KR101301462B1 - Pedestiran dead-reckoning apparatus using low cost inertial measurement unit and method thereof - Google Patents

Abstract

PURPOSE: A pedestrian inertia navigation device using a cheap inertia sensor and a navigation method are provided to accurately track the location of a pedestrian by using a cheap acceleration sensor and a gyroscope by calibrating the sensor and compensating temperature. CONSTITUTION: A pedestrian inertia navigation device (100) using a cheap inertia sensor includes an accelerator sensor (110), a gyroscope (140), a step detecting module (120), a stride estimation module (130), a posture and orientation angle estimation module (150), and a pedestrian inertia navigation module (160). The acceleration sensor is mounted on the waist of a pedestrian. The gyroscope is mounted on the waist of the pedestrian. The step detecting module detects steps of the pedestrian by using a measurement value of the acceleration sensor. The stride estimation module estimates strides of the pedestrian by using the detected steps. The posture and orientation angle estimation module estimates the posture and the orientation angle of the pedestrian. The pedestrian inertia navigation module calculates a traveling route of the pedestrian by using the estimated posture and strides. [Reference numerals] (110) Accelerator sensor; (120) Step detecting module; (130) Stride estimation module; (140) Gyroscope; (150) Posture and orientation angle estimation module; (160) Pedestrian inertia navigation module

Description

Pedestrian inertial navigation device using low cost inertial sensor and its navigation {PEDESTIRAN DEAD-RECKONING APPARATUS USING LOW COST INERTIAL MEASUREMENT UNIT AND METHOD THEREOF}

The present invention relates to a pedestrian inertial navigation device and its navigation, and more particularly, to a pedestrian inertial navigation device using a low-cost inertial sensor and its navigation.

Location-aware technology for pedestrians generally uses a global positioning system (GPS). GPS is a satellite-based radionavigation device, and has the advantage of providing a relatively accurate and absolute position.

However, GPS cannot be used alone in signal shadow areas such as inside buildings, indoors, and tunnels. In order to solve this problem, a location recognition system incorporating WLAN, UWB, ultrasound, and PDR has been studied.

WLAN and UWB have the advantage of being able to measure absolute position in the GPS shadow area, but have the disadvantage of building additional infrastructure. For this reason, pedestrian dead-reckoning (PDR) method without the construction of additional infrastructure has been studied a lot.

PDR is a dead reckoning system developed on the basis that pedestrians change their position by walking. By using the step information generated while the pedestrian moves, the distance and the direction of movement can be measured to know the position moved from the initial position. In this case, an inertial measurement unit (IMU) is used as the step information.

The inertial sensor is configured to obtain the pedestrian's stride length. There are two methods of obtaining the inertial sensor through double integration of acceleration or based on the pedestrian's step model. In the case of stride length estimation through double integration, the sensor is mainly mounted on the foot to obtain accurate acceleration, and the distance value is obtained by measuring the acceleration while walking from the acceleration sensor. In this case, a zero velocity update (ZUPT) technique is mainly used to minimize the accumulated error.

The gait model-based technique is mainly mounted on the waist and estimates the stride length through the characteristic modeling that occurs during the movement of pedestrians. The step model-based technique has the disadvantage of requiring model parameters, but it is more practical.

In the case of the step model-based technique, when the waist is mounted, accurate position tracking is possible depending on how to correct a large cumulative error. This cumulative error compensation technique is an important issue of position tracking.

An object of the present invention is to provide a pedestrian inertial navigation device using a low-cost inertial sensor.

Another object of the present invention is to provide a pedestrian inertial navigation using a low-cost inertial sensor.

The pedestrian inertial navigation device using the low-cost inertial sensor according to the object of the present invention, the acceleration sensor mounted on the waist of the pedestrian, the gyroscope mounted on the waist of the pedestrian, and the measured value of the acceleration sensor A step detection module for detecting a step of the step; a step estimation module for estimating a step length of the pedestrian using the detected step; and a posture of the pedestrian using the measured value of the gyroscope ( movement of the pedestrian using the attitude and direction angle estimation module for estimating attitude and heading angle, the stride estimated at the stride length estimation module, and the attitude and direction angle estimated at the attitude and direction angle estimation module. And a pedestrian inertial navigation module for calculating a path, wherein the attitude and direction angle estimation module is configured for the measured value of the gyroscope. It may be configured to compensate for the error ever. In this case, the acceleration sensor is an ADXL 345, which is a microelectro-mechanical system (MEMS) type 3-axis acceleration sensor of Analog Device, and the gyroscope is preferably ITG3200 of InvenSense. The step detection module or the attitude and direction angle estimation module may be configured to perform temperature compensation of a polynominal curver fitting method on the measured value of the acceleration sensor or the measured value of the gyroscope, respectively. . The step detection module may be configured to detect a step according to a moving average filter scheme.

Pedestrian inertial navigation using a low-cost inertial sensor according to another object of the present invention, the step of measuring by the acceleration sensor mounted on the waist of the pedestrian, the step of measuring by the gyroscope mounted on the waist of the pedestrian; Detecting a step of the pedestrian using the measured value of the acceleration sensor, estimating a step length of the pedestrian using the detected step, and measuring the measured value of the gyroscope. Estimating the attitude and heading angle of the pedestrian using the estimated stride length and the attitude and direction angle estimated by the attitude and direction angle estimation module. Calculating and using the measured value of the acceleration sensor, compensating for a cumulative error of the measured value of the gyroscope; . In this case, the acceleration sensor is an ADXL 345, which is a microelectro-mechanical system (MEMS) type 3-axis acceleration sensor of an analog device, and the gyroscope may be configured to use the ITG3200 of InvenSense. The measuring by the acceleration sensor mounted on the waist of the pedestrian may be performed by performing a temperature compensation of a polynominal curver fitting method on the measured value of the accelerometer and using the measured value of the gyroscope. The estimating the attitude and heading angle of the pedestrian may be configured to perform temperature compensation in a polynomial curve fitting method on the measured value of the gyroscope. Meanwhile, the detecting of the step of the pedestrian using the measured value of the acceleration sensor may be configured to detect the step according to a moving average filter method.

According to the pedestrian inertial navigation apparatus using the low-cost inertial sensor as described above and its navigation, the position of the pedestrian can be tracked using a low-cost acceleration sensor and gyroscope, but accurate estimation of the pedestrian can be performed through sensor compensation and temperature compensation. It's effective. There is an advantage that can increase the spread of pedestrian inertial navigation system and reduce the unit price.

1 is a block diagram of a pedestrian inertial navigation apparatus using a low-cost inertial sensor according to an embodiment of the present invention.
2 is a table showing specifications of an acceleration sensor and a gyroscope according to an embodiment of the present invention.
3 is a comparison graph before and after correction of an acceleration sensor according to an embodiment of the present invention.
4 is a detection graph of a filtered acceleration signal according to an embodiment of the present invention.
5 is a result table of step detection and stride length detection detected according to an embodiment of the present invention.
6 is a graph of a position estimation result shown through an experiment according to an embodiment of the present invention.
7 is a graph of position estimation results shown through experiments according to another embodiment of the present invention.
8 is a flowchart of a pedestrian inertial navigation using a low-cost inertial sensor according to an embodiment of the present invention.

While the invention is susceptible to various modifications and alternative forms, specific embodiments thereof are shown by way of example in the drawings and will herein be described in detail. It should be understood, however, that the invention is not intended to be limited to the particular embodiments, but includes all modifications, equivalents, and alternatives falling within the spirit and scope of the invention. Like reference numerals are used for like elements in describing each drawing.

The terms first, second, A, B, etc. may be used to describe various components, but the components should not be limited by the terms. The terms are used only for the purpose of distinguishing one component from another. For example, without departing from the scope of the present invention, the first component may be referred to as a second component, and similarly, the second component may also be referred to as a first component. And / or < / RTI > includes any combination of a plurality of related listed items or any of a plurality of related listed items.

When a component is referred to as being "connected" or "connected" to another component, it may be directly connected to or connected to that other component, but it may be understood that other components may be present in between. Should be. On the other hand, when an element is referred to as being "directly connected" or "directly connected" to another element, it should be understood that there are no other elements in between.

The terminology used herein is for the purpose of describing particular example embodiments only and is not intended to be limiting of the present invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In this application, the terms "comprise" or "have" are intended to indicate that there is a feature, number, step, operation, component, part, or combination thereof described in the specification, and one or more other features. It is to be understood that the present invention does not exclude the possibility of the presence or the addition of numbers, steps, operations, components, components, or a combination thereof.

Unless defined otherwise, all terms used herein, including technical or scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs. Terms such as those defined in commonly used dictionaries are to be interpreted as having a meaning consistent with the contextual meaning of the related art and are to be interpreted as either ideal or overly formal in the sense of the present application Do not.

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

1 is a block diagram of a pedestrian inertial navigation apparatus using a low-cost inertial sensor according to an embodiment of the present invention.

Referring to FIG. 1, the pedestrian inertial navigation apparatus 100 (hereinafter, referred to as a pedestrian inertial navigation apparatus) using a low-cost inertial sensor according to an embodiment of the present invention may include an acceleration sensor 110 and a step detection module 120. ), The stride estimation module 130, the gyroscope 140, the attitude and direction angle estimation module 150, and the pedestrian inertial navigation module 160.

The pedestrian inertial navigation device 100 tracks the position of the pedestrian using the low-cost acceleration sensor 110 and the gyroscope 140, but enables accurate estimation of the pedestrian through sensor correction and temperature compensation. Here, the acceleration sensor 110 is an ADXL 345 which is a microelectro-mechanical system (MEMS) type 3-axis acceleration sensor of Analog Devices, and the gyroscope 140 preferably uses ITG3200 of InvenSense. Do.

2 is a table showing specifications of an acceleration sensor and a gyroscope according to an embodiment of the present invention.

Referring to FIG. 2, detailed specifications of the ADXL 345, a MEM type three-axis acceleration sensor, and the ITG3200, a three-axis gyroscope, are shown.

A detailed configuration will be described with reference to FIG. 1 again.

Acceleration sensor 110 is mounted on the waist of the pedestrian to measure the three-axis acceleration of the pedestrian.

The step detection module 120 is configured to detect a step of a pedestrian using the measured value of the acceleration sensor 110. In this case, the step detection module 120 may be configured to detect a step according to a moving average filter method among various techniques. In addition, the step detection module 120 may perform temperature compensation of a polynominal curver fitting method on the measured value of the acceleration sensor 110 or the measured value of the gyroscope 140, respectively.

The stride length estimation module 130 is configured to estimate a step length of a pedestrian using the steps detected by the step detection module 120.

The gyroscope 140 is mounted on the waist of the pedestrian to measure the three-axis acceleration of the pitch, roll, and yaw.

The attitude and direction angle estimation module 150 is configured to estimate the attitude and heading angle of the pedestrian using the measured values of the gyroscope 140. In this case, the attitude and direction angle estimation module 150 may be configured to compensate for a cumulative error with respect to the measurement value of the gyroscope 140. Hereinafter, error compensation will be described in more detail.

 In addition, the attitude and direction angle estimation module 150 may perform temperature compensation in a polynominal curver fitting method on the measured value of the acceleration sensor 110 or the measured value of the gyroscope 140, respectively.

The pedestrian inertial navigation module 160 is configured to calculate a moving path of the pedestrian using the attitude and direction angle estimated by the stride length and attitude and direction angle estimation module 150 estimated by the stride length estimation module 130.

Hereinafter, a technique of modeling a sensor to compensate for temperature bias, an acceleration sensor correction technique, a gyroscope correction technique, step detection, stride length estimation, posture and direction angle estimation will be described in more detail.

First, sensor calibration is to calibrate the sensor's output to the same value specified on the datasheet. At this time, since the error is different for each sensor, it is necessary to remove the error value through the individual correction for each sensor in order to obtain accurate performance. In particular, when using a low-cost sensor, such an error causes a large position error, so correction is very important.

In general, the 3-axis MEMS inertial sensor can be represented by a linear model equation as shown in Equation 1 below.

는 센서의 측정값이고,

는 오차가 없는 센서의 실제값이고,

는 센서의 바이어스 오차값이고,

은 센서의 비정렬 오차(misalignment error)와 비직교 오차(non-orthogonal error)이다. 그리고

는 환산 계수(scale factor)를 나타낸다.here,

Is the measured value of the sensor,

Is the actual value of the sensor without error,

Is the bias error of the sensor,

Are misalignment and non-orthogonal errors of the sensor. And

Denotes a scale factor.

Equation 1 may be expressed as Equation 2 below.

는 환산 계수, 비정렬 오차, 비직교 오차가 결합된 3X3 행렬이다. 최종적으로 구하고자 하는 보정된 센서의

는 다음 수학식 3과 같다.From here,

Is a 3X3 matrix that combines the conversion coefficients, the misalignment error, and the non-orthogonal error. Of the calibrated sensor

Is as shown in Equation 3 below.

Next, the temperature compensation technique will be described.

One of the main causes of errors in inertial navigation using inertial sensors is the cumulative error due to bias error. Bias error is highly temperature dependent and requires temperature compensation.

In the present invention, temperature bias compensation is first performed on the acceleration sensor 110 and the gyroscope 140, and then the parameter of Equation 3 is obtained. Three-point compensation, linear fitting, and polynomial curve fitting are used for the temperature compensation technique. Three-point compensation or linear fit can be compensated by simple calculations, but only accurate calibration is possible when the temperature and sensor output values are always linear. Polynomial curve fitting can be compensated precisely even if the relationship between the two is nonlinear, but it requires a higher-order polynomial to increase the computation time.

는 다음 수학식 4와 같다.The present invention utilizes a second order polynomial curve fit that is also possible for nonlinearity and has a small amount of computation. Sensor value compensated for temperature bias

Is as shown in Equation 4 below.

이고,

From here,

ego,

이다.

to be.

대신에

를 대입하면 온도 바이어스를 보상한 수학식 5가 된다.In equation (3)

Instead of

Substituting into Equation 5 is obtained by compensating for the temperature bias.

Next, correction of the acceleration sensor 110 will be described.

과

를 구한다.In general, the acceleration sensor 110 has only a gravity acceleration, g value in a stationary state without movement. Based on this, the acceleration sensor 110 may be used for correction. The three-axis accelerometer has an acceleration value of ± 1 g each in six postures perpendicular to the Earth's direction of gravity. Place the acceleration sensor 110 horizontally in six postures, collect the respective acceleration values at that time, and then use the least-square method.

and

.

3 is a comparison graph before and after correction of an acceleration sensor according to an embodiment of the present invention.

Referring to FIG. 3, the acceleration signal before and after correction is shown, and the signal is obtained when the z axis is rotated in a state of 1 g and left at −1 g and then returned to its original position.

Next, correction of the gyroscope 140 will be described.

는 무시한다. 따라서 수학식 5는 다음 수학식 6으로 표현될 수 있다.The error parameter estimation of the gyroscope 140 estimates the error parameters by using the error information from the measured angular velocity value under the assumption that the reference value is known, so that rotation measurement such as a rate table is different from the acceleration sensor 110. It is possible to have a device. The bias error of the gyroscope 140 is obtained through an average of outputs of sensor data of each axis of the gyroscope 140 when the gyroscope 140 is in a stopped state. This value is equal to the value obtained through the temperature bias compensation, so the bias of the gyroscope 140

Is ignored. Therefore, Equation 5 may be expressed by Equation 6 below.

값의 추정은 레이트 테 이블(rate table)에서 나오는 각속도 실제 값,

와 각각의 자이로 축의 측정값,

를 이용하여 최소자승법을 사용한다.

The estimate of the value is the actual value of the angular velocity from the rate table,

And measurements of each gyro axis,

Use the least-square method using.

Hereinafter, a step detection technique will be described.

Since the pedestrian inertial navigation system estimates the position based on human's walking, step detection must be accurately implemented. Step detection methods mainly include peak value detection, constant range detection, and zero crossing detection. The range detection technique can be used only when the sensor is mounted on the foot. When mounted on the waist as in the present invention, a foot is detected using a pattern of the acceleration sensor 110.

의 크기 값인

를 사용한다.In the present invention, a peak value detection technique and a zero crossing detection technique are used in combination to detect a step. Since the acceleration sensor 110 is affected by gravity acceleration, the value varies according to the posture. 3-axis acceleration value corrected for robust posture detection

Is the size of

Lt; / RTI >

값에 이동평균 필터(moving average filter)를 적용한다. 도 4는 이동평균 필터의 적용 전후의 데이터를 나타낸다. 도 4를 통해 좀 더 설명한다.The peak value detection technique and the zero crossing detection technique are sensitive to noise of the acceleration sensor 110. If a momentary error value occurs due to noise of the value of the acceleration sensor 100, an incorrect step may be detected. Therefore, in the present invention, to improve the performance of the step detection

Apply a moving average filter to the values. 4 shows data before and after applying the moving average filter. This will be described further with reference to FIG. 4.

4 is a detection graph of a filtered acceleration signal according to an embodiment of the present invention.

Referring to FIG. 4, it is difficult to apply a step detection algorithm due to noise before applying the moving average filter. However, after applying the moving average filter, the step is changed into a signal form in which peak values and zero crossings are easily detected. This becomes possible. The conditions of the step detection algorithm are as follows.

1) Maximum Peak Value> Maximum Threshold & Minimum Peak Value <Minimum Threshold

2) If the time between the current maximum / minimum peak and the previous maximum / minimum peak is less than the time threshold, update to the larger / smaller maximum / minimum peak.

3) Zero crossing detection after maximum peak and minimum peak

The second condition is a condition for finding an accurate maximum / minimum peak even when the signal of the acceleration sensor 110 has several peak values, which is necessary for accurate stride estimation. If all of the above conditions are met, the step is detected at the zero crossing point. 4 shows an example of step detection through a maximum / minimum threshold, a maximum / minimum peak value, and a zero crossing point. The zero crossing point was detected at 1 g, which is the acceleration value of gravity.

Hereinafter, the stride length estimation will be described.

The stride length is the distance from the front heel to the hind heel when you walk. After the steps are detected, the distance traveled by the stride length estimation between each step and the steps must be obtained. The stride length estimation technique using the acceleration sensor 110 has various techniques such as a linear model, an inverted pendulum model, a nonlinear model (empirical model), and an artificial neural network model. Since the similar performance is shown in the four types of stride length estimation, the present invention estimates the stride length by applying a nonlinear model that is easy to implement and has one stride length estimation parameter. The equation of the non-linearity model for the estimation of the stride length is shown in Equation 8 below.

은 보폭,

는 보폭 결정 상수,

는 한 걸음에서 발생하는 최대 가속도,

은 최소 가속도 값이다. 보폭 결정 상수는 미리 실험을 통하여 구하여야 하는 값으로서 보행자의 다리 길이에 따라 달라지나 그 영향이 크지 않다. 그 최대/최소 가속도 값은 걸음 검출에서의 최대/최소 피크치를 사용한다.

Silver stride,

Is the stride determination constant,

Is the maximum acceleration in one step,

Is the minimum acceleration value. The stride length constant is a value that must be obtained through experiments, but it depends on the length of the pedestrian's legs, but its effect is not significant. The maximum / minimum acceleration value uses the maximum / minimum peak value in the step detection.

The posture estimation will be described below.

, 피치

, 요

의 조합으로 물리적으로 이해하기 쉽게 때문에 자세 계산에 많이 이용되고 있다. 각속도를 오일러 각으로 변환하는 변환 행렬은 다음의 수학식 9와 같다.The simplest method of estimating the attitude and direction angle from the acceleration sensor 110 and the gyroscope 140 is the integration of the angular velocity values of the gyroscope 140. Since the gyroscope 140 is an apparatus for measuring the angular velocity of the body, the body coordinate system changes according to the movement of the acceleration sensor 110 and the gyroscope 14. Coordinate transformation for rotation between the navigation frame and body frame of the acceleration sensor 110 and gyroscope 140 is required. The Euler angle is used as an attitude calculation algorithm for obtaining the coordinate transformation matrix. Euler angles are rolls of three angles for indicating the orientation of objects in three-dimensional space

, pitch

, Yo

Because it is easy to understand physically by the combination of, it has been widely used in posture calculation. The conversion matrix for converting the angular velocity into Euler angle is shown in Equation 9 below.

From here,

이다.

to be.

은 오일러 각 변화량이고,

은 각속도를 오일러 각으로 변환하기 위한 행렬이고,

은 자이로스코프(140)에서 측정한 각속도를 나타낸다.

Is the Euler angle change,

Is a matrix for converting angular velocity into Euler angles,

Denotes the angular velocity measured by the gyroscope 140.

In the posture measurement using only the gyroscope 140, a cumulative error occurs over time. In the case of a roll and a pitch, the cumulative error may be eliminated through the fusion of the acceleration sensor 110 and the gyroscope 140. The relationship between the output value of the acceleration sensor 110 and the Euler angle is as follows.

와

는 가속도 센서(110)로부터 측정된 롤과 피치의 각이다. 가속도 센서(110)와 자이로스코프(140)의 롤과 피치축 융합 알고리즘은 1차 상보 필터를 이용하였다. 1차 상보 필터는 그 구현이 간단하면서도 우수한 성능을 나타내며, 가속도 센서(110)에 저역 통과 필터를 사용하고 자이로스코프(140)에는 고역 통과 필터를 사용한다. 각도를 얻는데 있어서 가속도 센서(110)는 스테디 상태(steady state)일 때 특성이 좋고, 자이로스코프(140)는 고주파 특성에 우수한 특성을 가지기 때문에 두 센서를 융합하면 누적 오차가 없고 정확한 각도가 얻어진다.From here,

Wow

Is the angle of the roll and pitch measured from the acceleration sensor 110. The roll and pitch axis fusion algorithms of the acceleration sensor 110 and the gyroscope 140 used a first-order complementary filter. The first complementary filter is simple in its implementation and exhibits excellent performance, using a low pass filter for the acceleration sensor 110 and a high pass filter for the gyroscope 140. In order to obtain the angle, the acceleration sensor 110 has good characteristics in a steady state, and since the gyroscope 140 has excellent characteristics in high frequency characteristics, when the two sensors are fused, an accurate angle is obtained. .

Hereinafter, the direction angle estimation will be described.

을 나타낸다.In general, the yaw angle eliminates the cumulative error of the gyroscope 140 through fusion with the geomagnetic sensor. However, it is difficult to use a geomagnetic sensor in an indoor environment because it is affected by an external magnetic field. The yaw angle measurement using only the gyroscope 140 should be corrected because it generates a continuous cumulative error. Therefore, in the present invention, the error of the gyro bias is minimized through temperature compensation and sensor correction in order to minimize the cumulative error of the gyroscope 140. In addition, a method of minimizing the accumulation error by using the size value of the gyroscope 140 is proposed. Equation 11 is a size value of the gyroscope 140

.

값이 특정 임계치보다 낮을 경우 이를 정지 상태로 가정하면 이때 발생하는 노이즈로 인한 누적 오차와 보정되지 못한 추가적인 바이어스를 제거할 수 있다. 정지 상태에 대한 조건을 다음의 수학식 12와 같이 나타낼 수 있다.The gyroscope 140 has an angular velocity value of zero when there is no movement. However, noise and bias cause non-zero values. In standstill

If the value is below a certain threshold, assuming that it is stationary, the cumulative error due to noise and uncorrected additional bias can be eliminated. The condition for the stop state may be expressed as in Equation 12 below.

는 정지 상태에서의

의 임계치이고,

는 앞 조건이 참일 때부터 경과한 시간이고,

는 시간 임계치이다. 3D 각속도 값이 임계치보다 작고, 시간 경과가

보다 크면 정지 상태로 가정한다. 이때 발생하는 각속도 값은 0으로 만들고, 그 시간 동안의 각속도의 평균값을 바이어스로 두고 이동 상태에서 제거해 준다.here,

In the stopped state

Is the threshold of,

Is the time elapsed since the preceding condition was true,

Is the time threshold. The 3D angular velocity value is less than the threshold,

If greater than, it is assumed to be stopped. At this time, the angular velocity value is set to 0, and the average value of the angular velocity during the time is biased and removed from the moving state.

Hereinafter, the experimental results according to the embodiment of the present invention.

For the experiment, the pedestrian inertial navigation device 100 was mounted at the center behind the waist. The battery is designed to be separated and installed separately for light weight. Experiments were performed indoors, and experiments were performed to measure gait detection performance, stride estimation accuracy, and experiments for position error measurement. As the stride length estimation parameter used in the stride length estimation, a previously learned value was used. The steps were assumed to be normal walking patterns.

For the measurement of steps and the estimation of the stride length, the linear paths 20 m and 10 m were tested 10 times. The data obtained from the acceleration sensor 110 were calculated by using the step detection conditions of Equations 7 and 8, respectively, and the results are shown in FIG. 5.

5 is a result table of step detection and stride length detection detected according to an embodiment of the present invention.

 Referring to FIG. 5, the step was detected at 100% in the 20m experiment, and the maximum step detection error was 0.8% in the 100m experiment. In the case of the distance error, the average error was -0.138 m at 20 m, and the average error was 2.62 m at 100 m. It can be seen that the excellent performance with the stepped estimation error rate is within 2%.

Two types of walking tests were performed to confirm the positional performance of the pedestrian inertial navigation system 100. Each experiment analyzed the root mean square (RMS) error between the starting position and the ending position. Both experiments were performed 10 times in the same form. Experiment 1 was a walking test of a square shape of 25m x 7.24m. 6 shows the position estimation result and the position estimation error of Experiment 1. FIG. As a result of experiment 10, the average distance error was within 1% and the position error was within 2%.

In Experiment 2, a new straight line and rotation path were added in Experiment 1. 7 shows the position estimation result of Experiment 2. Experiment 2 also increased the error compared to experiment 1, but the distance error is less than 2%, the average RMS error is less than 3% can be reconfirmed that the excellent performance. Compared with the existing methods using relatively expensive sensors, it can be seen that similar performance with the expensive sensors can be obtained even though low-cost sensors are used.

8 is a flowchart of a pedestrian inertial navigation using a low-cost inertial sensor according to an embodiment of the present invention.

Referring to Figure 8, it is measured by the acceleration sensor 110 mounted on the waist of the pedestrian (S110). In this case, the temperature compensation of the polynominal curver fitting method may be performed on the measured value of the acceleration sensor 110. In addition, the acceleration sensor 110 may be configured as an ADXL 345 which is a microelectro-mechanical system (MEMS) type 3-axis acceleration sensor manufactured by an analog device.

And it is measured on the gyroscope 140 mounted on the waist of the pedestrian (S120). In this case, the gyroscope 140 may be configured to perform temperature compensation in a polynomial curve fitting method. The gyroscope 140 may be configured as an ITG3200 manufactured by InvenSense.

The step detection module 120 detects a step of the pedestrian using the measured value of the acceleration sensor 110 (S130). At this time, it is preferable to detect the step according to the moving average filter method.

The stride length estimation module 130 estimates the step length of the pedestrian using the previously detected steps (S140).

The attitude and direction angle estimation module 150 estimates the attitude and heading angle of the pedestrian using the measured values of the gyroscope 140 (S150).

The pedestrian inertial navigation module 160 calculates a moving path of the pedestrian using the attitude and the direction angle estimated by the stride length and the attitude and the direction angle estimation module 150 estimated at the stride length estimation module 130 (S160).

In addition, the pedestrian inertial navigation module 160 compensates for the cumulative error of the measured value of the gyroscope 140 using the measured value of the acceleration sensor 110 (S170).

It will be understood by those skilled in the art that various changes in form and details may be made therein without departing from the spirit and scope of the invention as defined in the appended claims. It will be possible.

Claims (8)

An acceleration sensor mounted on the waist of the pedestrian;
A gyroscope mounted on the waist of the pedestrian;
A step detection module for detecting a step of the pedestrian using the measured value of the acceleration sensor;
A stride length estimation module for estimating a step length of the pedestrian using the detected stride;
An attitude and direction angle estimation module for estimating an attitude and a heading angle of the pedestrian using the measured value of the gyroscope;
And a pedestrian inertial navigation module configured to calculate a moving path of the pedestrian by using the stride length estimated by the stride length estimation module and the posture and direction angle estimated by the posture and direction angle estimation module.
The attitude and direction angle estimation module,
Compensating for a cumulative error with respect to the measured value of the gyroscope,
The step detection module or the attitude and direction angle estimation module,
Perform temperature compensation of a second polynominal curver fitting method on the measured value of the acceleration sensor or the measured value of the gyroscope, respectively, and the measured value of the acceleration sensor after the temperature compensation is performed.

Is calculated by the following equation,

From here,

ego,

ego,
After performing the temperature compensation, the actual value of the sensor is calculated by the following equation,

Where

Is a 3X3 matrix that combines conversion factors, unalignment error, and non-orthogonal error.

Is the bias error value of the sensor,

And

Using the low-cost inertial sensor, characterized in that the acceleration sensor is placed in the six postures perpendicular to the earth's gravitational direction, each of which has an acceleration value of ± 1g, and the respective acceleration values are collected and then calculated by the least square method. Pedestrian Inertial Navigation Device. delete delete The method of claim 1, wherein the step detection module,
Pedestrian inertial navigation system using a low-cost inertial sensor, characterized in that the step of detecting the step according to the moving average filter (moving average filter) method. Measuring by an acceleration sensor mounted on the waist of the pedestrian;
Measuring by a gyroscope mounted on the waist of the pedestrian;
Detecting a step of the pedestrian using the measured value of the acceleration sensor;
Estimating a step length of the pedestrian using the detected steps;
Estimating the attitude and heading angle of the pedestrian using the measured value of the gyroscope;
Calculating a moving path of the pedestrian using the estimated stride length and the attitude and direction angle estimated by the attitude and direction angle estimation module;
Compensating for a cumulative error with respect to the measured value of the gyroscope using the measured value of the acceleration sensor,
Measuring by the acceleration sensor mounted on the waist of the pedestrian,
Performing temperature compensation of a polynominal curver fitting method on the measured value of the acceleration sensor,
Estimating the attitude and heading angle of the pedestrian using the gyroscope measurements,
Performing temperature compensation in a polynomial curve fitting method on the measured value of the gyroscope,
Compensating the cumulative error for the measured value of the gyroscope using the measured value of the acceleration sensor,
Perform temperature compensation of a second polynominal curver fitting method on the measured value of the acceleration sensor or the measured value of the gyroscope, respectively, and the measured value of the acceleration sensor after the temperature compensation is performed.

Is calculated by the following equation,

From here,

ego,

ego,
After performing the temperature compensation, the actual value of the sensor is calculated by the following equation,

Where

Is a 3X3 matrix that combines conversion factors, unalignment error, and non-orthogonal error.

Is the bias error value of the sensor,

And

Using the low-cost inertial sensor, characterized in that the acceleration sensor is placed horizontally in six postures, each of which is perpendicular to the earth's gravitational direction having an acceleration value of ± 1 g, and is collected through a least square method after collecting the respective acceleration values. Pedestrian Inertial Navigation Method. delete delete The method of claim 5, wherein the detecting of the step of the pedestrian using the measured value of the acceleration sensor comprises:
Pedestrian inertial navigation using low-cost inertial sensors, characterized in that the step of detecting the step according to the moving average filter (moving average filter) method.

Images