CN107389060A - The hypercompact combination indoor navigation method of IMU/Wi Fi signals based on CKF - Google Patents

Abstract

The present invention is a kind of hypercompact combination indoor navigation method of IMU/Wi Fi signals based on CKF, is mainly included：Establish the RSS Fingerprinting databases of Wi Fi signals；Positioned using Wi Fi signals, obtain pedestrian position P_Wi‑Fi；Positioned using IMU, obtain the position P where pedestrian_IMU, speed V_IMU, acceleration a and posture A_IMU；The drift of IMU accelerometers is corrected by zero-speed correction, and the speed V after being corrected_JIMU；Found respectively in RSS Fingerprinting databases and P_Wi‑FiAnd P_IMUN nearer reference point, then respectively with P_Wi‑FiAnd P_IMUMake the difference the distance between acquisition measurement point and reference pointAnd then range difference Δ d is obtained againⁱ；Information fusion is carried out using CKF, finally gives position, acceleration, speed and the attitude information of pedestrian.The inventive method using Wi Fi and IMU location information, with reference to volume Kalman filtering algorithm, has the characteristics of filtering accuracy height, fast convergence rate, strong robustness, high navigation accuracy simultaneously.

Description

Super-tight combined indoor navigation method of IMU/Wi-Fi signal based on CKF

Technical Field

The invention belongs to the technical field of navigation, and particularly relates to an IMU/Wi-Fi signal ultra-tight integrated indoor navigation method based on Cubature Kalman Filter (CKF).

Background

The existing pedestrian navigation positioning method is mainly implemented by a Global Positioning System (GPS). However, the GPS positioning has the disadvantages of poor dynamic performance, non-autonomy, and the like. Especially in the environment of urban streets, tunnels, indoor and the like which are easy to shield satellite navigation signals, signal loss can be caused, and even positioning failure can be caused.

Existing indoor navigation methods generally use Micro-electro mechanical systems (MEMS) inertial devices for navigation. However, the pedestrian navigation autonomous positioning system based on the MEMS inertial device has a certain disadvantage that the device error thereof is rapidly dispersed. In the navigation stage, if the error of the MEMS inertial device cannot be effectively compensated, the position error can be dispersed in the third power of time, and finally the navigation function of the system is failed. Therefore, effective correction of errors becomes the biggest difficulty of applying the inertial calculation algorithm based on the MEMS system to the pedestrian navigation system.

Usually, error correction and information fusion of the MEMS system are implemented in a filtering manner. Since a nonlinear model is involved, nonlinear filtering such as extended kalman filtering is generally used in filter selection. The extended kalman filtering is a typical nonlinear filtering method, and a first-order taylor series linear expansion is performed on a nonlinear equation to approximate the real state distribution, but the approximation result is only generated in a small neighborhood, otherwise, a large error is brought, and the filtering result is unstable and even diverged.

In summary, the existing indoor pedestrian combined navigation system has poor stability, low convergence rate, poor filtering precision and weak robustness, so that the pedestrian navigation precision is low.

Disclosure of Invention

The invention aims to provide an IMU/Wi-Fi signal ultra-tight combination indoor navigation method based on CKF, and aims to solve the problems of poor stability and poor filtering precision of the existing indoor pedestrian combination navigation method.

The implementation of the invention comprises the following steps:

the method comprises the following steps: selecting a certain density indoorsMeasure and record the signal strength received from the access point APs and the corresponding location information P for each reference point_DBEstablishing an RSS Fingerprint database;

step two: the IMU is fixed on the foot of the pedestrian, the IMU is used for measuring the acceleration a and the rotating speed omega of the motion information of the pedestrian during the advancing process, and the velocity V of the pedestrian is obtained through inertial navigation calculation_IMUPosition P_IMUAnd attitude A_IMU；

Step three: utilizing the acceleration a obtained in the step two to carry out zero-speed correction on the accelerometer, carrying out inertial navigation solution on the corrected motion information, and obtaining the corrected position velocity V through filtering of the extended filter_JIMUAnd the velocity V obtained in the step two_IMUObtaining a speed error delta V by difference;

step four: using the position information P obtained in step two_IMUPosition information P of reference point in RSS finger rprint database_DBComparing to screen out n points near the measuring pointAnd then respectively with the IMU measured position P_IMUMaking difference to obtain the distance between pedestrian and reference point

To which is relatedThe expression is as follows:

wherein, X_IMUAnd Y_IMUIs measured by IMU to obtain position coordinates P_IMUAnd has a component of_IMU＝{X_IMU,Y_IMU}；Andis the coordinates of the selected reference pointAnd has a component of

Step five: the method comprises the steps of receiving Wi-Fi signals by using a Wi-Fi antenna carried by a pedestrian, measuring the intensity of the received signals, calculating the distance d between a measuring point and an access point APs according to a mathematical model of the signal intensity and the distance, and calculating the position P of the measuring point by using a trilateration method_Wi-FiAnd making a difference with the position measured by the IMU to obtain a position error delta P;

the mathematical model between the referred Wi-Fi signal strength RSS and distance d is:

in the formula (d)₀For a known reference distance, RSS₀To be at a reference distance d₀The average signal strength, p is the signal attenuation exponent;

step six: using the position information P from step five_Wi-FiPosition information P of reference point in RSS finger rprint database_DBComparing to screen out n points near the measuring pointThen respectively locating the obtained positions P with Wi-Fi_Wi-FiMaking difference to obtain the distance between pedestrian and reference pointNamely:

to which is relatedThe expression is as follows:

wherein, X_Wi-FiAnd Y_Wi-FiIs the measured position coordinate P_Wi-FiAnd has a component of_Wi-Fi＝{X_Wi-Fi,Y_Wi-Fi}；Andis the selected reference point coordinatesAnd has a component of

Step seven: the distances obtained in the fourth step and the sixth stepMaking a difference to obtain a distance difference delta dⁱ；

Step eight: using the velocity error Δ V obtained in step three and the position error Δ P obtained in step five as the state quantity X of CKF ═ Δ V, Δ P]Using the difference Δ d obtained in step sevenⁱMeasuring Z ═ d as CKF quantityⁱ]Performing filtering calculation by using CKF;

step nine: correcting the speed V and the position P measured by the IMU by using the filtered speed error delta V and the filtered position error delta P obtained in the step eight, and comparing the filtered position error delta P obtained in the step eight with the position P of the measuring point in the step five_Wi-FiCorrecting to obtain final indoor navigation information of the pedestrian

In the first step, when a reference point is selected, one reference point is taken every 5 meters, one reference point is taken every 10 meters, 5 groups of data are measured in each direction at each reference point, 40 groups of data are measured for processing, and the Wi-Fi signal strength of the reference point is determined.

The invention has the beneficial effects that:

the method has the advantages of high convergence speed, high filtering precision and strong robustness, so that the indoor navigation precision of the pedestrian can be effectively improved.

Drawings

FIG. 1 is a flow chart of an IMU/Wi-Fi signal ultra-tight combination indoor navigation method based on CKF provided by the invention.

FIG. 2 is a block diagram of an IMU/Wi-Fi signal ultra-compact combination indoor navigation system based on CKF provided by the invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

The invention discloses an IMU/Wi-Fi signal ultra-tight combination indoor navigation method based on CKF, which comprises the following steps: establishing an RSS finger printing database of the Wi-Fi signals; utilizing IMU to carry out positioning to obtain the position P of the pedestrian_IMUAnd information such as speed, acceleration, and attitude; correcting the IMU accelerometer drift through zero speed correction, and obtaining corrected speed; positioning by utilizing Wi-Fi signals to obtain the position P of the pedestrian_Wi-Fi(ii) a Constructed numberRespectively finding P in database_IMUCloser n reference points and P_Wi-FiThe n closer reference points respectively obtain the distance between the measuring point and the reference point according to the positions of the known reference pointsThereby obtaining a distance difference Deltadⁱ(ii) a And performing information fusion by using a Cubature Kalman Filter (CKF) to finally obtain the position, acceleration, speed and attitude information of the pedestrian. According to the invention, information fusion is carried out on the pedestrian navigation system which is formed by tightly combining the indoor Wi-Fi signal and the IMU by using the CKF, the Wi-Fi is widely applied indoors and the Wi-Fi signal is relatively stable, and the pedestrian navigation system can be effectively assisted to navigate pedestrians by the inertial navigation system. Compared with EKF, UKF and other filtering algorithms, the CKF algorithm has the advantages of high filtering precision, high convergence speed and strong robustness, so that the method can effectively improve the navigation precision.

The IMU/Wi-Fi signal tight combination indoor navigation method based on the CKF comprises the following steps:

the method comprises the following steps: selecting reference Points with certain density indoors, measuring and recording the Received Signal Strength (Received Signal Strength RSS) and corresponding position information P of each reference point from access Points (Access Points aps)_DBEstablishing an RSS Fingerprint database;

step two: the IMU is fixed on one foot of the pedestrian, the IMU is used for measuring the acceleration a and the rotating speed omega of the movement information of the pedestrian during the advancing process, and the velocity V of the pedestrian is obtained through inertial navigation calculation_IMUPosition P_IMUAnd attitude A_IMU；

Step three: utilizing the acceleration a obtained in the step two to carry out zero-speed correction on the accelerometer, carrying out inertial navigation solution on the corrected motion information, and obtaining the corrected position velocity V through filtering of the extended filter_JIMUAnd the velocity V obtained in the step two_IMUObtaining a speed error delta V by difference;

step four: using the position information P obtained in step two_IMUAnd RSS Fingerposition information P of reference point in print database_DBComparing to screen out n points near the measuring pointAnd then respectively with the IMU measured position P_IMUMaking difference to obtain the distance between pedestrian and reference pointNamely:

wherein, X_IMUAnd Y_IMUIs measured by IMU to obtain position coordinates P_IMUAnd has a component of_IMU＝{X_IMU,Y_IMU}；Andis the coordinates of the selected reference pointAnd has a component of

Step five: the method comprises the steps of receiving Wi-Fi signals by using a Wi-Fi antenna carried by a pedestrian, measuring the intensity of the received signals, calculating the distance d between a measuring point and an access point APs according to a mathematical model of the signal intensity and the distance, and calculating the position P of the measuring point by using a trilateration method_Wi-FiAnd making a difference with the position measured by the IMU to obtain a position error delta P;

by the formula:

establishing a mathematical model between Wi-Fi signal strength RSS and distance d, where d₀For a known reference distance, RSS₀To be at a reference distance d₀Where p is the signal attenuation exponent, v is a gaussian distributed random variable with a mean of 0 and a mean square error σ_RSSIndoor obstacles. The maximum likelihood estimation is carried out on the formula to obtain a mathematical model between the Wi-Fi signal strength RSS and the distance d as follows:

the distance d between the measurement point and the APs can be obtained by the above formula, so that the position P of the measurement point is solved by using the trilateration method_Wi-FiAnd obtaining a position error delta P by subtracting the position measured by the IMU;

step six: using the position information P from step five_Wi-FiPosition information P of reference point in RSS finger rprint database_DBComparing to screen out n points near the measuring pointThen respectively locating the obtained positions P with Wi-Fi_Wi-FiMaking difference to obtain the distance between pedestrian and reference pointNamely:

wherein, X_Wi-FiAnd Y_Wi-FiIs the measured position coordinate P_Wi-FiAnd has a component of_Wi-Fi＝{X_Wi-Fi,Y_Wi-Fi}；Andis the selected reference point coordinatesAnd has a component of

Step seven: the distances obtained in the fourth step and the sixth stepMaking a difference to obtain a distance difference delta dⁱ；

Step eight: using the velocity error Δ V obtained in step three and the position error Δ P obtained in step five as the state quantity X of CKF ═ Δ V, Δ P]Using the difference Δ d obtained in step sevenⁱMeasuring Z ═ d as CKF quantityⁱ]Filtering by using CKF;

step nine: correcting the speed V and the position P measured by the IMU by using the filtered speed error delta V and the filtered position error delta P obtained in the step eight, and comparing the filtered position error delta P obtained in the step eight with the position P of the measuring point in the step five_Wi-FiAnd correcting to obtain final indoor navigation information of the pedestrian.

In the first step, when a reference point is selected, one reference point is taken every 5 meters in a room, one reference point is taken every 10 meters in a corridor, 5 groups of data are measured in each direction at each reference point, 40 groups of data are measured for processing, and the Wi-Fi signal strength of the reference point is determined.

In step eight, a volume kalman filter (CKF) is used for filtering, and the velocity error Δ V and the position error Δ P obtained in step three and step five are used as the state quantity X of the volume kalman filter, [ Δ V, Δ P ═ V]Distance difference Δ d obtained in step sevenⁱMeasurement of Z ═ d as a volumetric Kalman filterⁱ]The processing flow of the CKF algorithm is as follows:

1) time updating

Selecting state volume points:

wherein S is_k-1＝chol{P_k-1Is a matrix P_k-1Cholesky decomposition of (i.e. error covariance matrix)

State volume point propagated through state equation:

and (3) state prediction:

error covariance prediction:

2) measurement update

State volume point:

wherein,

measuring volume points:

Z_j,k＝h(X_j,k)

measurement and prediction:

innovation variance:

covariance matrix:

kalman gain:

and (3) updating the state:

error covariance:

FIG. 1 shows a flow of an IMU/Wi-Fi signal ultra-tight combination indoor navigation method based on CKF provided by the invention. For convenience of explanation, only portions relevant to the present invention are shown.

The IMU/Wi-Fi signal ultra-tight combination indoor navigation method based on the CKF comprises the following steps:

the method comprises the following steps: selecting reference Points with certain density indoors, measuring and recording the Received Signal Strength (Received Signal Strength RSS) and corresponding position information P of each reference point from access Points (Access Points aps)_DBEstablishing an RSS Fingerprint database;

step two: the IMU is fixed on one foot of the pedestrian, the IMU is used for measuring the acceleration a and the rotating speed omega of the movement information of the pedestrian during the advancing process, and the velocity V of the pedestrian is obtained through inertial navigation calculation_IMUPosition P_IMUAnd attitude A_IMU；

Step three: utilizing the acceleration a obtained in the step two to carry out zero-speed correction on the accelerometer, carrying out inertial navigation solution on the corrected motion information, and obtaining the corrected position velocity V through filtering of the extended filter_JIMUAnd the velocity V obtained in the step two_IMUObtaining a speed error delta V by difference;

step four: using the position information P obtained in step two_IMUPosition information P of reference point in RSS finger rprint database_DBComparing to screen out n points near the measuring pointAnd then respectively with the IMU measured position P_IMUMaking difference to obtain the distance between pedestrian and reference pointNamely:

wherein, X_IMUAnd Y_IMUIs measured by IMU to obtain position coordinates P_IMUAnd has a component of_IMU＝{X_IMU,Y_IMU}；Andis the coordinates of the selected reference pointAnd has a component of

Step five: the method comprises the steps of receiving Wi-Fi signals by using a Wi-Fi antenna carried by a pedestrian, measuring the intensity of the received signals, calculating the distance d between a measuring point and an access point APs according to a mathematical model of the signal intensity and the distance, and calculating the position P of the measuring point by using a trilateration method_Wi-FiAnd making a difference with the position measured by the IMU to obtain a position error delta P;

by the formula:

establishing a mathematical model between Wi-Fi signal strength RSS and distance d, where d₀For a known reference distance, RSS₀To be at a reference distance d₀Where p is the signal attenuation exponent, v is a gaussian distributed random variable with a mean of 0 and a mean square error σ_RSSIndoor obstacles. The maximum likelihood estimation is carried out on the formula to obtain a mathematical model between the Wi-Fi signal strength RSS and the distance d as follows:

the distance d between the measurement point and the APs can be obtained by the above formula, so that the position P of the measurement point is solved by using the trilateration method_Wi-FiAnd obtaining a position error delta P by subtracting the position measured by the IMU;

step six: by usingPosition information P obtained in step five_Wi-FiPosition information P of reference point in RSS finger rprint database_DBComparing to screen out n points near the measuring pointThen respectively locating the obtained positions P with Wi-Fi_Wi-FiMaking difference to obtain the distance between pedestrian and reference pointNamely:

wherein, X_Wi-FiAnd Y_Wi-FiIs the measured position coordinate P_Wi-FiAnd has a component of_Wi-Fi＝{X_Wi-Fi,Y_Wi-Fi}；Andis the selected reference point coordinatesAnd has a component of

Step seven: the distances obtained in the fourth step and the sixth stepMaking a difference to obtain a distance difference delta dⁱ；

Step eight: using the velocity error Δ V obtained in step three and the position error Δ P obtained in step five as the state quantity X of CKF ═ Δ V, Δ P]Using the difference Δ d obtained in step sevenⁱAs a measure of CKFAmount Z ═ dⁱ]Filtering by using CKF;

step nine: correcting the speed V and the position P measured by the IMU by using the filtered speed error delta V and the filtered position error delta P obtained in the step eight, and comparing the filtered position error delta P obtained in the step eight with the position P of the measuring point in the step five_Wi-FiAnd correcting to obtain final indoor navigation information of the pedestrian.

As an optimization scheme of the embodiment of the invention, in the step one, when the reference point is selected, one reference point is taken every 5 meters in a room, one reference point is taken every 10 meters in a corridor, 5 groups of data are measured in each direction at each reference point, 40 groups of data are measured for processing, and the Wi-Fi signal strength of the reference point is determined. In order to make the RSS Fingerprint database more accurate, the reference point is measured twice, once when someone walks indoors and once when no one exists indoors.

As an optimized solution of the embodiment of the present invention, in step eight, filtering is performed by using a volume kalman filter (CKF), and the speed error Δ V and the position error Δ P obtained in step three and step five are used as the state quantity X ═ Δ V, Δ P of the volume kalman filter]Distance difference Δ d obtained in step sevenⁱMeasurement of Z ═ d as a volumetric Kalman filterⁱ]The processing flow of the CKF algorithm is as follows:

1) time updating

Selecting state volume points:

wherein S is_k-1＝chol{P_k-1Is a matrix P_k-1Cholesky decomposition of (i.e. error covariance matrix)

State volume point propagated through state equation:

and (3) state prediction:

error covariance prediction:

2) measurement update

State volume point:

wherein,

measuring volume points:

Z_j,k＝h(X_j,k)

measurement and prediction:

innovation variance:

covariance matrix:

kalman gain:

and (3) updating the state:

error covariance:

as shown in fig. 1, the IMU/Wi-Fi signal ultra-tight combination indoor navigation method based on CKF of the embodiment of the present invention includes the following steps:

s101: establishing an RSS finger printing database of the Wi-Fi signals;

s102: the IMU is fixed on one foot of the pedestrian, the IMU is used for measuring the acceleration a and the rotating speed omega of the movement information of the pedestrian during the advancing process, and the velocity V of the pedestrian is obtained through inertial navigation calculation_IMUPosition P_IMUAnd attitude A_IMU；

S103: the obtained acceleration a is used for carrying out zero-speed correction on the accelerometer, inertial navigation calculation is carried out on the corrected motion information, and the corrected position velocity V is obtained through filtering of the expansion filter_JIMUAnd the velocity V obtained in the step two_IMUObtaining a speed error delta V by difference;

s104: using the position information P obtained in step two_IMUParticipating in RSS Fingerprint databasePosition information P of examination point_DBComparing, screening out the point near to the measurement point, and comparing with the position P measured by IMU_IMUMaking a difference to obtain the distance between the pedestrian and a reference point;

s105: the method comprises the steps of receiving Wi-Fi signals by using a Wi-Fi antenna carried by a pedestrian, measuring the intensity of the received signals, calculating the distance d between a measuring point and an access point APs according to a mathematical model of the signal intensity and the distance, and calculating the position P of the measuring point by using a trilateration method_Wi-FiAnd making a difference with the position measured by the IMU to obtain a position error delta P;

s106: using the position information P from step five_Wi-FiPosition information P of reference point in RSS finger rprint database_DBComparing, screening out the point close to the measuring point, and locating with Wi-Fi to obtain position P_Wi-FiMaking a difference to obtain the distance between the pedestrian and a reference point;

s107: the distances obtained in the fourth step and the sixth stepMaking difference to obtain distance difference;

s108: using the obtained speed error Δ V and the obtained position error Δ P as the state quantity X of CKF ═ Δ V, Δ P]Using the resulting difference in distance Δ dⁱMeasuring Z ═ d as CKF quantityⁱ]Filtering by using CKF;

s109: correcting the speed V and the position P measured by the IMU by using the filtered speed error delta V and the filtered position error delta P, and correcting the filtered position error delta P to the position P of the measuring point_Wi-FiAnd correcting to obtain final indoor navigation information of the pedestrian.

The method comprises the following specific steps:

the method comprises the following steps: selecting reference Points with certain density indoors, measuring and recording the Received Signal Strength (Received Signal Strength RSS) and corresponding position of each reference point from access Points (Access Points aps)Letter P_DBEstablishing an RSS Fingerprint database;

when the reference point is selected, one reference point is taken at every 5 steps in a room, one reference point is taken at every 10 steps in a hall, 5 groups of data are measured in each direction at each reference point, 40 groups of data are measured for processing, and finally the Wi-Fi signal strength of the reference point is determined. In order to make the RSS Fingerprint database more accurate, the reference point is measured twice, once when someone walks indoors and once when no one exists indoors.

Step two: the IMU is fixed on one foot of the pedestrian, the IMU is used for measuring the acceleration a and the rotating speed omega of the movement information of the pedestrian during the advancing process, and the velocity V of the pedestrian is obtained through inertial navigation calculation_IMUPosition P_IMUAnd attitude A_IMU；

Step three: utilizing the acceleration a obtained in the step two to carry out zero-speed correction on the accelerometer, carrying out inertial navigation solution on the corrected motion information, and obtaining the corrected position velocity V through filtering of the extended filter_JIMUAnd the velocity V obtained in the step two_IMUObtaining a speed error delta V by difference;

step four: using the position information P obtained in step two_IMUPosition information P of reference point in RSS finger rprint database_DBComparing to screen out n points near the measuring pointAnd then respectively with the IMU measured position P_IMUMaking difference to obtain the distance between pedestrian and reference pointNamely:

wherein, X_IMUAnd Y_IMUIs measured by IMU to obtain position coordinates P_IMUAnd has a component of_IMU＝{X_IMU,Y_IMU}；Andis the coordinates of the selected reference pointAnd has a component of

Step five: the method comprises the steps of receiving Wi-Fi signals by using a Wi-Fi antenna carried by a pedestrian, measuring the intensity of the received signals, calculating the distance d between a measuring point and an access point APs according to a mathematical model of the signal intensity and the distance, and calculating the position P of the measuring point by using a trilateration method_Wi-FiAnd making a difference with the position measured by the IMU to obtain a position error delta P;

by the formula:

establishing a mathematical model between Wi-Fi signal strength RSS and distance d, where d₀For a known reference distance, RSS₀To be at a reference distance d₀Where p is the signal attenuation exponent, v is a gaussian distributed random variable with a mean of 0 and a mean square error σ_RSSIndoor obstacles. The maximum likelihood estimation is carried out on the formula to obtain a mathematical model between the Wi-Fi signal strength RSS and the distance d as follows:

the distance d between the measurement point and the APs can be obtained by the above formula, so that the measurement point is solved by using the trilateration methodPosition P of_Wi-FiAnd obtaining a position error delta P by subtracting the position measured by the IMU;

step six: using the position information P from step five_Wi-FiPosition information P of reference point in RSS finger rprint database_DBComparing to screen out n points near the measuring pointThen respectively locating the obtained positions P with Wi-Fi_Wi-FiMaking difference to obtain the distance between pedestrian and reference pointNamely:

wherein, X_Wi-FiAnd Y_Wi-FiIs the measured position coordinate P_Wi-FiAnd has a component of_Wi-Fi＝{X_Wi-Fi,Y_Wi-Fi}；Andis the selected reference point coordinatesAnd has a component of

Step seven: the distances obtained in the fourth step and the sixth stepMaking a difference to obtain a distance difference delta dⁱ；

Step eight: using the velocity obtained in step threeThe error Δ V and the position error Δ P obtained in step five are used as the state quantity X of CKF ═ Δ V, Δ P]Using the difference Δ d obtained in step sevenⁱMeasuring Z ═ d as CKF quantityⁱ]Filtering by using CKF, wherein the CKF algorithm flow is as follows;

1) time updating

Selecting state volume points:

wherein S is_k-1＝chol{P_k-1Is a matrix P_k-1Cholesky decomposition of (i.e. error covariance matrix)

State volume point propagated through state equation:

and (3) state prediction:

error covariance prediction:

2) measurement update

State volume point:

wherein,

measuring volume points:

Z_j,k＝h(X_j,k)

measurement and prediction:

innovation variance:

covariance matrix:

kalman gain:

and (3) updating the state:

error covariance:

step nine: using the filtered speed error Δ V obtained in step eight,Correcting the speed V and the position P measured by the IMU by the position error delta P, and comparing the position error delta P obtained in the step eight after filtering with the position P of the measuring point in the step five_Wi-FiAnd correcting to obtain final indoor navigation information of the pedestrian.

Claims (2)

1. The super-tight combined indoor navigation method of the IMU/Wi-Fi signal based on the CKF is characterized by comprising the following steps:

the method comprises the following steps: selecting reference points with certain density indoors, measuring and recording the signal intensity received by each reference point from the APs and the corresponding position information P_DBEstablishing an RSS Fingerprint database;

step two: the IMU is fixed on the foot of the pedestrian, the IMU is used for measuring the acceleration a and the rotating speed omega of the motion information of the pedestrian during the advancing process, and the velocity V of the pedestrian is obtained through inertial navigation calculation_IMUPosition of replacementPut P_IMUAnd attitude A_IMU；

Step three: utilizing the acceleration a obtained in the step two to carry out zero-speed correction on the accelerometer, carrying out inertial navigation solution on the corrected motion information, and obtaining the corrected position velocity V through filtering of the extended filter_JIMUAnd the velocity V obtained in the step two_IMUObtaining a speed error delta V by difference;

step four: using the position information P obtained in step two_IMUPosition information P of reference point in RSS finger rprint database_DBComparing to screen out n points near the measuring pointAnd then respectively with the IMU measured position P_IMUMaking difference to obtain the distance between pedestrian and reference pointTo which is relatedThe expression is as follows:

<mrow> <msubsup> <mi>d</mi> <mrow> <mi>I</mi> <mi>M</mi> <mi>U</mi> </mrow> <mi>i</mi> </msubsup> <mo>=</mo> <msqrt> <mrow> <msup> <mrow> <mo>(</mo> <msub> <mi>X</mi> <mrow> <mi>I</mi> <mi>M</mi> <mi>U</mi> </mrow> </msub> <mo>-</mo> <msubsup> <mi>X</mi> <mrow> <mi>D</mi> <mi>B</mi> </mrow> <mi>i</mi> </msubsup> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mo>+</mo> <msup> <mrow> <mo>(</mo> <msub> <mi>Y</mi> <mrow> <mi>I</mi> <mi>M</mi> <mi>U</mi> </mrow> </msub> <mo>-</mo> <msubsup> <mi>Y</mi> <mrow> <mi>D</mi> <mi>B</mi> </mrow> <mi>i</mi> </msubsup> <mo>)</mo> </mrow> <mn>2</mn> </msup> </mrow> </msqrt> </mrow>

wherein, X_IMUAnd Y_IMUIs measured by IMU to obtain position coordinates P_IMUAnd has a component of_IMU＝{X_IMU,Y_IMU}；Andis the coordinates of the selected reference pointAnd has a component of

Step five: the method comprises the steps of receiving Wi-Fi signals by using a Wi-Fi antenna carried by a pedestrian, measuring the intensity of the received signals, calculating the distance d between a measuring point and an access point APs according to a mathematical model of the signal intensity and the distance, and calculating the position P of the measuring point by using a trilateration method_Wi-FiAnd making a difference with the position measured by the IMU to obtain a position error delta P;

the mathematical model between the referred Wi-Fi signal strength RSS and distance d is:

<mrow> <mi>d</mi> <mo>=</mo> <msub> <mi>d</mi> <mn>0</mn> </msub> <mo>&amp;CenterDot;</mo> <msup> <mn>10</mn> <mfrac> <mrow> <msub> <mi>RSS</mi> <mn>0</mn> </msub> <mo>-</mo> <mi>R</mi> <mi>S</mi> <mi>S</mi> </mrow> <mrow> <mn>10</mn> <mo>&amp;CenterDot;</mo> <mi>p</mi> </mrow> </mfrac> </msup> </mrow>

in the formula (d)₀For a known reference distance, RSS₀To be at a reference distance d₀The average signal strength, p is the signal attenuation exponent;

step six: using the position information P from step five_Wi-FiPosition information P of reference point in RSS finger rprint database_DBComparing to screen out n points near the measuring pointThen respectively locating the obtained positions P with Wi-Fi_Wi-FiMaking difference to obtain the distance between pedestrian and reference pointNamely: to which is relatedThe expression is as follows:

<mrow> <msubsup> <mi>d</mi> <mrow> <mi>W</mi> <mi>i</mi> <mo>-</mo> <mi>F</mi> <mi>i</mi> </mrow> <mi>i</mi> </msubsup> <mo>=</mo> <msqrt> <mrow> <msup> <mrow> <mo>(</mo> <msub> <mi>X</mi> <mrow> <mi>W</mi> <mi>i</mi> <mo>-</mo> <mi>F</mi> <mi>i</mi> </mrow> </msub> <mo>-</mo> <msubsup> <mi>X</mi> <mrow> <mi>W</mi> <mi>i</mi> <mo>-</mo> <mi>F</mi> <mi>i</mi> </mrow> <mi>i</mi> </msubsup> <mo>)</mo> </mrow> <mn>2</mn> </msup> <mo>+</mo> <msup> <mrow> <mo>(</mo> <msub> <mi>Y</mi> <mrow> <mi>W</mi> <mi>i</mi> <mo>-</mo> <mi>F</mi> <mi>i</mi> </mrow> </msub> <mo>-</mo> <msubsup> <mi>Y</mi> <mrow> <mi>W</mi> <mi>i</mi> <mo>-</mo> <mi>F</mi> <mi>i</mi> </mrow> <mi>i</mi> </msubsup> <mo>)</mo> </mrow> <mn>2</mn> </msup> </mrow> </msqrt> </mrow>

wherein, X_Wi-FiAnd Y_Wi-FiIs the measured position coordinate P_Wi-FiAnd has a component of_Wi-Fi＝{X_Wi-Fi,Y_Wi-Fi}；Andis the selected reference point coordinatesAnd has a component of

Step seven: the distances obtained in the fourth step and the sixth stepMaking a difference to obtain a distance difference delta dⁱ；

Step eight: using the velocity error Δ V obtained in step three and the position error Δ P obtained in step five as the state quantity X of CKF ═ Δ V, Δ P]Using the difference Δ d obtained in step sevenⁱMeasuring Z ═ d as CKF quantityⁱ]Performing filtering calculation by using CKF;

step nine: correcting the speed V and the position P measured by the IMU by using the filtered speed error delta V and the filtered position error delta P obtained in the step eight, and comparing the filtered position error delta P obtained in the step eight with the position P of the measuring point in the step five_Wi-FiAnd correcting to obtain final indoor navigation information of the pedestrian.

2. The method as claimed in claim 1, wherein in the step one, when selecting the reference point, one reference point is taken every 5 meters, one reference point is taken every 10 meters, and at each reference point, four directions are measured at each reference point, 5 groups of data are measured in each direction, and 40 groups of data are measured for processing, so as to determine the Wi-Fi signal strength of the reference point.