KR20020001257A - Positioning module and navigation method for personal navigation system using gps and inertial sensors - Google Patents

Abstract

PURPOSE: An apparatus and a method for measuring a personal location using a GPS(Global Positioning System) and an inertial sensor are provided to obtain location information of a person by using a walking coefficient method. CONSTITUTION: A GPS receiver(1) receives a satellite signal and detects information about an absolute location and a direction of a person from the received satellite signal. An acceleration meter(2) and a pitch gyro(3) are used for detecting information about walking of the person and a step of the person. A magnetometer(4) is used for obtaining an absolute azimuth angle. A heading gyro(5) is used for obtaining angular velocity. A velocity filter(6) is used for correcting the velocity of the person by combining the velocity information from the GPS receiver(1) with the walking and step information from the acceleration meter(2) and the pitch gyro(3). A direction filter(7) is used for correcting the direction of the person by combining the direction information of the GPS receiver(1) with the information from the magnetometer(4) and the heading gyro(5). A location filter(8) is used for correcting the location of the person by combining the location information from the GPS receiver(1) with the information from the velocity filter(6) and the direction filter(7).

Description

Personal positioning device and method using GPS and inertial sensor {POSITIONING MODULE AND NAVIGATION METHOD FOR PERSONAL NAVIGATION SYSTEM USING GPS AND INERTIAL SENSORS}

The present invention relates to a personal positioning device and method using a GPS (Global Positioning System) and an inertial sensor, and in particular, a new form of combined with GPS using a new dead reckoning of a step counting method that can be used by a person walking. It is a personal portable navigation system that can track the location of people moving anywhere anytime, including military operations, mountaineering and indoor walking.

Navigation techniques that have been conventionally used include GPS, Inertial Navigation System (INS), Dead Reckoning (DR), and the like.

GPS can easily obtain the user's location and speed information anywhere, there is no error accumulation over time, but there is a problem that can not be used in the situation that can not receive satellite signal, that is, in the city, forest, tunnel, indoors. In the case of inertial navigation, it is impossible to expect the desired navigation performance due to the accumulated error over time unless expensive sensors are used. And if you apply the dead reckoning to a person traveling on foot, unlike a vehicle, you can't attach an odometer to find out the moving distance, and if you use a low-cost sensor to find the speed in the same way as inertial navigation, There is a problem that can not be used due to speed error.

The present invention is to solve the problems of the prior art, to develop a new personal hand-held navigation system combined with GPS by using a new dead reckoning of the step count method that can be used while walking by a person, location information to the user The purpose of the present invention is to provide a personal positioning device and method using a GPS and an inertial sensor that can provide always.

The present invention for achieving the above object is a GPS receiver receiving information about the absolute position, direction, and velocity of the individual receiving a satellite signal, an accelerometer and pitch gyro to obtain a step detection and stride determination, and an absolute azimuth angle A speed filter that corrects the speed by combining the magnetometer obtainable, the heading gyro obtaining angular velocity, the speed information obtained from the GPS receiver with the steps and strides detected from the accelerometer and the pitch gyro, and the direction obtained from the GPS receiver. A direction filter for correcting an individual's direction using information and information obtained from a magnetometer and a heading gyro, and a position filter for correcting an individual's location using the location information of the GPS receiver and information obtained from a speed filter and a direction filter. Characterized in that the configuration.

In addition, the object of the present invention includes a parameter determining step of calculating the basic variable values such as the basic step frequency, average stride length, average speed, etc. through the pitch gyro angular velocity, initial slope, acceleration; When the absolute value of the acceleration obtained from the accelerometer is above the threshold and the step period is over the set time, the step is judged as a step and counts the pulse made by dividing the step into the rising part and the falling part, and detecting the step. A stride length determining step of determining a stride length by integrating the corrected speed every step after the correction; A speed estimating step of estimating the speed of the individual by combining the speed information provided by the GPS with the speed information by the step detection and the stride length determination; A direction estimating step of estimating an individual's direction by selecting and combining a GPS, a magnetometer, and a heading gyro; A speed filter obtains speed information necessary for dead reckoning and calculates a moving distance, and then performs a location estimation step of estimating a person's location by combining the estimated direction with GPS location information.

1 is a schematic block diagram of a personal portable navigation system applied to the present invention.

2 is a schematic diagram illustrating a calibration process for step detection and stride length correction.

3 is a schematic diagram of a stride determination algorithm using a walking feature and a method of detecting a step using an inertial sensor;

4 is a schematic diagram of a peak to peak detection algorithm of pitch angular velocity and forward directional acceleration applied to pitch angle and forward direction velocity determination.

5 is a speed filter for combining the speed provided by the GPS and the speed obtained from the dead reckoning result according to the case where GPS can be used.

FIG. 6 is a schematic diagram of setting three modes and estimating directions in consideration of whether GPS is available and a distortion of a surrounding magnetic field.

7 is a navigation algorithm system diagram of a GPS / DR combining method using a Kalman filter.

<Description of Main Parts of Drawing>

One ; GPS receiver 2; Accelerometer

3; Pitch gyro 4; Magnetometer

5; Heading gyro 6; Speed filter

7a, 7b, 7c; Directional filter 8; Position filter

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

1 is a schematic block diagram of the present invention, in which the personal portable navigation system of the present invention is a GPS receiver (1), an accelerometer (2), a pitch gyro (3), a magnetometer (4), a heading gyro (heading) GPS, magnetometer data acquisition uses RS232 serial communication, inertial sensor accelerometer (2), gyro (Analog / Digital converter). At this time, the absolute position, direction, and speed can be obtained from the GPS receiver 1.

The speed in dead reckoning can be obtained by detecting the steps from the accelerometer 2 and the pitch gyro 3 and determining the stride length. The direction may be determined through the direction filter 7 in which the absolute azimuth angle of the magnetometer 4 and the angular velocity of the heading gyro 5 are combined.

On the other hand, the personal portable navigation system can be largely divided into a speed filter 6, a direction filter 7 and a position filter 8 for obtaining a navigation solution. The speed is calibrated in FIG. 3, the direction filter 7 selects the GPS / magnet coupling method, the magnetometer / gyro coupling method, and the restricted walking method according to the surrounding conditions, and the position filter 8 7 applies a method of combining the absolute position of the GPS and the relative position of the dead reckoning.

2 is a schematic diagram of a speed calculation calibration algorithm for stride length determination. Calculating the stride in the usual way of integrating acceleration is difficult to apply due to many errors. Prior to the speed filter 6 for correcting such an error, the basic parameter values are calculated in the following order.

First, the pitch gyro (3) measures the angular velocity (101), calculates the initial slope (θ ₀ = arcsin (y ₀ / g)) (102), and then uses the acceleration FFT to determine the peak frequency, that is, the fundamental step frequency (f). ₀ ) is calculated (103).

Then, calculate the average stride length by experiment (average stride length l ₀ = distance / steps) (104), and calculate the average speed using the average stride length (average speed v ₀ = f ₀ * l ₀ ) (105). .

3 is a schematic diagram of a method for detecting a step using an inertial sensor and a step determination method using a walking feature. The personal portable navigation system is a system in which a step counting method using a step counting method is combined with a GPS. Using the method to detect the step divided into rising and falling. Step detection is from (208) to (213) in Fig. 3 and each part is as follows.

In the vertical accelerometer 2, acceleration is obtained through the A / D converter (201), and low pass filtering is performed to improve the step detection performance (208). At this time, since the accelerometer 2 is affected by gravity, the bias is corrected by a recursive average equation to obtain a desired acceleration (209).

When the absolute value of the acceleration is greater than or equal to the threshold value, it is determined as a step division condition (210), and when the step period is more than the minimum time, the step is determined as step (211), and the step is divided into a rising part and a falling part to form a pulse (212). ). The pulse is then counted and counted (213).

At this time, if the speed of the pedestrian is determined only by the step detection, an error accumulates due to the fixed step length due to the dynamic step characteristics. Therefore, the stride length can be determined at every step to minimize the error. That is, the stride determination is composed of a process of gravitational compensation using a corrected pitch angle and a correction of acceleration in a forward direction using a walking feature. The procedure below is pitch angle correction and forward acceleration gravity compensation.

First, an initial slope, a basic step frequency, an average stride length, and an average speed are calculated through a calibration process for determining the stride length of FIG. 2 (100).

The pitch gyro 3 measures the angular velocity through the A / D converter (201), and calculates a pitch angle (tilt) (= ∫ angular velocity * dt) (202).

Thereafter, as shown in FIG. 4, the peak-to-peak of the pitch angle is detected (220), the low point of the pitch angle is determined as one step, the pitch angle is corrected every step (203), and the correction value is calculated. (Compensation value = mid-to-peak slope-initial slope (θ ₀ ) (204).

At this time, the pitch angle is corrected (corrected pitch angle = pitch angle − correction value) 205, and gravity acceleration compensation (f = yg * sin (θ)) is performed to remove the influence of gravity on acceleration (206). ), Calculate the forward direction velocity (= ∫f * dt) (207).

The calculated forward direction velocity accumulates with time due to the bias error of the accelerometer 2. In order to minimize this error, a correction algorithm using the gait feature of acceleration and deceleration is applied as follows.

That is, the calculated peak-to-peak of the forward speed is detected (220), the low point of the forward speed is determined as one step, and the forward speed correction is performed every step (214). The correction at this time is as follows.

Compensation Speed = Peak and Low Intermediate Velocity-Average Velocity (v ₀ ) (215)

Calibrated Speed = Forward Speed-Calibrated Speed (216)

Stride = ∫Compensated speed * dt (217)

4 is a schematic diagram of a peak-to-peak detection algorithm of pitch angle and forward velocity, in which the inclination of the human body has a repetitive feature of decreasing and increasing, and the forward direction acceleration also includes a repetitive feature of acceleration and deceleration. This feature is consistent with the general gait feature. The corrected slope required for gravitational compensation of acceleration and the speed correction required for forward direction speed correction require the following peak-to-peak detection process.

That is, the corrected pitch angle or forward velocity data is obtained (221), and the vertex and the low point are detected (222), where the vertex is data having a larger value than the previous data and the subsequent data, and the low point has a small value. Data.

Then, it is determined whether the difference between the peak and the low point is greater than or equal to the threshold (223), the next peak and the low point are detected (224), and the low point is stored at the next low point (225).

Then, it is determined whether the vertex is smaller than the next vertex (226) and if the vertex is smaller than the next vertex, the vertex becomes the next vertex (227), when the period of the vertex is longer than a predetermined time (228) to determine the vertex step cycle Store at vertex (229).

After that, the next vertex and the low point are detected (230), the vertex is stored as the next vertex (231), and the difference between the vertex and the low point is determined to be greater than or equal to the threshold (232) and the next vertex and the low point are detected (233). The next vertex is stored at the vertex (234).

The low point is compared with the next low point (235) and when the low point is smaller than the next low point as the low point (236), when the low period is more than the minimum time is determined as the low point (237) and the low point is stored in the step cycle low ( 238).

FIG. 5 is a schematic diagram of an algorithm for obtaining velocity information necessary for dead reckoning. When the velocity information provided by the GPS receiver 1 and the velocity information by the step detection / strate determination method are combined using a Kalman filter, FIG. In addition, the speed of the user can be estimated continuously.

According to this, after determining whether the GPS speed information is available (301), GPS speed (V _GPS ) information is obtained (302). The walking speed V _DR is calculated (207), and the Kalman filter is used to combine the two speed information (V _GPS, V _DR ). In this case, the measurement equation is Z = V _GPS -V _DR + w (303).

FIG. 6 is a schematic diagram of an algorithm for obtaining direction information necessary for dead reckoning. As shown in FIG. 1, a direction is obtained from a GPS receiver 1, a magnetometer 4, and a heading gyro 5, and a direction filter according to a surrounding situation. Select the mode (7) to determine the direction. In the outdoor where GPS can be used, the first direction filter 7a combining the GPS receiver 1 and the magnetometer 4 is used.In a room where GPS cannot be used and the distortion of the surrounding magnetic field is not severe, the magnetometer 4 and the heading gyro ( The second direction filter 7b combined with the 5) is used, and the third direction filter 7c of the limited walking type using the heading gyro 5 is used in a room where the distortion of the surrounding magnetic field is severe.

First, in determining whether the GPS can be used (401), since the GPS signal cannot be received in indoors, downtowns, tunnels, forests, etc., the magnetometer 4 or the heading gyro 5 is used to determine the direction.

When the GPS can be used outdoors, the first direction filter 7a estimates the direction by applying the GPS / magnet coupling method using the Kalman filter (402).

Then, it is determined whether GPS is available in the outdoor mode (403), the GPS direction 404 and the magnetometer direction are measured (405), and the direction is estimated using the Kalman filter (406). The measurement equation is Z = H _GPS -H _DR + w.

When the GPS cannot be used in the outdoor mode, the direction is determined only by the direction of the magnetometer 4 (407), and in the room, the magnetometer 4 determines whether or not the surrounding magnetic field is distorted (408).

The second direction filter 7b estimates the direction by combining the magnetometer 4 and the heading gyro 5 using the Kalman filter when the surrounding magnetic field is not distorted more than a predetermined value indoors (409), and the heading gyro ( 5) The direction (H _Gyro ) is calculated by integrating the output, that is, the angular velocity (410), and the magnetometer direction (H _Mag ) is measured (411).

The Kalman filter is then used to combine the two directions of information (412). The measurement equation is Z = H _Mag -H _Gyro + w.

When the surrounding magnetic field is distorted to a predetermined value or more in the room, the third direction filter 7c estimates the direction by applying a limited walking method composed of straight and rotation using only the heading gyro 5 (413).

The direction of the magnetometer is measured to set an initial direction (414), and the heading gyro angular velocity is measured (415). If the angular velocity is greater than or equal to the threshold, it is determined as the first condition of the rotation section.

When the continuous rotation time is 0.2 seconds or more when the first condition of the rotation section is satisfied, it is determined as the second condition of the rotation section (417), and when the rotation section ends, the rotation angle is calculated by integrating the angular velocity during the rotation section (418). , Accumulate direction (419).

Thereafter, it is determined whether the continuous straight time is 2 seconds or more (421), a gyro bias is estimated using the recursive average equation (422), and the direction estimated in the selected mode is determined (423).

7 is a position filter 8 which combines the position of the dead reckoning by the step counting method and the GPS positional information by using the Kalman filter.

First, speed information necessary for dead reckoning is obtained from the speed filter 6 of FIG. 5 (300).

The movement distance in the dead reckoning is calculated by multiplying the speed and the time (501), and the direction mode is determined in consideration of the conditions (401) and (408) in the direction filter (7) of FIG.

Then, when the first direction filter 7a mode is selected, the GPS / magnetometer combined direction is estimated 503, and the estimated position is estimated by dead reckoning. Position P _DR is x = s * sin (H), y = s * cos (H) (504).

Thereafter, GPS position information (P _GPS ) is obtained (505), and the two position information are combined (506) using a Kalman filter. The measurement equation is z = P _DR + P _GPS + w.

On the other hand, when the second direction filter 7b mode in which GPS cannot be used is selected, the magnetometer / gyro is estimated in a combined direction (508), and the gyro is selected when the third direction filter 7c mode in which GPS cannot be used is selected. The direction is estimated by applying the limited walking method consisting of straight and rotation using only the bay (509).

Then, the position estimated by dead reckoning is calculated (510). The position P _DR is x = s * sin (H), y = s * cos (H).

Therefore, such a location tracking device and method can provide a continuous location information even in the poor reception of GPS, such as indoors and downtown can be used for a wide range of applications, such as leisure, military operations, personal location tracking, emergency rescue, etc. .

As described above, according to the present invention, the GPS and the dead reckoning are combined to measure and process walking features having forward acceleration, vertical acceleration and deceleration, tilting forward and backward, and left and right shaking according to the movement of the foot when walking. By detecting the steps and determining the stride length, it is possible to minimize the sudden error increase that is a problem in the general inertial navigation method, so that even if a GPS signal cannot be received, the individual's position and speed information can be continuously provided.

Claims (7)

A GPS receiver receiving satellite signals and receiving information about the absolute position, direction, and speed of the individual; Accelerometer and pitch gyro capable of step detection and stride determination; Magnetometer capable of obtaining an absolute azimuth; Heading gyro to obtain angular velocity; A speed filter configured to correct the speed by combining the speed information obtained by the GPS receiver with the steps and stride lengths detected from an accelerometer and a pitch gyro; A direction filter for correcting an individual's direction using direction information obtained from the GPS receiver and information obtained from a magnetometer and a heading gyro; And a position filter for correcting the position of the individual using the position information of the GPS receiver and the information obtained from the speed filter and the direction filter. The method of claim 1, wherein the direction filter Where the GPS signal can be received, the first direction filter combines the GPS and the magnetometer to estimate the direction, and when the GPS signal cannot be received and the distortion of the surrounding magnetic field is below a certain value, the magnetometer and the heading gyro are used to adjust the direction. And a third direction filter for estimating a direction using a heading gyro when the distortion of the surrounding magnetic field is greater than or equal to a predetermined value. The personal position measuring device using a GPS and an inertial sensor according to claim 1 or 2, wherein the direction filter is a Kalman filter. A parameter determining step of calculating basic variable values such as basic step frequency, average stride length, and average speed through pitch gyro angular velocity, initial slope, and acceleration; When the absolute value of the acceleration obtained from the accelerometer is above the threshold and the step period is over the set time, the step is judged as a step and counts the pulse made by dividing the step into the rising part and the falling part, and detecting the step. A stride determining step of determining a stride length by integrating the corrected speed at every step after the correction; A speed estimating step of estimating the speed of the individual by combining the speed provided by the GPS with the speed information by the step detection and the stride length determination; A direction estimating step of estimating an individual's direction by selecting and combining a GPS, a magnetometer, and a heading gyro; GPS and inertial sensors, characterized in that the position estimation step of estimating the position of the individual by obtaining the speed information necessary for dead reckoning in the speed filter, combined with the GPS position information along with the estimated direction Personal location measurement method using. The method of claim 4, wherein the step of determining the stride comprises a gravity compensation process using a corrected pitch angle and a forward direction acceleration correction process using a walking feature. 5. The method of claim 4, wherein the step of detecting the step is performed by detecting the step of moving up and down using vertical acceleration. 5. The method of claim 4, wherein the direction estimating step uses a GPS and a magnetometer in a place where GPS signals can be received, a magnetometer and a heading gyro in a place where GPS signals cannot be received and the distortion of the surrounding magnetic field is relatively low. Personal location measurement method using GPS and inertial sensors, characterized in that the heading gyro is used to estimate the direction in a place where the magnetic field is severely distorted.