KR101217138B1 - Apparatus and method for estimating position of mobile unit - Google Patents

Abstract

The present invention relates to an apparatus and method for estimating a position of a moving object. The apparatus for estimating position of a moving object includes a GPS unit for collecting coordinates of the moving object, an acceleration measuring unit for measuring an acceleration of the moving object, and a rotation of the moving object. A gyro sensor unit for measuring an angle, a discrete wavelet transform unit for outputting an output signal of the acceleration measurement unit by at least one or more frequency bands to generate a transform signal for each frequency band, and a converted signal according to the discrete wavelet transform Set the variable threshold by comparing the wavelet coefficient with the preset threshold, remove the noise included in the converted signal by using the variable threshold, and output the noise-free acceleration signal by integrating the converted signals generated for each frequency band. Noise canceling unit and the acceleration signal By combining the coordinates of the moving object provided from each of the GPS section and the acceleration of the moving body, the rotation of the mobile member, and includes a Unscented Particle Filter to estimate the position of the moving object according to the result of the combination.

Description

Apparatus and method for estimating position of moving object {APPARATUS AND METHOD FOR ESTIMATING POSITION OF MOBILE UNIT}

The present invention relates to an apparatus and method for estimating a position of a moving object, and more particularly, a position of a moving object which estimates an outdoor precise position of a moving object by combining an inertial navigation system (INS) and a global positioning system (GPS). An apparatus and method for estimating are provided.

Recently, research has been actively conducted to determine the precise position by combining the Inertial Navigation System (INS) and the Global Positioning System (GPS).

Inertial navigation systems can provide precise location information for short periods of time, including accelerometer sensors, gyroscope sensors, and magnetic compasses. However, the cumulative error occurs in estimating the absolute position and relative position of the moving object when used for a long time due to the characteristic error of the sensor itself used in the inertial navigation system and the disturbance caused by the external environment.

On the other hand, GPS has a large error for a short time and there may be a disconnection of the signal in an environment that does not receive GPS signals. Provide location information. Therefore, if the characteristics of the inertial navigation system and GPS are complemented, more accurate location estimation may be possible outdoors.

However, according to the related art, in measuring a signal change of an inertial navigation system according to a change in speed of a moving body, a high pass filter or a low pass filter of a specific band is used for filtering the signal.

The noise included in the signals of the sensors of the inertial navigation system varies depending on the speed of the moving object and the ground state. The noise is removed using a specific high pass and low pass filter without considering the deceleration, acceleration, and constant velocity motion of the moving object. The filtering method of the signal does not actively filter the signal according to the speed change of the moving object, resulting in distortion of the signal.

In addition, the inertial navigation system and GPS combination according to the prior art do not properly reflect the nonlinear motion characteristics of the moving object and combine the position information under the assumption that the moving trajectory of the moving object is linear, thereby generating the driving trajectory and error of the actual moving object. There was a problem that precise position estimation became impossible.

An object of the present invention to solve the above problem is to use an Inertial Navigation System (INS) and GPS (Global Positioning System) using an Unscented Particle Filter suitable for nonlinear or nonnormal distribution analysis. It is to provide an apparatus and method for estimating the position of the combined moving object.

Another object of the present invention is to provide a device and method for estimating the position of a moving object that is robust to the change in the speed of the moving object and the influence of the external environment by removing noise included in the measurement result of the acceleration sensor of the inertial navigation system.

The objects of the present invention are not limited to the above-mentioned objects, and other objects not mentioned can be clearly understood from the following description.

According to an aspect of the present invention, an apparatus for estimating a position of a moving object includes a GPS unit for collecting coordinates of a moving object, an acceleration measuring unit measuring acceleration of the moving object, and a gyro sensor unit measuring a rotation angle of the moving object. And a discrete wavelet transform unit for performing discrete wavelet transform on the output signal of the acceleration measurement unit for at least one frequency band to generate a transform signal for each frequency band, and in the converted signal, a wavelet coefficient according to the discrete wavelet transform. A noise canceller and a noise canceller that set a variable threshold in comparison with the signal, remove the noise included in the converted signal using the variable threshold, and output the noise-free acceleration signal by integrating the converted signals generated for each frequency band. The acceleration of the moving object obtained from the acceleration signal provided from the rejection, By combining the coordinates of the moving object provided from the full-width, and the GPS section and includes a Unscented Particle Filter to estimate the position of the moving object according to the result of the combination.

According to another aspect of the present invention, a method for estimating a position of a moving body includes measuring acceleration of a moving body, removing noise from at least one frequency band of an acceleration measurement signal including noise as a measurement result of the acceleration sensor, and performing an acceleration signal. Outputting, collecting coordinates of the moving object, measuring the rotation angle of the moving object, and accelerating the moving object obtained from the noise-free acceleration signal, the rotating angle of the moving object and the moving object coordinates Combining using a filter, and estimating the position of the moving body according to the result of the combining.

According to the present invention, the inertial navigation system (INS) and the global positioning system (GPS) can be combined to estimate the outdoor precision position of the moving object.

More specifically, the apparatus and method for estimating the moving object according to the embodiments of the present invention removes noise included in the measurement result of the acceleration sensor of the inertial navigation system, thereby being robust to changes in the speed of the moving object and the influence of the external environment. It is possible to estimate the position of. In addition, by using the Unscented Particle Filter, which is suitable for nonlinear and non-normal distribution analysis, the inertial navigation system and GPS are combined to enable accurate position estimation of the moving object.

In addition, the apparatus and method for estimating the moving object position according to the embodiments of the present invention can remove noise for each frequency band in removing noise of the acceleration measurement signal, thereby eliminating noise caused by acceleration, deceleration, and constant velocity movement of the moving object. Therefore, since distortion of the acceleration data can be minimized, more accurate position estimation of the moving object is possible.

1 is a flow chart for explaining a moving position estimation method according to an embodiment of the present invention.
2 is a block diagram illustrating an apparatus for estimating a moving position according to an embodiment of the present invention.
3 is a block diagram for explaining a mobile position estimation apparatus according to another embodiment of the present invention.
4 is a block diagram illustrating an apparatus for estimating a moving object according to still another embodiment of the present invention.
5 is a diagram illustrating decomposition of an original signal into a multi-level converted signal according to one-dimensional discrete wavelet transform;
FIG. 6 is a diagram illustrating an acceleration measurement signal including noise that is a result of measuring an acceleration of a moving direction in an acceleration measurement unit. FIG.
7 is a diagram illustrating a plurality of levels of converted signals a ₅ , d ₅ , d ₄ , d ₃ , d ₂ , and d ₁ generated by the discrete wavelet transform unit performing discrete wavelet transform of an acceleration measurement signal.

BRIEF DESCRIPTION OF THE DRAWINGS The advantages and features of the present invention, and the manner of achieving them, will be apparent from and elucidated with reference to the embodiments described hereinafter in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below, but will be implemented in various forms, and only the embodiments are to make the disclosure of the present invention complete, and those skilled in the art to which the present invention pertains. As the invention is provided to fully inform the scope of the invention, the invention is defined only by the description of the claims. It is to be understood that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. In the present specification, the singular form includes plural forms unless otherwise specified in the specification. Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

Hereinafter, a moving object estimation apparatus and a moving object estimation method according to embodiments of the present invention will be described with reference to FIGS. 1 to 7. 1 is a flowchart illustrating a moving position estimation method according to an embodiment of the present invention, Figure 2 is a block diagram illustrating a moving position estimation apparatus according to an embodiment of the present invention.

First, referring to FIG. 2, the moving object position estimating apparatus 10 according to an exemplary embodiment of the present invention may include an acceleration measuring unit 110, a discrete wavelet transform unit 120, a noise removing unit 130, and a gyro sensor unit ( 200, the GPS unit 300, and the ascended particle filter unit 400.

Here, the acceleration measuring unit 110 and the gyro sensor unit 200 may be an acceleration sensor and a gyro sensor as part of an inertial navigation system (INS).

The acceleration measuring unit 110 measures acceleration in at least one traveling direction among the X, Y, and Z axes of the moving object (S100).

Meanwhile, the acceleration measuring unit 110 may be a three-axis acceleration sensor.

Since the measurement of the acceleration in the traveling direction is made when the moving body travels in an arbitrary traveling direction, a change in speed of the moving body and noise due to the external environment are generated. In the acceleration measurement of the moving object, if the accurate acceleration cannot be measured due to the influence of noise, in the calculation of the displacement of the moving object using the inertial navigation system and the estimation of the position of the moving object, a large error occurs as the moving distance is proportional to the square of the acceleration. Can be.

Therefore, the apparatus 10 for estimating the position of the moving object according to the embodiment of the present invention removes the noise included in the acceleration measurement in the direction of acceleration of the acceleration measuring unit 110 and uses the acceleration signal from which the noise is removed to estimate the position of the moving object. The discrete wavelet transform unit 120 and the noise canceller 130 are included.

The discrete wavelet transform unit 120 performs discrete wavelet transform on the output signal of the acceleration measuring unit 110 by at least one or more frequency bands to generate a converted signal for each frequency band (S110).

That is, the discrete wavelet transform unit 120 may remove noises of the inertial navigation system or the acceleration measurement unit 110 differentially generated according to the frequency change of the moving object for each frequency region to minimize distortion of the acceleration data. The acceleration measurement signal is decomposed and a converted signal is generated for each frequency band (S110).

를 이용하여 가속도 측정 신호를 재구성한다.The discrete wavelet transform unit 120 uses a signal transformation technique called discrete wavelet transform to analyze the acceleration measurement signal. Specifically, a finite length basis function is used to decompose an acceleration measurement signal into a collection of appropriate signals, and each of the decomposed Basis signals or building blocks.

Reconstruct the acceleration measurement signal using.

Meanwhile, the discrete wavelet transform unit 120 may perform discrete wavelet transform on the acceleration measurement signal as shown in Equation (1).

(1)

f (t): acceleration measurement signal

: 변환 신호로서의 빌딩블록

: Building block as a conversion signal

: 스케일 계수

Scale factor

)을 모든 시간에 대하여 이동시켜 가속도 측정 신호의 저주파 성분은 팽창(High scales), 고주파 성분은 수축(Low scales)시켜 이동시킴으로써, 가속도 측정 신호의 고주파, 저주파 성분을 모두 분석할 수 있다. 또한, 특정 해상도를 갖는 낮은 해상도(resolution)에서 출발하여 원하는 해상도에 이르기까지 다중해상도로 가속도 측정 신호를 해석할 수 있다. 즉, 이산 웨이블릿 변환부(120)는 가속도 측정 신호를 다중해상도 해석(Multi Resolution Analysis, MRA)할 수 있다. On the other hand, the discrete wavelet transform unit 120 is a wavelet (

) Can be analyzed for all the time so that the low frequency component of the acceleration measurement signal is expanded (High scales) and the high frequency component is contracted (Low scales) to move both high and low frequency components of the acceleration measurement signal. In addition, acceleration measurement signals can be interpreted in multiple resolutions starting from a lower resolution with a specific resolution up to the desired resolution. That is, the discrete wavelet transform unit 120 may perform a multi resolution analysis (MRA) on the acceleration measurement signal.

) 및 스케일 함수(

)를 이용하여 가속도 측정 신호를 다중 해상도 해석할 수 있다.The discrete wavelet transform unit 120 may perform a wavelet function according to Equations 2 and 3

) And scale function (

) Can be used for multi-resolution analysis of acceleration measurement signals.

(2)

(3)

t: Time

n: sampling number

와

은 신호의 저주파 통과 필터 및 고주파 통과 필터의 신호 필터링을 의미한다.In addition, in Formula (2) and Formula (3),

Wow

Means the signal filtering of the low pass filter and the high pass filter of the signal.

The discrete wavelet converter 120 may decompose the output signal into signals of a plurality of high frequency bands and a low frequency band to generate a plurality of converted signals corresponding to signals of the plurality of high frequency bands and the low frequency band. Here, the transform signal may be a dot product of the scale function and the wavelet function, and the scaling function on one level may be expressed as a convolution of the scale function and the wavelet function on one level. This means that the acceleration measurement signal can be decomposed into a plurality of high frequency bands and low frequency bands.

That is, the discrete wavelet transform unit 120 may subdivide the acceleration measurement signal into a high frequency band signal and a low frequency band signal through discrete wavelet transformation, and divide the high frequency band signal and the low frequency band signal corresponding to a plurality of levels into a converted signal. Can be generated.

Hereinafter, referring to FIG. 5, the discrete wavelet transform unit 120 decomposes the acceleration measurement signal into a plurality of high frequency band signals and low frequency band signals. 5 is a diagram illustrating decomposition of an original signal into a multi-level transformed signal according to one-dimensional discrete wavelet transform.

5, the acceleration measurement signal (Signal) may be decomposed into a low frequency band signal (cA ₁₎ and high-frequency band signal (cD ₁₎ of the first level (1 ^st level), the first level (1 ^st level ) it can be disassembled back into the low-frequency signal (cA ₂₎ and a high frequency band signal (cD ₂₎ of the second level (2 ^nd level), the low-frequency signal (cA _1). In the same manner, the acceleration measurement signal Signal may be decomposed into a third level or higher level according to the discrete wavelet transform.

Hereinafter, with reference to FIGS. 6 and 7, the noise removing unit 130 removes noise from the converted signal by using a variable threshold method.

FIG. 6 is a diagram illustrating an acceleration measurement signal including noise that is a result of measuring an acceleration in a moving direction of the moving object by the acceleration measuring unit 110.

FIG. 7 illustrates a plurality of levels of converted signals a ₅ , d ₅ , d ₄ , d ₃ , d ₂ , and d ₁ generated by the discrete wavelet transform unit 120 performing discrete wavelet transform of the acceleration measurement signal. Drawing.

Referring to FIG. 6, the acceleration signal measured by the acceleration measuring unit 110 includes noise due to a speed change and an external environment in an acceleration section and a constant velocity section (velocity 1 and velocity 2). Assuming the noise-free acceleration signal, it is preferable that the measured signal value is maintained at a constant level (0 m / s ^{2 in} FIG. 6) at least in the constant velocity section, but as shown in FIG. The measured value measured by the measuring unit 110 may not maintain a constant level due to noise.

Accordingly, in order to remove noise from the acceleration measurement signal without distortion of the acceleration data, the moving object position estimation apparatus 10 according to the exemplary embodiment of the present invention performs discrete wavelet transform of the acceleration measurement signal by discrete wavelet transform to convert the converted signal. The noise canceller 130 removes noise of the converted signal from the multiple levels of the converted signals a ₅ , d ₅ , d ₄ , d ₃ , d ₂ , and d ₁ .

In detail, the noise removing unit 130 sets a variable threshold in the converted signal by comparing the wavelet coefficient according to the discrete wavelet transform with a preset threshold (S120), and includes the noise included in the converted signal using the variable threshold. In operation S130, noise is removed by integrating the converted signals generated for each frequency band.

Referring to FIG. 7, the lowest low frequency converted signal a ₅ among the multiple levels of converted signals a ₅ , d ₅ , d ₄ , d ₃ , d ₂ , and d ₁ has noise in acceleration and constant velocity sections. Accelerated at constant velocity for d ₅ , d ₄ , d ₃ , d ₂ , and d ₁ (especially d ₄ , d ₃ , d ₂ ), which are not large relative to the original signal components, but are converted signals of relatively high frequency bands There are relatively many values corresponding to noise whose value does not converge to zero. Therefore, in order to remove noise while reducing distortion of acceleration data, it is desirable to remove noise of the converted signal for each frequency band.

The noise removing unit 130 may remove noise by converging a portion (or a portion not included) included in the variable threshold to a predetermined value, and may remove noise included in at least one of the plurality of converted signals. In operation S130, the acceleration signal may be output by integrating the converted signal and the plurality of converted signals from which the noise is removed.

In addition, the noise removing unit 130 may decompose and process the acceleration measurement signal output according to the change in the speed of the moving object into high frequency and low frequency components, and may remove noise while minimizing distortion of the acceleration data (S130). .

The threshold may be set based on at least one of the standard deviation of the converted signal and the sampling number of the converted signal, and the noise removing unit 130 may remove noise of the converted signal that exceeds the standard deviation range of the converted signal. The variable threshold may be set in comparison with the threshold (S120).

)를 설정할 수 있다(S120).On the other hand, the noise removing unit 130 is a variable threshold (

) Can be set (S120).

(Equation 4)

: 이산 웨이블릿 변환에 따라 계산된 웨이블릿 계수

: Wavelet coefficient calculated according to the discrete wavelet transform

: 임계값

: Threshold

Here, since the noise removing unit 130 sets the variable threshold by comparing the wavelet coefficient according to the discrete wavelet transform with a predetermined threshold value, the change of the setting value of the threshold may affect the noise removing performance. Therefore, the threshold value can be set to select an optimal value according to the speed change of the moving object and the external environment.

)은 수학식 5에 따라 설정될 수 있다.On the other hand, the threshold value (

) May be set according to Equation 5.

(5)

: 신호의 표준 편차

: Standard deviation of the signal

: 신호의 샘플 개수

: Number of samples in the signal

Gyro sensor unit 200 measures the rotation angle of the moving object (S200). On the other hand, the gyro sensor unit 200 may be a three-axis gyro sensor.

The GPS unit 300 collects a position, a speed, and the like including the X coordinate and the Y coordinate of the moving object (S300).

The ascending particle filter unit 400 is an inertial navigation system that acquires the motion information of the moving object at a time shorter than the GPS update time from the positional information (the X coordinate and the Y coordinate of the moving object) updated every second in the global positioning system (GPS). (Inertial Navigation System, INS) is corrected by the information (acceleration of the moving body and the rotation angle of the moving body), and the combined information of each, and the position of the moving body is estimated according to the result of the combination.

In more detail, the unscented particle filter 400 may use an unscented particle filter to determine the acceleration of the moving object, the rotation angle of the moving object, and the coordinates of the moving object obtained from the acceleration signal from which the noise is removed. By using the combination (S400), the position of the moving body is estimated according to the result of the combination (S500).

Each particle of the ascendant particle filter unit 400 receives the rotation angle of the moving object (the pitch axis angular velocity and the yaw axis angular velocity) obtained from the gyro sensor unit 200 and the acceleration in the advancing direction obtained from the noise removing unit 130 as an input. This value (rotation angle and acceleration) is accumulated and the next state is estimated. At this time, the uncentralized particle filter 400 more accurately propagates the average value and covariance of the next state of the nonlinear system by using an unscented transform (UT).

Here, the propagation of the mean value and the covariance using the ascendant transform may be performed according to Equations 6 and 7 below.

(6)

: 시그마 포인트

Sigma Points

: 평균값

: Mean value

: 공분산(Covariance)

Covariance

: 상태(State) 크기

: State size

: 시그마 포인트 수(1~2L-1))

: Number of sigma points (1 ~ 2L-1))

(7)

: 시그마 포인트

Sigma Points

: 전파된 시그마 포인트(Transferred Sigma Point)

: Transferred Sigma Point

By reflecting the weights based on the measured values of the GPS unit 300 in the mean value and the covariance according to the unsented transformation, the mean of the actual distribution of the moving object, the estimated mean value very close to the covariance, and the covariance can be obtained. It is possible to estimate the position of the moving object using the estimated average value and covariance.

Here, the weight may be calculated according to Equation 8.

(8)

: 상태 변수(이동체의 가속도, 피치각, 요각, 속도 및 위치에 대응)

: State variable (corresponding to acceleration, pitch angle, yaw angle, speed and position of moving body)

: 가중치

: weight

: 측정값(GPS가 제공하는 이동체의 X 좌표 및 Y 좌표에 대응)

: Measured value (corresponding to X coordinate and Y coordinate of moving object provided by GPS)

: 시간

: time

: 파티클의 개수(1~N)

: Number of particles (1 ~ N)

Meanwhile, the uncentralized particle filter 400 performs sequential importance sampling (SIS) for estimating the position of the moving object and weights the particles by reflecting at least one measurement value of the acceleration of the moving object and the coordinates of the moving object. Recalculate.

Here, sequential importance sampling may be performed according to Equation 9.

(9)

: 상태 변수(이동체의 가속도, 피치각, 요각, 속도 및 위치에 대응)

: State variable (corresponding to acceleration, pitch angle, yaw angle, speed and position of moving body)

: 측정값(GPS가 제공하는 이동체의 X 좌표 및 Y 좌표에 대응)

: Measured value (corresponding to X coordinate and Y coordinate of moving object provided by GPS)

: 시간(Time)

: Time

: 파티클의 개수(1~N)

: Number of particles (1 ~ N)

If sequential importance sampling is repeated over time, particles with a very small weight can be ignored, which reduces the total number of particles. As a result, the ascentd particle filter unit 400 may perform re-sampling on the particles so that the position estimation of the moving object can be efficiently performed. When re-sampling is performed, the weight of each particle can be evenly distributed and an appropriate number of particles can be maintained for efficient position estimation of the moving object.

Meanwhile, the uncentralized particle filter 400 may update the position estimate of the moving body according to the recalculated weight and the result of the re-sampling.

As described above, the uncentralized particle filter unit 400 is a probability distribution corresponding to the position of the moving object in the probability space and the probability space consisting of state variables related to the position of the moving object using the acceleration of the moving object and the rotation angle of the moving object. Constituting a plurality of particles, calculating the weights of the particles based on the coordinates of the moving object, and calculating the x-axis displacement, y-axis distance, y-axis distance, At least one of a pitch axis angle of rotation and a yaw axis angle of rotation is calculated and output as a result value.

That is, the combination of the acceleration of the moving object, the rotation angle of the moving object, and the coordinates of the moving object in the ascendant particle filter unit 400 is expressed as a plurality of particles having a probability distribution, and according to the weighted sum of the particles whose weight is reflected, It is output in the form of x axis distance, y axis distance, pitch axis angle of rotation and yaw axis angle of rotation. Meanwhile, in the ascended particle filter unit 400, the x axis displacement, the y axis distance, the pitch axis angle of rotation, and the yaw axis rotation angle of the moving object are used. The position of the movable body can be estimated using the result value of any one of angle of rotation and the coordinates of the movable body at a past time.

Hereinafter, a moving object estimation apparatus according to another embodiment of the present invention will be described with reference to FIG. 3. 3 is a block diagram illustrating an apparatus for estimating a moving object according to another embodiment of the present invention.

Here, the same reference numerals are used for components that perform the same functions as the components illustrated in FIG. 2, and a detailed description of the corresponding components will be omitted.

Referring to FIG. 3, the moving object position estimating apparatus 20 according to an exemplary embodiment of the present invention may include an acceleration measuring unit 110, a discrete wavelet transform unit 120, a noise removing unit 130, and a gyro sensor unit 200. In addition, the GPS unit 300 and the ascended particle filter unit 400 may further include a magnetic compass unit 500 capable of measuring an absolute angle.

The magnetic compass unit 500 measures the magnetic angle with respect to the pitch axis or yaw axis of the movable body.

The magnetic compass unit 500 may be a magnetic compass as part of an Inertial Navigation System (INS).

On the other hand, the magnetic angle of the moving object measured in the magnetic compass unit 500 may be used to calculate or recalculate the weight of the particles in the ascendant particle filter 400.

Therefore, the moving object position estimating apparatus 20 according to an embodiment of the present invention can estimate the position of the moving object reflecting the absolute angle of the moving object using the magnetic angle of the moving object measured by the magnetic comparator 500.

Hereinafter, an apparatus for estimating a moving object according to still another embodiment of the present invention will be described with reference to FIG. 4. 4 is a block diagram illustrating an apparatus for estimating a moving object according to still another exemplary embodiment of the present invention.

Here, the same reference numerals are used for components that perform the same functions as the components illustrated in FIG. 2, and a detailed description of the corresponding components will be omitted.

Referring to FIG. 4, the moving object position estimating apparatus 30 according to another exemplary embodiment of the present invention may include an acceleration measuring unit 110, a discrete wavelet transform unit 120, a noise removing unit 130, and a gyro sensor unit 200. ), GPS unit 300 and the ascended particle filter unit 400, the X-axis displacement calculation unit 610, Y-axis displacement calculation unit 620, yaw axis rotation angle calculation unit 630 and pitch axis The rotation angle calculation unit 640 further includes.

In the mobile body position estimation apparatus 30 according to the embodiment of the present invention, the ascended particle filter unit 400 allows the inertial navigation system and the GPS to be combined, and the x-axis displacement, the y-axis displacement, and the pitch of the movable body. The axis rotation angle and yaw axis rotation angle are calculated and output by the X axis displacement calculation unit 610, the Y axis displacement calculation unit 620, the yaw axis rotation angle calculation unit 630, and the pitch axis rotation angle calculation unit 640, respectively. do.

More specifically, the ascented particle filter unit 400 is a probability distribution corresponding to the position of the moving object in the probability space and the probability space consisting of state variables related to the position of the moving object using the acceleration of the moving object and the rotation angle of the moving object. It constructs a plurality of particles, calculates the weight of the particles based on the coordinates of the moving object, the x-axis displacement, y-axis displacement, pitch axis rotation angle and yaw axis of the moving object which is a result of the weighted sum of the particles reflected weight The rotation angle is calculated by the X-axis displacement calculator 610, the Y-axis displacement calculator 620, the yaw axis rotation angle calculator 630, and the pitch axis rotation angle calculator 640, respectively.

Accordingly, the X-axis displacement calculator 610 calculates the x-axis displacement of the moving body according to the weighted sum of the particles reflected by the weight of the uncentralized particle filter 400.

In addition, the Y axis displacement calculator 620 calculates a y axis distance according to the weighted sum of the particles in which the weight of the uncentralized particle filter 400 is reflected.

In addition, the yaw axis rotation angle calculator 630 calculates a pitch axis angle of rotation according to the weighted sum of the particles reflected by the weight of the uncentral particle filter unit 400.

In addition, the pitch axis rotation angle calculator 640 calculates a yaw axis angle of rotation according to the weighted sum of the particles in which the weight of the uncentralized particle filter 400 is reflected.

It will be understood by those skilled in the art that the present invention may be embodied in other specific forms without departing from the spirit or essential characteristics thereof. It is therefore to be understood that the above-described embodiments are illustrative in all aspects and not restrictive. The scope of the present invention is shown by the following claims rather than the detailed description, and all changes or modifications derived from the claims and their equivalents should be construed as being included in the scope of the present invention.

Claims (11)

An acceleration measuring unit measuring an acceleration of the moving body;
A gyro sensor unit measuring a rotation angle of the moving body;
A GPS unit for collecting coordinates of the moving object;
A discrete wavelet transform unit for performing discrete wavelet transform on the output signal of the acceleration measuring unit for at least one frequency band to generate a converted signal for each frequency band;
In the converted signal, a variable threshold is set by comparing the wavelet coefficients according to the discrete wavelet transform with a preset threshold, and the noise included in the converted signal is removed using the variable threshold, and generated for each frequency band. A noise canceller configured to output the acceleration signal from which the noise is removed by integrating the converted signals; And
Combining the acceleration of the movable body obtained from the acceleration signal provided from the noise canceling unit with the rotation angle of the movable body provided from the gyro sensor unit and the coordinates of the movable body provided from the GPS unit, and as a result of the combining Including; an ascendant particle filter for estimating the position of the moving body;
The ascended particle filter unit,
Comprising a probability space by including at least one of the x coordinate, y coordinate, pitch angle and yaw angle of the moving object as a state variable
Moving object position estimation device. The method of claim 1, wherein the discrete wavelet transform unit,
Decomposing an output signal of the acceleration measurement unit into a signal of a plurality of high frequency bands and a low frequency band to generate a plurality of converted signals corresponding to the signals of the plurality of high frequency bands and the low frequency band,
The noise removing unit may remove the noise included in at least one converted signal among the plurality of converted signals, and output the acceleration signal by integrating the converted noise signal and the plurality of converted signals.
Moving object position estimation device. The method of claim 1,
The threshold is set based on at least one of the standard deviation of the converted signal and the sampling number of the converted signal,
The noise removing unit sets the variable threshold in comparison with the threshold value so that the noise of the converted signal is removed based on at least one of the standard deviation of the converted signal and the sampling number of the converted signal.
Moving object position estimation device. The method of claim 1, wherein the ascend particle filter,
By using the acceleration of the moving object and the rotation angle of the moving object to form a plurality of particles of a probability space consisting of a state variable relating to the position of the moving object and a probability distribution corresponding to the position of the moving object in the probability space, Calculating weights of the particles based on coordinates, and estimating the position of the moving body according to the weighted sum of the particles reflecting the weights
Moving object position estimation device. delete The method of claim 1, wherein the ascend particle filter,
Outputting at least one of x-axis displacement, y-axis displacement, yaw axis rotation angle, and pitch axis rotation angle as a result of estimating the position of the movable body;
Moving object position estimation device. Measuring the acceleration of the moving object using an acceleration sensor;
Outputting an acceleration signal by removing noise from at least one frequency band of an acceleration measurement signal including noise as a measurement result of the acceleration sensor;
Collecting coordinates of the moving object using GPS;
Measuring a rotation angle of the moving body using a gyro sensor; And
Combining the acceleration of the moving object obtained from the acceleration signal from which the noise is removed, the rotation angle of the moving object and the coordinates of the moving object using an unscented particle filter, and estimate the position of the moving object according to the result of the combining. Including;
Estimating the position of the moving body,
Defining an average and a covariance of the acceleration of the mobile body and the rotation angle of the mobile body;
Constructing a plurality of particles having a state variable relating to the position of the moving object, a probability space consisting of the state variable, and a probability distribution corresponding to the position of the moving object in the probability space using at least one of the mean and the covariance; And
Calculating weights of the particles based on the coordinates of the moving object and estimating the position of the moving object according to the weighted sum of the particles reflecting the weights;
Moving object position estimation method. The method of claim 7, wherein the outputting the acceleration signal,
Generating a converted signal for each frequency band by discrete wavelet transforming an acceleration measurement signal of a moving object including noise as a measurement result of the acceleration sensor by at least one or more frequency bands;
In the converted signal, setting a variable threshold by comparing a wavelet coefficient according to the discrete wavelet transform with a preset threshold;
Removing the noise included in the converted signal using the variable threshold; And
Integrating the converted signals generated for each frequency band and outputting the noise-free acceleration signal.
Moving object position estimation method. The method of claim 8, wherein generating the converted signal comprises:
Decomposing the output signal into signals of a plurality of high frequency bands and a low frequency band to generate a plurality of converted signals corresponding to the signals of the plurality of high frequency bands and the low frequency band,
The step of removing the noise includes:
Removing the noise from at least one level of the converted signal of the plurality of levels
Moving object position estimation method. delete The method of claim 7, wherein estimating the position of the moving body,
Recalculating the weights of the particles by reflecting at least one measurement value of the acceleration of the mobile body, the coordinates of the mobile body, and the magnetic angle of the mobile body; And
Updating the position of the moving body in accordance with the recalculated weights.
Moving object position estimation method.

Images