed deviation value D due to slip need not be corrected. This is a problem even if there is some error since the wheel speed deviation value D may not exceed the threshold value when the slip deviation value A is small when the rear wheels (drive wheels) are depressurized. This is because the inclination A becomes almost zero in any case when the front wheel (rolling wheel) is depressurized, so that it is not necessary to correct the wheel speed deviation value D due to slip. Therefore, by excluding the case where the slip deviation value A is small from the correction target, it is possible to exclude the case where the necessity of the correction during the rear wheel depressurization and the case where the front wheel depressurization is low from the correction target.

[0102] In the above embodiments, after the average value D _AVE of the wheel speed deviation D, Betaid the average value D _AVE
= F (A) to obtain the corrected wheel speed deviation value D ′ _AVE by projecting the curve on the curve represented by F = (A). Each of the wheel speed deviation values D is represented by βid = F (A). After projecting on a curved line, an average value thereof may be taken.

In each of the above embodiments, each time the number of operations N reaches n ₀ , the wheel speed deviation values D and the front-rear wheel speed ratio β for n ₀ data stored up to that point are calculated.
The average value DAVE and the average value βAVE are obtained, and the absolute value | ΔD ′ _AVE | of the differential pressure determination value ΔD ′ _AVE is obtained. However, in such a case, the tire air pressure determination cannot be performed while n ₀ data is accumulated. For this reason, in the wheel speed deviation value storage unit and the front and rear wheel speed ratio storage unit, the oldest stored wheel speed deviation value D and front and rear wheel speed ratio β are newly calculated wheel speed deviation value D and front and rear wheel speed ratio. The moving average is set so that the tire pressure is updated to β as appropriate, and the average value DAVE and the average value βAVE are obtained each time the updating is performed, so that the tire pressure can be determined every short time. In the embodiment, the case where the rotation state value (wheel speed deviation value D) is corrected based on the ideal traveling state value βid = F (A) has been described as an example. In the adopted one, it is also possible to perform the tread lift correction processing described in each embodiment.

[Brief description of the drawings]

FIG. 1 is a diagram showing a schematic configuration of a tire pressure detecting device according to a first embodiment of the present invention.

FIG. 2 is a flowchart of control executed by the tire pressure detecting device shown in FIG.

FIG. 3 is a flowchart of control executed by the tire pressure detecting device following FIG. 2;

FIG. 4 is a flowchart of a tread lift correction process shown in FIG. 2;

FIG. 5 shows wheel speed deviation values D before and after tread lift correction,
It is the figure which plotted D '.

6 is a diagram showing the relationship between the wheel speed deviation mean D _{AV E} before correction in the tire air pressure detecting device and the corrected wheel speed deviation D _'AVE shown in FIG.

FIG. 7 is a flowchart of control executed by a tire pressure detection device according to a second embodiment of the present invention.

FIG. 8 is a flowchart of control executed by the tire pressure detecting device following FIG. 7;

FIG. 9 is a diagram showing a relationship between a slip deviation value A and a corrected wheel speed deviation value D ′ ″ _AVE in each speed region.

FIG. 10 is a diagram showing a relationship between a wheel speed deviation value D before correction and a wheel speed deviation value D ′ after correction in a conventional tire pressure detection device.

FIG. 11 is a diagram for explaining the relationship between a regression line obtained under the influence of a tread lift and an actual regression line.

[Explanation of symbols]

1a to 1d: wheels, 2a to 2d: wheel speed sensors, 3 ...
Arithmetic processing unit, 3a: wheel speed calculating unit, 3b: wheel speed deviation value processing unit, 3c: front and rear wheel speed ratio processing unit, 3d: average wheel speed processing unit, 3e: tread lift correction processing unit,
3f: Slip deviation value calculation unit, 3g: Ideal traveling state value calculation unit, 3h: Wheel speed deviation value correction processing unit, 3j: Air pressure drop judgment unit, 4: Alarm device.

 ────────────────────────────────────────────────── ─── Continuing from the front page (71) Applicant 000000011 Aisin Seiki Co., Ltd. 2-1-1 Asahi-cho, Kariya-shi, Aichi (72) Inventor Motonori Tominaga 14 Iwatani, Shimoba-Kakucho, Nishio-shi, Aichi Japan Automotive Parts Co., Ltd. Inside the Research Institute (72) Inventor Taguchi Health 1-1-1, Showa-cho, Kariya, Aichi Prefecture Inside DENSO Corporation (72) Inventor Masaro Tabata 1-Toyota-cho, Toyota-shi, Aichi Prefecture Inside Toyota Motor Corporation (72) Invention Mori Yukio 2-1-1 Asahicho, Kariya City, Aichi Prefecture Inside Aisin Seiki Co., Ltd.

Claims (6)

[Claims] 1. Wheel speed detecting means (2a-2) for detecting the wheel speed of each wheel of a front-wheel drive or rear-wheel drive vehicle.
d, 3a) and the rotational state obtained by relating the wheel speeds detected by the wheel speed detecting means so as to cancel the deviation of the wheel speed between the left and right wheels caused by the turning of the vehicle. Rotation state value calculation means (3) for calculating the value (D)
b), first rotation state value correction means (3e) for correcting the rotation state value obtained by the rotation state value calculation means using the vehicle speed as a parameter, and a wheel speed detected by the wheel speed detection means. A slip state value calculating means (3c) for calculating a slip state value (β) depending on a degree of a slip state between the driving wheel and the driven wheel; and a rotation corrected by the first rotation state value correcting means. The regression calculating means (3f) for deriving a regression line by regressing the state value (D ') and the slip state value calculated by the slip state value calculating means to a linear function, Second rotation state value correction means (3) for further correcting the rotation state value corrected by the first rotation state value correction means based on the regression line.
h) and a pneumatic pressure drop judging means (3j) for judging a decrease in the tire air pressure of each wheel based on the corrected turning state value (D '' _AVE ) obtained by the second turning state value correcting means. A tire pressure detection device comprising: 2. The method according to claim 1, wherein the first rotation state value correction unit obtains a change amount of the rotation state value with respect to the vehicle speed based on a correlation between the rotation state value obtained by the rotation state value calculation unit and the vehicle speed. The rotation state value obtained by the rotation state value calculation means is corrected by adding the correction amount to the rotation state value using the amount of change as a correction amount. Tire pressure detector. 3. A rotation state value storage means (3b) for storing a rotation state value obtained by the rotation state value calculation means, wherein the first rotation state value correction means stores the rotation state value in the rotation state value storage means. Based on the data of the rotation state value obtained,
A correlation between the rotation state value and the vehicle speed is obtained, and before the correlation is obtained, the rotation state is determined based on a default value indicating a correlation between the rotation state value and the vehicle speed set in advance. After the rotation state value calculated by the value calculation unit is corrected, and after the correlation is calculated, the rotation state value calculated by the rotation state value calculation unit is corrected based on the correlation. The tire pressure detecting device according to claim 2, wherein: 4. An ideal state value (βid = Fid) corresponding to the slip state value in an ideal running state without slip.
(A)), and the second rotation state value correction means calculates the regression line calculated by the regression calculation means and the ideal running state value calculation means. The tire pressure detecting device according to any one of claims 1 to 3, wherein a rotation state value in an ideal running state is determined from the ideal state value. 5. A wheel speed detecting means (2a-2) for detecting a wheel speed of each wheel of a front-wheel drive or rear-wheel drive vehicle.
d, 3a) and the rotational state obtained by relating the wheel speeds detected by the wheel speed detecting means so as to cancel the deviation of the wheel speed between the left and right wheels caused by the turning of the vehicle. Rotation state value calculation means (3) for calculating the value (D)
b) a slip state value calculation means (β) that calculates a slip state value (β) depending on the degree of slip state between the drive wheel and the driven wheel based on the wheel speed detected by the wheel speed detection means. 3c) dividing the vehicle speed into a plurality of speed ranges, and classifying the rotation state value obtained by the rotation state value calculation means and the slip state value obtained by the slip state value calculation means for each speed area. After that, for each of the divided speed regions, the rotation state value and the slip state value are regressed to a linear function to derive a regression line. A rotation state value correction means (3h) for correcting a rotation state value obtained by the rotation state value calculation means for each of the vehicle speed regions based on the regression line. With a regression curve from a rotating state value correcting means the rotational state value corrected for each vehicle speed range determined (D '' _AVE), based on the regression curve, the rotation after correction for each vehicle speed range A determination value (D "" _AVE ) is obtained from the state value,
A tire pressure detection device comprising: a pressure drop determining means (3j) for determining a drop in tire pressure of each wheel based on the determination value. 6. A storage unit for storing data of the rotation state value and the slip state value in the plurality of speed regions, and a time when the data is stored in the storage unit. 6. The tire pressure detection according to claim 5, wherein a counter for counting each time is provided, and data in which the count value of the counter reaches a predetermined value is deleted from the storage means. apparatus.