CN106092095A - A kind of smart phone user indoor orientation method of earth magnetism correction inertial navigation - Google Patents

Abstract

The present invention relates to the smart phone user indoor orientation method of a kind of earth magnetism correction inertial navigation, step is as follows: according to building plane map, area to be targeted is divided into little grid；Smart phone user holds the smart mobile phone of built-in magnetometer in each grid element center locality magnetic intensity data a period of time；The absolute force data gathering each grid are averaged, and build off-line earth magnetism fingerprint base and store；User holds smart mobile phone and starts to walk about, and by the process of the accelerometer data to smart mobile phone, step-length is also estimated by detection paces, by gyro data processes the angle of direction and the direct north obtaining motion, current position magnetic intensity；Obtain the predictive value of current state；Calculate covariance matrix and the Kalman gain of forecast period；Calculating observation result；More new state.The present invention only i.e. can reach higher indoor position accuracy by smart mobile phone.

Description

A kind of smart phone user indoor orientation method of earth magnetism correction inertial navigation

Art

The invention belongs to mobile terminal application, utilize smart mobile phone to obtain the multiple sensors data carried to user Carry out indoor positioning.

Background technology

Global positioning system (GPS) can provide high-precision outdoor location service, but blocking due to building, Gps signal is the faintest the most visible in indoor, and therefore GPS alignment system may not apply to indoor positioning.And nowadays people Increasing activity be all to carry out in indoor, location Based service is increasingly paid close attention to by people.Along with radio The fast development of technology and becoming increasingly popular of mobile intelligent terminal, indoor positioning airmanship supplements skill as the one of GPS Art is fast-developing.Owing to mobile intelligent terminal equipment (smart mobile phone) has become the necessary that people are daily, there is arithmetic speed Hurry up, the feature such as Portable belt, and be integrated with the multiple sensors such as such as magnetometer, acceleration transducer, gyroscope.Hence with The indoor positioning problem of smart phone user increasingly becomes the focus of research and has been achieved for good achievement.

Existing various indoor solution major part is all based on the location technology of radio signal propagation, but this method is all Need deployment base facility, limit system large-scale application, and easily receive due to wireless signal communication environments interference (as Multipath, non line of sight interference etc.) and the human body absorption etc. to wireless signal so that the positioning performance of these technology is greatly reduced. Inertial navigation technology, as a kind of relative localization technology, has the advantages that short distance inner position precision is high, but there are cumulative errors Bigger shortcoming.And the hybrid location system of fusing wireless location technology and inertial navigation has the spy that precision is high, facilitate implementation Point is approved by numerous researcheres.

Nowadays, increasing researcher has turned to earth's magnetic field sight.Earth's magnetic field, as the intrinsic resource of the earth, is one Individual vector field, has round-the-clock, the round-the-clock and feature of full region.Therefore, utilize earth's magnetic field to have and carry out high accuracy indoor calmly The potentiality of position.Different from wireless signal, earth's magnetic field is not affected by human body, and the distribution in indoor earth's magnetic field is mainly tied by building Structure, the impact of material.Iron and steel Deng Cheng branch magnetic field over the ground in construction material produces interference steady in a long-term and produces a kind of ground Magnetic anomaly field.This GEOMAGNETIC FIELD carries out coupling location as a kind of finger print information corresponding with position.In fact, earth's magnetic field is Through being widely used in indoor positioning.A kind of method is to utilize magnetic field to distinguish the direction of motion in inertial navigation system.Other one The method of kind is to utilize fingerprinting to position as a kind of fingerprint geomagnetic field intensity.Some researcheres are in inertial navigation system Under help, use dynamic time programming (DTW) algorithm that the geomagnetic field intensity information sequence in continuous moment carries out coupling location.This The method of kind can reach the highest positioning precision, but the region that this method is only applicable to corridor the longest and the narrowest positions.

Although earth's magnetic field has been widely used in indoor positioning, but still problematic not process is fine, first earth magnetism Field intensity the faintest (only about tens uT), secondly for fingerprinting, utilize solely magnetic field intensity as fingerprint The resolution distinguishing diverse location is the lowest.Although three axle magnetometers can obtain the earth's magnetic field data of three-dimensional, thinking naturally To utilize the geomagnetic field intensity of whole three axles to improve the resolution of earth magnetism fingerprint, but it practice, magnetometer gather three numbers According to changing along with the change of sensor coordinate system, therefore, the most total utilizable magnetic field intensity in indoor.

Summary of the invention

It is an object of the invention to provide a kind of outside that do not relies on and set up extra equipment, only carried with by user Smart mobile phone can realize the method for degree of precision individual's indoor positioning.The present invention utilizes building to produce earth magnetic effect Abnormal and stable geomagnetic field intensity signal, as a kind of fingerprint corresponding with position, merges earth magnetism fingerprint by Kalman filtering User's indoor positioning is carried out with inertial navigation information.Technical scheme is as follows:

A kind of smart phone user indoor orientation method of earth magnetism correction inertial navigation, including off-line phase and on-line stage,

The off-line data collecting stage comprises the following steps:

1) according to building plane map, area to be targeted is divided into little grid, the w grid representation B_w, its grid Center RP is expressed as

2) smart phone user holds the smart mobile phone of built-in magnetometer in each grid element center locality magnetic intensity data A period of time；

3) the absolute force data gathering each grid are averaged, and build off-line earth magnetism fingerprint base and store, off-line Earth magnetism fingerprint base is by N^fIndividual fingerprint is constituted, and each fingerprint is stored asWhereinRefer to for off-line earth magnetism The position of stricture of vagina,Geomagnetic field intensity meansigma methods for this fingerprint；

The tuning on-line stage comprises the following steps:

1) initializing, smart mobile phone obtains initial absolute force information m by magnetometer₀, use m₀With in fingerprint base The geomagnetic field intensity meansigma methods of all fingerprints compare, the meansigma methods of the several fingerprint positions that will be closest to is as initialization Position (x_o,y₀)；

2) user holds smart mobile phone and starts to walk about, by the process of the accelerometer data to smart mobile phone, detection step Cut down and to step-length d_kEstimate, by gyro data being processed the included angle of direction and the direct north obtaining motion_k, Before regulation, the result of one-time positioning is designated as (x_k-1,y_k-1), current positioning result is designated as (x_k,y_k), currently available absolute force is m_k；

3) predictive value of acquisition current state:

$X\left(k\right|k-1)=A\·X(k-1|k-1)=\left[\begin{array}{cccc}1& 0& {\mathrm{sin\φ}}_k& 0\\ 0& 1& {\mathrm{cos\φ}}_k& 0\\ 0& 0& 1& 0\\ 0& 0& 0& 1\end{array}\right]\·\left[\begin{array}{c}{x}_{k-1}\\ y_{k-1}\\ d_k\\ {\mathrm{\φ}}_k\end{array}\right],$

Wherein, and X (k-1 | k-1) it is the last state positioned, X (k-1 | k-1)=[x_k-1,y_k-1,d_k,φ_k]^T, X (k | k- 1) being this state matrix positioned, A is state matrix:

$A=\left[\begin{array}{cccc}1& 0& {\mathrm{sin\φ}}_k& 0\\ 0& 1& {\mathrm{cos\φ}}_k& 0\\ 0& 0& 1& 0\\ 0& 0& 0& 1\end{array}\right];$

4) the covariance matrix P (k | k-1) of forecast period is calculated:

$P\left(k\right|k-1)=A\·P(k-1|k-1)\·A^{\′}+Q=A\·P(k-1|k-1)\·A^{\′}+\left[\begin{array}{cccc}\frac{q}{3}& 0& \frac{q}{2}& 0\\ 0& \frac{q}{3}& 0& \frac{q}{2}\\ \frac{q}{2}& 0& q& 0\\ 0& \frac{q}{2}& 0& q\end{array}\right],$

Wherein P (k-1 | k-1) is the covariance matrix after updating in a upper moment, if k=0, then making P (k | k)=Q, q is one Individual confidence parameter, if not believing that, user is uniform motion, setting big；

5) Kalman gain is calculated；

6) calculating observation resultWithCentered by, 2 × d_kFor radius, referring to Stricture of vagina storehouse calculates and obtains absolute force m with current_iThree most like fingerprints, it adds and meansigma methods conduct

7) more new state, obtains current position location (x_k,y_k)；

8) covariance matrix: P (k | k)=(1-Kg (k)) P (k | k-1) is updated.

The localization method of the present invention realizes in smart mobile phone, and developing operation, at the APP of android system, utilizes The magnetometer of mobile phone obtains absolute force data, utilizes acceleration transducer to obtain and carries out paces detection and estimate with step-length Meter, utilizes gyroscope to obtain the direction of motion and the angle of direct north.For judging precision and the robust of alignment system of the present invention Property, employ five personal hand-held Android smartphone to carry out positioning experiment in an indoor environment, region, location is divided into 1m takes advantage of the grid of 1m, every volunteer's random walk about 500 meters, the average localization error of the present invention about at about 1.20m, and And maximum mean error about 1.35m, the minimum about 1.13m of five volunteer location, substantially meet the need of high accuracy indoor positioning Ask, and easy to spread.

Accompanying drawing explanation

Fig. 1 is intelligent mobile phone sensor coordinate axes schematic diagram.

Fig. 2 is the FB(flow block) of system.

Detailed description of the invention

Below in conjunction with the accompanying drawings the smart phone user indoor orientation method of the earth magnetism correction inertial navigation of the present invention is done further Description.

Smart mobile phone, as a kind of electronic equipment carried with, carries magnetometer, accelerometer and gyroscope etc. many Plant sensor.The present invention obtains geomagnetic field intensity data by the magnetometer that smart mobile phone carries, by accelerometer and top Spiral shell instrument obtains the exercise data of user, utilizes Kalman filtering algorithm merging motion information and Geomagnetism Information, passes through Geomagnetism Information Coupling revises the cumulative errors of inertial navigation.

Mobile phone sensor coordinate is as it is shown in figure 1, with user's hand-held intelligent mobile phone in front, mobile phone y-axis points to the direction of motion And screen upwards illustrates for instantiation, 3-axis acceleration, magnetometer and gyroscope in mobile phone are with about 25Hz's Speed obtains data.Can use peakvalue's checking mechanism that Z axis acceleration information is processed and carry out paces detection, Y-axis is accelerated Degrees of data processes estimates paces length.Off-line phase area grid to be set to size can be between 0.6m to 1m, each finger The time that stricture of vagina gathers ensures that absolute force data are more than 100 groups.The localization method of the present invention is real-time according to these data Ground is estimated customer location and shows, as in figure 2 it is shown, step is as follows:

The tuning on-line stage comprises the following steps:

1) initializing, system obtains initial absolute force information m by magnetometer₀, use m₀With the institute in fingerprint base There is the geomagnetic field intensity meansigma methods of fingerprintComparing, the meansigma methods of three fingerprint positions that will be closest to is as initialization bit Put (x_o,y₀)。

2) user holds mobile phone and starts to walk about, and alignment system can be by the process of accelerometer data and observation, detection Paces to step-length d_kEstimate, by gyro data being processed the angle of direction and the direct north obtaining motion φ_k.The result of regulation one-time positioning is designated as (x_k-1,y_k-1), current positioning result is designated as (x_k,y_k), currently available absolute force For m_k。

3) predictive value of acquisition current state:

$X\left(k\right|k-1)=A\·X(k-1|k-1)=\left[\begin{array}{cccc}1& 0& {\mathrm{sin\φ}}_k& 0\\ 0& 1& {\mathrm{cos\φ}}_k& 0\\ 0& 0& 1& 0\\ 0& 0& 0& 1\end{array}\right]\·\left[\begin{array}{c}{x}_{k-1}\\ y_{k-1}\\ d_k\\ {\mathrm{\φ}}_k\end{array}\right],$

The wherein state of the last location of X (k-1 | k-1), and X (k-1 | k-1)=[x_k-1,y_k-1,d_k,φ_k]^T。X(k|k-1) The state matrix positioned for this.A is state matrix:

$A=\left[\begin{array}{cccc}1& 0& {\mathrm{sin\φ}}_k& 0\\ 0& 1& {\mathrm{cos\φ}}_k& 0\\ 0& 0& 1& 0\\ 0& 0& 0& 1\end{array}\right].$

4) covariance of calculating forecast period:

$P\left(k\right|k-1)=A\·P(k-1|k-1)\·A^{\′}+Q=A\·P(k-1|k-1)\·A^{\′}+\left[\begin{array}{cccc}\frac{q}{3}& 0& \frac{q}{2}& 0\\ 0& \frac{q}{3}& 0& \frac{q}{2}\\ \frac{q}{2}& 0& q& 0\\ 0& \frac{q}{2}& 0& q\end{array}\right].$

Wherein P (k-1 | k-1) is the covariance matrix after updating in a upper moment.In this example, q=1 is set.

5) Kalman gain is calculated:

$Kg\left(k\right)=\frac{P\left(k\right|k-1)\·H^{\′}}{H\·P\left(k\right|k-1)\·H^{\′}+R}.$

Wherein H is for measuring sytem matrix, and R is the covariance matrix of measurement system:

$H=\left[\begin{array}{cccc}1& 0& 0& 0\\ 0& 1& 0& 0\\ 0& 0& 1& 0\\ 0& 0& 0& 1\end{array}\right],R=\left[\begin{array}{cccc}3& 0& 0& 0\\ 0& 3& 0& 0\\ 0& 0& 0.1& 0\\ 0& 0& 0& \frac{20}{180}\mathrm{\π}\end{array}\right].$

6) calculating observation resultWithCentered by, 2 × d_kFor radius, referring to Stricture of vagina storehouse calculates and obtains absolute force m with current_iThree most like fingerprints, it adds and meansigma methods conduct

7) more new state:

X (k | k)=X (k | k-1)+Kg (k) (Z (k)-H X (k | k-1))=[x_k y_k d_i φ_i]^T,

Obtain current position location (x_k,y_k)。

8) covariance matrix: P (k | k)=(1-Kg (k)) P (k | k-1) is updated.

9) pedestrian movement one step again detected when system, repeat (3) to (8) step and obtain position.

Claims (1)

1. a smart phone user indoor orientation method for earth magnetism correction inertial navigation, including off-line phase and on-line stage,

The off-line data collecting stage comprises the following steps:

1) according to building plane map, area to be targeted is divided into little grid, the w grid representation B_w, its grid element center RP is expressed as

2) smart phone user holds the smart mobile phone of built-in magnetometer at each grid element center locality magnetic intensity data one section Time；

3) the absolute force data gathering each grid are averaged, and build off-line earth magnetism fingerprint base and store, off-line earth magnetism Fingerprint base is by N^fIndividual fingerprint is constituted, and each fingerprint is stored asWhereinFor off-line earth magnetism fingerprint Position,Geomagnetic field intensity meansigma methods for this fingerprint；

The tuning on-line stage comprises the following steps:

1) initializing, smart mobile phone obtains initial absolute force information m by magnetometer₀, use m₀With the institute in fingerprint base The geomagnetic field intensity meansigma methods having fingerprint compares, and the meansigma methods of the several fingerprint positions that will be closest to is as initialized location (x_o,y₀)；

2) user holds smart mobile phone and starts to walk about, and by the process of the accelerometer data to smart mobile phone, detection paces are also To step-length d_kEstimate, by gyro data being processed the included angle of direction and the direct north obtaining motion_k, it is stipulated that The result of front one-time positioning is designated as (x_k-1,y_k-1), current positioning result is designated as (x_k,y_k), currently available absolute force is m_k；

3) predictive value of acquisition current state:

$X\left(k\right|k-1)=A\·X(k-1|k-1)=\left[\begin{array}{cccc}1& 0& {\mathrm{sin\φ}}_k& 0\\ 0& 1& {\mathrm{cos\φ}}_k& 0\\ 0& 0& 1& 0\\ 0& 0& 0& 1\end{array}\right]\·\left[\begin{array}{c}{x}_{k-1}\\ y_{k-1}\\ d_k\\ {\mathrm{\φ}}_k\end{array}\right],$

Wherein, and X (k-1 | k-1) it is the last state positioned, X (k-1 | k-1)=[x_k-1,y_k-1,d_k,φ_k]^T, X (k | k-1) be The state matrix of this location, A is state matrix:

$A=\left[\begin{array}{cccc}1& 0& {\mathrm{sin\φ}}_k& 0\\ 0& 1& {\mathrm{cos\φ}}_k& 0\\ 0& 0& 1& 0\\ 0& 0& 0& 1\end{array}\right];$

4) the covariance matrix P (k | k-1) of forecast period is calculated:

$P\left(k\right|k-1)=A\·P(k-1|k-1)\·A^{\′}+Q=A\·P(k-1|k-1)\·A^{\′}+\left[\begin{array}{cccc}\frac{q}{3}& 0& \frac{q}{2}& 0\\ 0& \frac{q}{3}& 0& \frac{q}{2}\\ \frac{q}{2}& 0& q& 0\\ 0& \frac{q}{2}& 0& q\end{array}\right],$

Wherein P (k-1 | k-1) is the covariance matrix after updating in a upper moment, if k=0, then makes P (k | k)=Q, q be one and put Letter parameter, if not believing that, user is uniform motion, setting big；

5) Kalman gain is calculated；

6) calculating observation resultWithCentered by, 2 × d_kFor radius, at fingerprint base Middle calculating obtains absolute force m with current_iThree most like fingerprints, it adds and meansigma methods conduct

7) more new state, obtains current position location (x_k,y_k)；

8) covariance matrix: P (k | k)=(1-Kg (k)) P (k | k-1) is updated.