JPH06229772A - Integration sensing unit - Google Patents

Abstract

PURPOSE:To enhance reliability by allowing the correction of error contained in the differentiated amount of measurements and reducing the error which is otherwise accumulated through integration. CONSTITUTION:The integration sensing unit comprises a first means, a second means, means 3 for operating the difference of output between the first and second means, and means 5 for operating the variation rate of the difference of output and estimating the error of the first means 1. The first means 1 comprises a differentiated amount measuring means 6, means 7 for correcting the error based on the amount of error estimated by an error estimating means 5, and means 8 for time integrating the signal. The error of a physical amount 9 increases through integration. In order to prevent the increase of error, the means 3 produces a difference with reference to the second means 2 and the means 5 calculates the variation rate thereof and estimates the error, and then the means 7 corrects the error of the differentiated amount of measurements.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an integral type sensing device which measures a differential amount and integrates it to calculate a required amount.

[0002]

2. Description of the Related Art Integral type sensing devices for measuring a required amount by measuring a differential amount and integrating it are used in various fields.

The prior art of the integral type sensing device will be described below by taking an on-vehicle navigation device as an example.

Conventionally, as a method for detecting the position of a vehicle traveling at an arbitrary point in a road traffic network, a distance sensor, a direction sensor, and a processing device for performing necessary processing on output signals from both sensors. A dead reckoning navigation system has been proposed that obtains the current position data of the vehicle by integrating the distance change amount and the direction change amount that occur with the traveling of the vehicle, but the distance sensor and the direction sensor necessarily have them. There is a problem that the present error is accumulated with the traveling distance, and the error included in the obtained current position data is also accumulated.

In particular, in the direction sensor, a geomagnetic sensor capable of detecting the absolute direction of the vehicle, and
The traveling direction of the vehicle is detected by using it in combination with a gyro sensor capable of detecting a change in the relative direction of the vehicle, but each has its own error characteristics.

The geomagnetic sensor produces short-term noise in a place where the magnetic field environment is bad due to the influence of high-voltage lines and buildings, but outputs a correct azimuth in the long term. On the other hand, the gyro sensor is completely unaffected by the magnetic field environment,
In the short term, the correct relative azimuth change is output, but in the long term, the azimuth error accumulates with the passage of time due to the influence of drift (variation of angular velocity bias).

As a method for solving such a problem, there is a method of filtering a sensor signal as disclosed in Japanese Patent Laid-Open No. 61-258112, which will be described with reference to FIG. The differential odometer 20 is a sensor that detects the vehicle rotation angle from the rotation difference between the left and right wheels, and has the same error characteristics as a gyro sensor. The compass 22 is an absolute azimuth sensor represented by a geomagnetic sensor.
Reference numeral 1 denotes a digital road database installed in a vehicle. When the current position is specified on the road, the road direction can be output as the vehicle direction like the sensor.

The difference between the azimuth from the compass 22 or the map 21 and the azimuth from the differential odometer 20 is calculated, and a low pass filter is applied to the difference to remove high frequency noise contained in the compass 22 or the like (geomagnetic sensor). By subtracting the signal from the output of the differential odometer 20, the bias error included in the differential odometer 20 and the like (gyro sensor) is removed. Therefore, the finally output azimuth signal is a signal in which the error component of each sensor is corrected.

There is also an example of Japanese Patent Laid-Open No. 3-188316 in which a Kalman filter is applied by a similar filtering method.

As another solution, on the assumption that the vehicle travels on the road, the current position data obtained based on the dead reckoning is compared with the road traffic network data stored in the memory in advance. Then, the deviation amount of the current position data from the road is calculated as a cumulative error, and the current position data is corrected by the cumulative error so that the current position data matches the road data. 63-
A map matching method as disclosed in Japanese Patent No. 148115 has been proposed.

[0011]

In the case of the integral type sensing device which measures the differential amount and calculates the required amount by integrating the differential amount as in the above example, the error included in the differential amount is the integration operation. Therefore, even if the error is very small, a large error may occur.

In addition, in the case of the above-mentioned on-vehicle navigation device, in the filtering method, for example, when traveling on a continuous structure such as a highway, a steady error occurs in the geomagnetic sensor output. Therefore, it is difficult to accurately correct the bias of the gyro sensor based on this.

Further, in the map matching method, the final display shown to the driver is only corrected, and it is not a fundamental solution to the occurrence of sensor error.

That is, for example, when a rate sensor such as a gyro is used as the azimuth sensor, if the measured angular velocity includes a bias error, the angular error increases with time. Even if the final angle display is corrected, a similar error will occur immediately after that.

Further, for example, when a vehicle speed sensor for periodically measuring the number of rotations of a wheel is used as the distance sensor, a scale factor for calculating the vehicle speed from the number of rotations of the wheel is changed due to a change in tire air pressure or the like, resulting in a distance error. Increases as the mileage increases. Also in this case, even if the position is corrected by the map matching method, the same error will occur immediately thereafter.

As described above, this kind of error is not limited to the in-vehicle navigation device, but in the method of measuring the differential amount of the physical quantity to be sensed and then integrating it to calculate the required quantity. It is a common issue.

The present invention has been made in view of the above problems, and it is an object of the present invention to provide a more accurate integral sensing device by reducing the error accumulated by integration as much as possible.

[0018]

In order to achieve the above object, there is provided a first means for obtaining a physical quantity to be sensed, and the first means comprises a measuring means for measuring a differential quantity of the physical quantity, An integrating-type sensing device having an integrating unit that integrates the differential amount;
Means for estimating the error of the measured differential quantity in the first means from the rate of change of the difference between the physical quantities obtained by the first and second means, and the error estimating means based on the error. And means for correcting the measured differential amount in the first means.

The rate of change of the physical quantity difference may be a rate of change over time.

The error may be a bias error of the measured differential amount in the first means.

The change of the difference of the physical quantities with respect to time may be represented by a regression line, and the slope of the regression line may be used as a constant bias error of the measured differential amount in the first means.

The physical quantity is a traveling azimuth angle of the vehicle, the first means is a traveling azimuth detection by a gyro sensor, and the second means is a corresponding road stored in a road map memory mounted on the vehicle. The azimuth angle may be output as the traveling azimuth angle.

The second means may be a GPS device that detects the position and direction on the earth based on satellite signals.

The second means may detect the azimuth by a geomagnetic sensor that detects the earth's magnetic field.

The physical quantity is the traveling distance of the vehicle, the first means periodically measures the vehicle speed to detect the traveling distance, and the second means is stored in the road map memory mounted on the vehicle. Detected the corresponding road length as the traveled distance,
The scale factor error of the measured differential amount in the first means is calculated from the rate of change of the difference between the physical quantities obtained by the first means and the second means according to the travel distance,
The measurement differential amount in the first means may be corrected based on the scale factor error.

The second means may be a traveling distance detection by a GPS device that detects the position and direction on the earth based on a satellite signal.

The physical quantity is the traveling speed of the vehicle, the first means detects the traveling speed by an acceleration sensor fixed to the vehicle body, and the second means periodically measures the rotational speed of the wheels. It may be one that detects the traveling speed.

[0028]

With the above integral type sensing device, the differential quantity, which is the rate of change of the physical quantity to be sensed, is measured,
In an integral sensing device including a first means for obtaining the physical quantity by integrating a signal representing the differential quantity, a second means for obtaining the physical quantity, which is different from the first means, is provided. The error of the measured differential quantity in the first means is estimated from the change rate of the difference between the physical quantities obtained by the first and second means, and the measured differential quantity in the first means is corrected based on the error. be able to.

That is, when the rate of change of the physical quantity difference is taken as the rate of change with time, the bias error of the measured differential quantity can be estimated by the rate of change with time. For example, when the change of the physical quantity difference with respect to time is represented by a regression line, a constant slope of the regression line can be regarded as a constant bias error of the measured differential amount in the first means.

The physical quantity is the traveling azimuth of the vehicle, the first means detects the traveling azimuth by a gyro sensor fixed to the vehicle body, and the second means is stored in a road map memory mounted on the vehicle. When the corresponding road azimuth is used as the traveling azimuth, or the traveling azimuth is detected by a GPS device that detects the absolute position and absolute azimuth on the earth based on satellite signals, or by the geomagnetic sensor that detects the earth's magnetic field. When the azimuth is detected, the bias error in the angular velocity of the traveling azimuth can be calculated, and the traveling azimuth with high accuracy can be obtained by correcting the error, as in the above.

The physical quantity is used as the traveling distance of the vehicle, the first means measures the vehicle speed to detect the traveling distance, and the second means corresponds to the stored distance in the road map memory mounted on the vehicle. Even when the road length is used as the travel distance, or when the travel distance is detected by the GPS device that detects the absolute position and absolute direction on the earth based on the satellite signal, the first means and the second means are used. The scale factor error of the measured differential amount in the first means is calculated from the obtained change rate of the difference between the respective physical quantities according to the travel distance, and the measured differential amount in the first means is corrected based on the scale factor error. It is possible.

Even when the physical quantity is the traveling speed of the vehicle, the first means detects the traveling speed with an acceleration sensor solidly mounted on the vehicle body, and the second means measures the vehicle speed to detect the traveling speed. It is possible to achieve exactly the same accuracy improvement as described above.

[0033]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENT An embodiment of an apparatus according to the present invention will be described below with reference to the drawings. First, the principle of the device according to the present invention will be described with reference to FIGS.

FIG. 1 is a block diagram of an integral type sensing device to which the present invention is applied. This device comprises a first means 1,
The second means 2, the output difference calculation means 3 for calculating the output difference between the first means and the second means, the change rate of the output difference are calculated, and the error of the first means 1 is calculated by the change rate. Error estimation means 5 for estimating

The first means 1 is a differential quantity measuring means 6 for measuring a time differential quantity of a target physical quantity and an error estimating means 5.
Error correction means 7 for correcting an error based on the error estimation amount of
It has a differential amount measuring means 6 and a time integration means 8 for time-integrating the signal processed by the error correction means 7. By the above process, the target physical quantity 9 is obtained.

That is, the first means 1 obtains the target physical quantity 9 by measuring the differential quantity by the differential quantity measuring means 6 and integrating it by the time integrating means 8, so that an error occurs in the measured differential quantity. The error of the target physical quantity 9 gradually increases due to integration. Therefore, the difference is calculated by the output difference calculating means 3 with another second means 2 as a reference, and the rate of change is calculated by the error estimating means 4 to estimate the error, and then the error correcting means 7 calculates the measured differential amount. By correcting the error, the accuracy of the final target physical quantity 9 is improved.

FIG. 3 shows an error (noise) necessary for calculating the Kalman filter gain in the method of calculating the physical quantity of interest by subjecting the signals of the first means 31 and the second means 2 to Kalman filtering 13. It is a block diagram which applied this invention to the means to measure a component.

Similar to FIG. 1, the error (noise) component in the first means is measured based on the rate of change calculated by the error measuring means 10, and the Kalman filter gain calculating means 11 calculates the Kalman filter gain. By subjecting the target physical quantity obtained from the means to a weighting process based on the Kalman filter gain, the target physical quantity can be estimated by the required physical quantity estimating means 12.

Next, the Kalman filter processing will be specifically described.

First, the equation of state established by modeling the target system is as follows:

..

## EQU1 ## X (k + 1) = AX + BU (K) + v (k) where X (k) is a state vector, which is an estimated amount at time k including the required physical amount. A is the state transition matrix, B
Is the driving matrix, U (k) is blunt noise, and v (k) is system noise.

An observation equation showing the relationship between the observed amount from each sensor and the above state amount is as follows.

[0043]

## EQU00002 ## Y (k) = f (x (k)) + W (k) where Y (K) is an observation vector, f is a functional expression expressing the relationship with the state quantity, and W (k) is an observation noise. Is.

Kalman filtering is applied to the above model. Optimal estimate X (k |
k) is calculated by the following recurrence formula. If the observation equation is non-linear, the extended Kalman filter algorithm is used.

[0045]

[Equation 3] X (k + 1 | k) = AX (k | k) + BU (k)

[0046]

[Formula 4] X (k | k) = X (k | k-1) + K (k) [Y (k) -f (X (k | k-1))]

[0047]

[Equation 5] P (k + 1 | k) = AP (k | k) A '+ V (k)

[0048]

[Equation 6] P (k | k) = P (k | k-1) -K (k) C (k) P (k | k-1)

[0049]

[Equation 7] K (k) = P (k | k-1) C '(k) [C (k) P (k | k-1) C' (k) + W (k)] ~ ¹ where The subscript (k | k) is the optimum value by the Kalman filter based on the observation value at time k, and (k | k-1) is the value at time k-1 predicted by the state equation from the value at time K-1. Is. Further, V (k) and W (k) are the covariance matrices of system noise and observation noise, respectively.

P is a covariance matrix of the estimation error and serves as a measure of the accuracy of estimation by the Kalman filter. k is a Kalman filter gain, which is calculated when the observed quantity is obtained, and is a quantity that determines whether to weight the equation of state or the observed value by comparing the statistics of the system noise and the observed noise. Equation 4 is an equation for calculating the optimum estimated value. Time k-
According to the state equation using the optimal estimate in 1, time k
Predict the value at. The optimal estimated value is the internal division point of this predicted value and the Kalman filter gain k of the observed value at time k. The matrix C (k) is used to linearize the observation equation.
It is the Jacobian of f (X (k | k-1)).

In the above Kalman filter processing, the component corresponding to V (k), W (k) or P is appropriately changed based on the magnitude of the noise component obtained from the error measuring means 10. As a result, it is possible to calculate the Kalman filter gain K for performing optimal weighting under the conditions.

Next, the first embodiment will be described.

FIG. 4 is an example in which the configuration of FIG. 1 is applied to detection of a flight distance in an inertial navigation system of an airplane. The first means 45 has a speed sensor 14,
Measure at. The second means 402 has a position detecting means 16 by radio navigation such as Loran C. Based on the flight distance calculated from the position output of the second means 402,
The bias error estimating means 42 examines the time change of the flight distance due to the output of the speed sensor 14. Further, the bias error estimation means 42 calculates the time change rate at that time by statistical processing such as regression analysis described later, regards it as the bias error of the speed sensor, and the bias error correction means 41 corrects the bias error of the measured speed. To do. Therefore, the accuracy of the finally output flight distance 15 can be improved by the amount by which the bias error is corrected.

Next, the second embodiment will be described.

FIG. 14 is a block diagram of a vehicle navigation system to which the present invention is applied, showing the traveling distance of a wheel which is composed of a photoelectric or electromagnetic sensor which outputs a pulse signal proportional to the rotational speed of the wheel. A distance sensor 601 that generates a signal in accordance with the direction, a geomagnetic sensor 602 that outputs a signal in the absolute direction corresponding to the traveling direction of the vehicle, and a signal that is proportional to the direction or the amount of change in the direction according to the traveling direction of the vehicle. It has a rate type gyro sensor 603, and these outputs are configured to be input to the controller 604.

The controller 604 has the functions of the error correction means 7 and the time integration means 8 of the means 1 of 1 in FIG. 1, and further has the functions of the output difference calculation means 3 and the error estimation means 5. .

On the other hand, the controller 604 counts the number of pulse signals from the distance sensor 601 to detect the traveling distance of the vehicle, and the traveling direction of the vehicle based on the direction signals output from the geomagnetic sensor 602 and the gyro sensor 603. Is detected, and the position on the two-dimensional coordinates of each unit traveled distance of the vehicle is calculated according to the detection results.

A reference numeral 605 is a CRT for renewing and displaying the current position of the vehicle on the basis of the position data on the two-dimensional coordinate which is changed by the controller 604.
The display device includes a display device and a liquid crystal display device.

FIG. 5 is an example in which the configuration of FIG. 1 is applied to the detection of the traveling direction in a vehicle navigation system. First
The means 501 has the gyro sensor 40, and the angular velocity is measured by the gyro sensor 40. The second means 502 is
A map 21, a geomagnetic sensor 22 similar to the conventional example of FIG.
A GPS device 45 that detects an absolute position and an absolute azimuth on the earth based on a satellite signal is combined, and an address selector (ADS SEL) 23 selects the most accurate sensor according to the situation and outputs the azimuth. It works. Based on the azimuth output of the second means 502,
The bias error estimating means 42 examines the time change of the azimuth due to the output of the gyro sensor. The rate of change over time at that time is calculated by statistical processing such as regression analysis described later, and is regarded as the bias error of the gyro sensor, and the bias error correction means 41 corrects the bias error of the measured angular velocity. Therefore, the traveling direction 44 finally output can be improved in accuracy by the amount by which the bias error is corrected.

Next, a third embodiment will be described.

FIG. 6 is an example in which the configuration of FIG. 3 is applied to the detection of the traveling direction in a vehicle navigation system. Figure 5
Similarly, the first means 601 has a gyro sensor 40,
The angular velocity is measured by the gyro sensor 40. The second means 502 is three combinations of the map 21, the geomagnetic sensor 22 and the GPS device 545 for detecting the absolute position and the absolute azimuth on the earth based on the satellite signal, which are the same as those in the conventional example of FIG.
With the DS SEL 23, the most accurate sensor is selected according to the situation and the azimuth is output.

Based on the rate of change calculated by the bias error correction means 642, the error (noise) component in the first means 601 is measured, and then the Kalman filter gain calculation means 611 calculates the Kalman filter gain.
And the traveling direction obtained from the second means 601 and 502,
By the weighting processing based on the Kalman filter gain by the required physical quantity estimating means 612, the traveling direction 44
Can be estimated.

Next, a fourth embodiment will be described.

FIG. 7 is an example in which the configuration of FIG. 1 is applied to the detection of the traveling distance in a vehicle navigation system. First
The means 701 has a vehicle speed measuring means 750 that measures the vehicle speed equivalent amount by periodically counting the number of rotations of the wheels, and calculates the traveling distance by time integration by the time integration means 78. The second means 702 has a mechanism for selecting either the map data 21 or the GPS device 545 with high accuracy by the ADS SEL 723 and outputting the traveling distance. Based on the distance output of the second means 702, of the distance output of the first means 701,
The scale error estimating means 75 measures the change with the increase of the traveling distance.
Check by 2. The distance change rate (scale factor error) α at that time is calculated by statistical processing such as regression analysis described later, and is regarded as the scale factor error of the first means 1. And by the following number 1

[0065]

[Formula 1] Vehicle speed = Vehicle speed equivalent amount × (1 + α) Scale error for vehicle speed calculation is corrected. Therefore, the accuracy of the finally output mileage 54 can be improved by the amount by which the scale error is corrected.

Next, the fifth embodiment will be described.

FIG. 8 shows an example in which the configuration of FIG. 1 is applied to the traveling speed detection in a vehicle navigation system. First
The acceleration sensor 860 is applied to the means 801 and the acceleration is measured by the acceleration sensor 860. Second means 802
Is a combination of a wheel speed sensor 62 that counts the number of rotations of wheels and calculates a vehicle speed and a GPS device 545 that detects the speed of a moving object on the earth based on a satellite signal.
The SEL 823 is configured to select the most accurate sensor according to the situation and output the vehicle speed. Based on the vehicle speed output of the second means 802, the bias error estimating means 8 is used to determine the time change of the vehicle speed due to the acceleration sensor output.
Check at 42. The rate of change over time at that time is calculated by statistical processing such as regression analysis described later, which is regarded as the bias error of the acceleration sensor, and the bias error correction means 841 corrects the bias error of the measured acceleration. Therefore, the accuracy of the finally output traveling speed 61 can be improved by the amount by which the bias error is corrected.

The above example is only one example, and the present invention can be commonly applied to integral type sensing devices.

Next, the operation of the present invention will be described with reference to the flow charts of FIGS. 9 to 13 for the case of improving the accuracy of the traveling direction as shown in FIG.

When the system is started, initial processing is first performed (step 100), and when the current position is set by key input (step 102), the current position of the vehicle and a map of the surrounding area are displayed on the display device 605. (Step 104).

Next, in step 106, step 1
Interrupts 08 and below are permitted, and the main loop below is entered.

In this main loop, the running amount for each unit distance is integrated and added to update the vehicle current position and the map, and the running vector integration interrupt process (step 10).
8), the azimuth calculation interrupt process (step 110) and the bias error calculation interrupt process (step 112) move the current position (Yes in step 114),
The current vehicle position and the surrounding map are updated (step 116).

FIG. 10 is a flowchart showing the processing procedure of the traveling vector integration interrupt processing in step 108, which processing is executed every time the vehicle travels a predetermined distance ΔD and a vehicle speed pulse is generated.

That is, first, the vehicle direction θ is read in step 200, and this θ is performed by reading the one detected and stored in the process of step 110 described later.

As a result, the X-direction component and the Y-direction component of the distance ΔD are obtained (step 202), and each component is X-valued.
The accumulated distance is added to the direction accumulated distance and the Y direction accumulated distance to update the accumulated distance (step 204).

When the above-described processing of FIG. 10 is performed, the coordinate position indicated by the latest X-direction integrated distance and Y-direction integrated distance is treated as the current traveling position of the vehicle in the processing of FIG. 9 described above.

The vehicle traveling azimuth θ used in step 108 is determined by the azimuth calculation interrupt processing in step 110. This process is a timer interrupt process performed every time a predetermined time Δt elapses, and the process procedure will be described with reference to the flowchart of FIG. 11.

In this process, first, the gyro sensor 60
The angular velocity ω of the vehicle is read from 3 (step 300).

This is because a voltage proportional to the angular velocity ω is output from the gyro sensor 603, so after A / D conversion,
It is obtained by multiplying by a predetermined coefficient.

Next, at step 302, the bias error of the gyro sensor 603 is read. This is calculated in step 112 described later, and the bias error b = 0 is set until it is calculated.

As described above, using the read ω, b and Δt,
The azimuth change amount Δθg from the previous processing to the current processing is calculated by the calculation in step 304. Then, the other azimuth output and the azimuth after filtering are output at 306.

Next, the processing procedure of the bias error calculation interrupt processing shown in step 112 will be described with reference to FIGS.

FIG. 12 shows a process for calculating a gyro bias error in a relatively short time (several tens of seconds), which is a timer interrupt process activated every time a fixed time T elapses.

First, in step 402, the number of interrupts is counted. Next, in steps 404 and 406, the current outputs from the geomagnetic sensor and the gyro are read, and the traveling azimuth is calculated based on the gyro rotation angle (gyro azimuth) based on the geomagnetic sensor output. In step 408, the current position is calculated. Read the direction of a road. In step 410, the angular velocity ω is calculated by taking the difference from the previous value for each of the gyro azimuth and the road azimuth, and the gyro angular velocity difference based on the current road azimuth and the gyro angular velocity difference based on the road angular velocity are respectively Δ.
It is calculated as θ and Δω. In step 412, it is determined whether or not the predetermined interrupt count N (memory size) has been exceeded. If the predetermined interrupt count N has been exceeded, the memory size is full and the following processing is performed.

In step 414, the memory positions of the data up to the previous time are shifted one by one, and the new data is input to the Nth data. In step 416, the case where the gyro angular velocity and the road angular velocity are small is selected (a large number means that the turning angle is large, and the error due to the scale error, not the bias error, becomes dominant), and the step 418 is performed.
In, the case where Δθ changes to some extent or more is selected (because the case of a straight line is excluded). In addition, in step 420, it is determined that the bias error is generated in one direction (if it is not in one direction, the phenomenon caused by the steering wheel operation may be mistaken for an error), and in step 422, the fluctuation width of Δω Is determined to be small. If it is large, there is a possibility of operating the steering wheel.

If the above judgment is satisfied, the value obtained by dividing the change amount of the gyro angle difference Δθ by the elapsed time NT is regarded as the bias error b (step 424) and the interrupt processing is ended.

FIG. 13 shows a process for calculating a gyro bias error for a relatively long time (tens of minutes), which is a timer interrupt process activated every time a fixed time T elapses.

Steps 502-514 are shown in FIG.
The process is the same as steps 402 to 414 in step 2. In step 516, the long-term change in the gyro angle difference Δθ (calculated in step 510) based on the road angle is statistically processed, the slope of the regression line is calculated, and this is regarded as the bias error b. Then, in step 518,
When the fluctuation range of the bias error b is small, the average of the maximum and the minimum is newly set as the bias error b as in step 520.

The procedure following step 518 was determined for the following reasons. Since the slope b (see FIG. 15) representing the gyro bias is obtained from the statistical calculation (regression line), it cannot be calculated accurately unless the amount of information is accumulated to some extent. Therefore,
Information is accumulated up to N−k, and calculation of b is started from N−k. After that, the information increased up to N is added every time, and b (n) is calculated every time. Then, in step 518, it is determined whether or not b (n), which fluctuates with time, converges to a certain width.
At 0, the maximum value and the minimum value are averaged to obtain the gyro bias b.

This is because the calculation of b may have an adverse effect unless it is performed when the variation of the inclination is small.

The bias error calculation interrupt processing as described above can be directly applied to the change rate calculation 4 of the error measuring means 10 in the filtering method of FIG.

[0092]

According to the integral type sensing device of the present invention, in the method of measuring the differential quantity of the target physical quantity and then integrating it to calculate the target quantity, the error included in the measured differential quantity. , Estimating the bias error etc.,
It becomes possible to correct, and the error originally supposed to be accumulated by integration can be reduced.

As a result, since the error of the measurement signal itself by the sensor can be reduced, unlike the final stage correction such as the error correction of the map matching method in the vehicle navigation device, for example, a similar error may occur after the correction. It does not happen immediately. Therefore, the present invention is equivalent to using a highly accurate sensor,
System reliability can also be improved.

[Brief description of drawings]

FIG. 1 is a block diagram illustrating the principle of an integral sensing device of the present invention.

FIG. 2 is a block diagram for explaining a conventional example of a vehicle navigation device.

FIG. 3 is a block diagram illustrating the principle of the integral type sensing device (filter process) of the present invention.

FIG. 4 is a block diagram showing an embodiment in which the configuration of FIG. 1 is applied to flight distance detection in an aircraft inertial navigation system.

FIG. 5 is a block diagram showing an embodiment in which the configuration of FIG. 1 is applied to detection of traveling direction in a vehicle navigation device.

FIG. 6 is a block diagram showing an embodiment in which the configuration of FIG. 3 is applied to detection of traveling direction in a vehicle navigation device.

FIG. 7 is a block diagram showing an embodiment in which the configuration of FIG. 1 is applied to detection of travel distance in a vehicle navigation device.

FIG. 8 is a block diagram showing an embodiment in which the configuration of FIG. 1 is applied to detection of traveling speed in a vehicle navigation device.

FIG. 9 is a general flow of a vehicle navigation device to which the present invention is applied.

10 is a flowchart of a travel vector integration interrupt process in FIG. 7.

11 is a flowchart of a direction calculation interrupt process in FIG.

12 is a flowchart of a bias error calculation interrupt process (short term) in FIG.

13 is a flowchart of a bias error calculation interrupt process (long-term) in FIG.

FIG. 14 is a block diagram showing a navigation system to which the present invention is applied.

FIG. 15 is an explanatory diagram of how to obtain a bias error.

[Explanation of symbols]

 1 1st means 2 2nd means 3 Adder (subtractor) 5 Error estimation part 6 Differential amount measurement part 7 Error correction part 8 Time integration part 9 Physical quantity 10 Error measurement means 11 Filter gain calculation means 12 Target Physical quantity estimating means 13 Kalman filter 21 Map 22 Geomagnetic sensor 24 Filter 40 Angular velocity measuring unit 41 Bias error correcting unit 44 Traveling direction 45 GPS device

 ─────────────────────────────────────────────────── ─── Continuation of the front page (72) Inventor Shigeru Kakumoto 1-280, Higashi Koikekubo, Kokubunji, Tokyo Inside the Central Research Laboratory, Hitachi, Ltd.

Claims (11)

[Claims] 1. A first means for obtaining a physical quantity to be sensed, the first means having a measuring means for measuring a differential quantity of the physical quantity and an integrating means for integrating the differential quantity. In the integral type sensing device, an error of the measured differential quantity in the first means is estimated from the second means for obtaining the physical quantity and the change rate of the difference between the respective physical quantities obtained by the first and second means. And an error estimating unit for correcting the measured differential amount in the first unit based on the error. 2. A first means for obtaining a physical quantity to be sensed, a second means for obtaining the physical quantity, and two physical quantities obtained from the first and second means,
A physical quantity estimating means for estimating a physical quantity by applying a Kalman filter for reducing an influence of an error of the physical quantity; and the first means for measuring a differential quantity of the physical quantity, and integrating the differential quantity. Integrating type sensing device having an integrating means for performing an error for extracting an error component of the first and second means based on a rate of change of a difference between respective physical quantities obtained by the first and second means. It has a measuring means and a filter gain calculating means for calculating a Kalman filter gain using the error component measured by the error measuring means, and the physical quantity estimating means calculates the physical quantity obtained from the first and second means. An integral sensing device, wherein a physical quantity is estimated by performing a weighting process based on the Kalman filter gain. 3. The integral type sensing device according to claim 1, wherein the rate of change of the physical quantity difference is a rate of change with time. 4. The integral type sensing device according to claim 1, 2 or 3, wherein the error is a bias error of a measured differential amount in the first means. 5. The integral type sensing device according to claim 4, wherein the change of the difference of the physical quantities with respect to time is represented by a regression line, and the slope of the regression line is defined as a bias error of the measured differential amount in the first means. An integral sensing device characterized by: 6. The integral-type sensing device according to claim 1, wherein the physical quantity is a traveling azimuth angle of the vehicle, and the first means detects the traveling azimuth by a gyro sensor. The second sensing means detects a position and outputs a corresponding road azimuth angle stored in a road map memory as a traveling azimuth angle. 7. The integral sensing device according to claim 6, wherein the second means is a GPS device that detects a position and a direction based on a satellite signal. 8. The integral type sensing device according to claim 6, wherein the second means is a geomagnetic sensor for detecting the earth's magnetic field. 9. The integral type sensing device according to claim 1, wherein the physical quantity is a traveling distance of a vehicle, the first means obtains a traveling distance from a vehicle speed, and the second means is a position. Is detected and the corresponding road length stored in the road map memory is output as a travel distance, and the change rate according to the travel distance of the difference between the respective physical quantities obtained by the first means and the second means An integral sensing device, comprising: a means for calculating a scale factor error of a measurement differential amount in the first means; and a means for correcting the measurement differential amount in the first means based on the scale factor error. 10. The integral sensing device according to claim 9, wherein the second means is a GPS device that detects a position and a direction based on a satellite signal. 11. The integral type sensing device according to claim 3, 4 or 5, wherein the physical quantity is a traveling speed of the vehicle, and the first means detects an traveling speed by an acceleration sensor, Means for measuring the rotational speed of the wheel to detect the traveling speed, or to detect the traveling speed with a GPS device for detecting the speed of the moving body by a signal from a satellite. .