CN114710745A - Indoor positioning method with Bluetooth and PDR information deeply fused - Google Patents

Indoor positioning method with Bluetooth and PDR information deeply fused Download PDF

Info

Publication number
CN114710745A
CN114710745A CN202210378318.8A CN202210378318A CN114710745A CN 114710745 A CN114710745 A CN 114710745A CN 202210378318 A CN202210378318 A CN 202210378318A CN 114710745 A CN114710745 A CN 114710745A
Authority
CN
China
Prior art keywords
bluetooth
particles
particle
positioning
estimated
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
CN202210378318.8A
Other languages
Chinese (zh)
Other versions
CN114710745B (en
Inventor
武畅
孔孝童
袁翼飞
刘禹宏
刘思言
夏堃
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
University of Electronic Science and Technology of China
Original Assignee
University of Electronic Science and Technology of China
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by University of Electronic Science and Technology of China filed Critical University of Electronic Science and Technology of China
Priority to CN202210378318.8A priority Critical patent/CN114710745B/en
Publication of CN114710745A publication Critical patent/CN114710745A/en
Application granted granted Critical
Publication of CN114710745B publication Critical patent/CN114710745B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/02—Services making use of location information
    • H04W4/021—Services related to particular areas, e.g. point of interest [POI] services, venue services or geofences
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
    • G01S5/02—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using radio waves
    • G01S5/0257—Hybrid positioning
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
    • G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
    • G01S5/02—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using radio waves
    • G01S5/14—Determining absolute distances from a plurality of spaced points of known location
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/02—Services making use of location information
    • H04W4/023—Services making use of location information using mutual or relative location information between multiple location based services [LBS] targets or of distance thresholds
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/02—Services making use of location information
    • H04W4/025—Services making use of location information using location based information parameters
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/30—Services specially adapted for particular environments, situations or purposes
    • H04W4/33—Services specially adapted for particular environments, situations or purposes for indoor environments, e.g. buildings
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/80—Services using short range communication, e.g. near-field communication [NFC], radio-frequency identification [RFID] or low energy communication
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Landscapes

  • Engineering & Computer Science (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Signal Processing (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Traffic Control Systems (AREA)
  • Position Fixing By Use Of Radio Waves (AREA)

Abstract

The invention provides an indoor positioning method for deep fusion of Bluetooth and PDR information, which belongs to the technical field of Bluetooth positioning and comprises the following steps: collecting Bluetooth data in a scene; processing the acquired RSSI of the Bluetooth, and sequencing and ranging the RSSI to obtain the ranging amount; generating a novel measurement error model, a state equation of particle filtering and an observation equation by using the distance measurement; performing first heavy particle filtering by using estimated particles of particle filtering at the previous moment to generate M position particles; calculating the advancing direction and the step length through sensor data, generating N step particles by using a PDR algorithm through second particle filtering, and completely predicting the number of the particles N × M; and randomly resampling all the predicted particles to generate estimated particles, and weighting the positions of the estimated particles to obtain the positioning positions. Compared with a Bluetooth triangulation algorithm, the fusion positioning algorithm has the advantages of better robustness, higher positioning precision and lower implementation cost.

Description

Indoor positioning method with Bluetooth and PDR information deeply fused
Technical Field
The invention belongs to the technical field of Bluetooth indoor positioning, and particularly relates to an indoor positioning method for deep fusion of Bluetooth and PDR information.
Background
Currently, the commonly used bluetooth indoor positioning algorithm includes a position fingerprint method and a Triangulation (TRI) method. The position fingerprinting method needs to collect the RSSI (Received Signal Strength) of bluetooth in a positioning scene at an off-line stage and establish a fingerprint database, which results in that a great deal of manpower is consumed to establish the fingerprint database at an earlier stage. The triangulation method needs to receive at least three pieces of Bluetooth information for each positioning, and when the deployment density of the Bluetooth is low, the number of the received Bluetooth is less than three, and the positioning cannot be realized. Aiming at the problems of the positioning algorithm, the invention provides an indoor positioning method with the Bluetooth and PDR information depth fusion, in order to solve the positioning problem of scenes with low Bluetooth deployment density and realize high-precision positioning of various practical scenes. Through an adaptive dual particle filter algorithm formed by position particle filtering and step particle filtering, the Bluetooth information (Bluetooth position and RSSI ranging quantity) and the PDR algorithm are deeply fused, and high-precision positioning in a scene with low Bluetooth deployment density is realized.
Disclosure of Invention
Aiming at the defects of the traditional positioning algorithm, the indoor positioning method with the Bluetooth and PDR information depth integration provided by the invention has the advantage that the positioning accuracy is improved compared with that of the traditional triangulation positioning algorithm.
In order to achieve the above purpose, the invention adopts the technical scheme that:
the scheme provides an indoor positioning method with Bluetooth and PDR information depth fusion, which is characterized by comprising the following steps:
s1, collecting Bluetooth data in a positioning scene through the mobile terminal;
s2, processing the collected RSSI data of the Bluetooth, and sequencing and ranging the RSSIs to obtain the ranging amount;
s3, taking the distance measurement obtained by calculation as input, and generating a novel measurement error model, a state equation and an observation equation of particle filtering;
s4, taking the estimated particles of the particle filter at the previous moment as input, and carrying out first heavy particle filter to generate M position particles;
s5, calculating the advancing direction and the step length through the sensor data of the mobile terminal, taking the first heavy position particles, the advancing direction and the step length as input, generating N step particles by the second heavy particle filter by using a PDR algorithm, and completely predicting the number N M of the particles;
and S6, randomly resampling all the predicted particles to generate estimated particles, and weighting and calculating the positions of the estimated particles to obtain the positioning positions.
Further, the data of bluetooth collected in step S1 is as follows:
in the off-line stage, the information of the bluetooth deployed in the positioning scene and the information in the environment are collected through the mobile terminal, (UUID, Major, Minor, alpha, P)cali，η，PBLE(x, y)), wherein UUID, Major and Minor are Bluetooth IDs for distinguishing different Bluetooth, alpha is local declination, and P is local declinationcaliIs the received power measured at a distance of 1 meter from the Bluetooth, η is the path loss factor in the positioning scenario, PBLE(x, y) are the position coordinates of Bluetooth.
Still further, the step S2 includes the steps of:
s201, in an online stage, a mobile terminal scans Bluetooth in a positioning scene, and acquires data of the corresponding Bluetooth and the positioning scene from a database;
s202, filtering the Received RSSI (Received Signal Strength) data of the bluetooth, where the size of the filtering window is n, as shown in the following formula:
Figure BDA0003591076830000021
s203, sorting the filtered RSSI values, selecting the Bluetooth with the maximum RSSI value, and calculating the distance d between the mobile terminal and the Bluetooth by using an RSSI ranging formulaBLEThe Bluetooth coordinate is PBLE(x，y)：
Figure BDA0003591076830000031
PcaliA received power at a distance of 1 m from the transmitting antenna isBLEHas a received power of PrAnd η is the path loss factor.
Still further, the step S3 includes the steps of:
s301, resolving the advancing direction and the step length of the Pedestrian by using sensor data of the mobile terminal, and performing PDR (Pedestrian Dead Reckoning) estimation to obtain the relative position of the Pedestrian, wherein the relative position is shown as the following formula:
Figure BDA0003591076830000032
Figure BDA0003591076830000033
wherein SLkIs the step size, yaw, of the pedestrian at the k-th momentkIs the advancing direction of the pedestrian at the k-th time, F (S)k) Is the displacement of the pedestrian at the k-th moment;
s302, a state equation and an observation equation of Particle Filter (PF) are shown as follows:
Figure BDA0003591076830000034
wherein, Pk+1＝[Xk+1 Yk+1]TIs a priori estimate of the pedestrian location, i.e. the predicted particle,
Figure BDA0003591076830000035
is the estimated particle of the dual particle filter at time k, Q is the process noise, obeys zero mean, and has a variance of
Figure BDA0003591076830000036
Gaussian distribution of (Z)k+1Is an observed value, is a Bluetooth RSSI ranging value dBLE，PBLEIs a bluetooth coordinate, R is the observation noise;
s303, in order to make the particle filter algorithm more robust, the invention provides a novel measurement error model which can adaptively adjust the variance of observation noise according to an observation value, so that the robustness and the precision of the algorithm are improved, and the probability density of the observation noise is shown as the following formula:
Figure BDA0003591076830000041
wherein Z is an independent variable of the probability density of the observed noise, and is a distance measurement quantity, ZmaxIs the maximum Bluetooth RSSI ranging value, μRIn order to observe the mean value of the noise,
Figure BDA0003591076830000042
to observe the variance of the noise.
Still further, the expression that the first heavy particle filter of step S4 generates M position particles is as follows:
Figure BDA0003591076830000043
Figure BDA0003591076830000044
wherein the content of the first and second substances,
Figure BDA0003591076830000045
is performed by a dual particle filtering algorithm at the kth timeAnd generating estimated particles, sorting according to the weight of the particles, and selecting the first M particles as position particles of the first heavy particle filter at the k +1 th moment.
Still further, the step S5 includes the steps of:
s501, calculating the advancing direction yaw of the k-th time of the pedestrian by using the sensor information of the mobile terminalkSum step SLk；
S502, particles in advancing direction
Figure BDA0003591076830000046
And step size particle
Figure BDA0003591076830000047
Are respectively obeys
Figure BDA0003591076830000048
And
Figure BDA0003591076830000049
is generated from the normal distribution random number to generate second-order particle-filtered step particles
Figure BDA00035910768300000410
As shown in the following formula:
Figure BDA00035910768300000411
s503, generating the predicted particle at the k +1 th time
Figure BDA00035910768300000412
Figure BDA00035910768300000413
For the M position particles of the first heavy particle filter, N step particles are fused, respectively, to generate L new predicted particles, as shown in the following formula:
Figure BDA0003591076830000051
s504, updating the weight of the particles
Figure BDA0003591076830000052
Estimating weights of the particles for the first M times at the k-th time, and normalizing the weights of the particles, wherein
Figure BDA0003591076830000053
To observe the probability density function of the noise R, the following equation is shown:
Figure BDA0003591076830000054
still further, the step S6 includes the steps of:
s601, in order to prevent the particle degradation, randomly resampling all the predicted particles at the k +1 th time to generate estimated particles at the k +1 th time
Figure BDA0003591076830000055
Generating a section { [ cdf (i-1), cdf (i) } according to the following formula]l1, 2, …, L, and each time a uniformly distributed random number between (0, 1) is generated, the particles of the corresponding interval are copied according to the interval in which the random number is located, and finally, all the particle weights are set to be 1, 2, …, L
Figure BDA0003591076830000056
Figure BDA0003591076830000057
S602, counting the number of times each particle is copied, and recording the number of times of copying as { clAnd L is 1, 2 … L, which is sorted from large to small, the first M estimated particles at the k +1 th time are selected as the position particles of the first-weight particle filter at the next time, and the weights of the first M particles are stored as the current estimated particle weight multiplied by the number of times of particle copy, as shown in the following formula:
Figure BDA0003591076830000058
Figure BDA0003591076830000059
s603, calculating the final estimation Position of the particle Position at the k +1 th momentk+1And also the estimated position of the pedestrian at the k +1 th time, as shown in the following formula:
Figure BDA0003591076830000061
the invention has the beneficial effects that: the invention provides an indoor positioning method for deep fusion of Bluetooth and PDR information. The Bluetooth information, the Bluetooth distance measurement amount and the PDR calculation result are subjected to deep fusion through an adaptive dual particle filtering algorithm formed by position particle filtering and step particle filtering, the dual particles are formed by the position particle filtering and the step particle filtering, the position state and the step state of walking of pedestrians are simulated, and the Bluetooth, the advancing direction of the pedestrians and the step information are deeply fused. Meanwhile, the observation noise variance is adaptively adjusted by utilizing the Bluetooth distance measurement amount, so that the walking state of a pedestrian is better simulated by a particle filter algorithm, and the high-precision positioning of a Bluetooth deployment sparse scene is realized.
Drawings
FIG. 1 is a flow chart of the method of the present invention.
Fig. 2 is a specific flow of an indoor positioning method with bluetooth and PDR information deeply integrated.
Fig. 3 is a schematic diagram of a positioning result using the fusion algorithm of the present application in this embodiment.
Fig. 4 is a positioning error map using the fusion algorithm of the present application in this embodiment.
Detailed Description
The following description of the embodiments of the present invention is provided to facilitate the understanding of the present invention by those skilled in the art, but it should be understood that the present invention is not limited to the scope of the embodiments, and it will be apparent to those skilled in the art that various changes may be made without departing from the spirit and scope of the invention as defined and defined in the appended claims, and all matters produced by the invention using the inventive concept are protected.
Examples
The invention provides an indoor positioning method with deep fusion of Bluetooth and PDR information. The method of the invention has two stages: an offline phase and an online phase. In the off-line stage, the bluetooth information and the environmental information in the positioning scene are collected and stored in the database. And in the online positioning stage, filtering the received RSSI data of the Bluetooth, and then ranging by using a ranging formula to obtain an observed value of the distance between the Bluetooth and the mobile terminal. Secondly, the advancing direction and the step length of the pedestrian are obtained through a mobile terminal sensor. And finally, generating predicted particles by utilizing an adaptive dual particle filter algorithm and combining the estimated particles at the previous moment and the step particles at the moment, fusing the Bluetooth RSSI ranging quantity, correcting the predicted particles, randomly resampling the predicted particles, generating the estimated particles, weighting the estimated particles, and finally obtaining the estimated position of the pedestrian. Meanwhile, in order to improve the robustness of the algorithm, the invention also provides a novel measurement error model, the noise variance is observed through adaptive adjustment of particle filtering, and the invention can utilize the information of one or more blueteeth in each positioning, thereby reducing the number of the blueteeth in a positioning scene and reducing the positioning cost. As shown in fig. 1, the present invention provides an indoor positioning method with deep integration of bluetooth and PDR information, and the implementation method thereof is as follows:
s1, collecting Bluetooth data in the positioning scene through the mobile terminal, wherein the Bluetooth data collected in S1 are as follows:
in an off-line stage, the information of the Bluetooth deployed in a positioning scene and the information in the environment are collected through a mobile terminal, (UUID, Major, Minor, alpha, P)cali，η，PBLE(x, y)), wherein UUID, Major and Minor are Bluetooth identifications for distinguishing different Bluetooth, alpha is local declination, and P is local declinationcaliIs measured at a distance of 1 meter from BluetoothEta is the path loss factor in the positioning scenario, PBLE(x, y) are the position coordinates of Bluetooth.
S2, processing the collected RSSI data of the Bluetooth, sequencing and ranging the RSSI to obtain the ranging amount, wherein the realization method comprises the following steps:
s201, in an online stage, a mobile terminal scans Bluetooth in a positioning scene, and acquires data of the corresponding Bluetooth and the positioning scene from a database;
s202, filtering the Received RSSI (Received Signal Strength) data of the bluetooth, where the size of the filtering window is n, as shown in the following formula:
Figure BDA0003591076830000081
s203, sorting the filtered RSSI values, selecting the Bluetooth with the maximum RSSI value, and calculating the distance d between the mobile terminal and the Bluetooth by using an RSSI ranging formulaBLEThe Bluetooth coordinate is PBLE(x，y)：
Figure BDA0003591076830000082
PcaliA received power at a distance of 1 m from the transmitting antenna isBLEHas a received power of PrAnd η is the path loss factor.
S3, taking the distance measurement obtained by calculation as input, generating a novel measurement error model, a state equation and an observation equation of particle filtering, and the implementation method comprises the following steps:
s301, resolving the advancing direction and the step length of the Pedestrian by using sensor data of the mobile terminal, and performing PDR (Pedestrian Dead Reckoning) estimation to obtain the relative position of the Pedestrian, wherein the relative position is shown as the following formula:
Figure BDA0003591076830000083
Figure BDA0003591076830000084
wherein SLkIs the step size, yaw, of the pedestrian at the k-th momentkIs the advancing direction of the pedestrian at the k-th time, F (S)k) Is the displacement of the pedestrian at the k-th moment;
s302, a state equation and an observation equation of Particle Filter (PF) are shown as follows:
Figure BDA0003591076830000085
wherein, Pk+1＝[Xk+1 Yk+1]TIs a priori estimate of the pedestrian location, i.e. the predicted particle,
Figure BDA0003591076830000091
is the estimated particle of the dual particle filter at time k, Q is the process noise, obeys zero mean, and has a variance of
Figure BDA0003591076830000092
Gaussian distribution of (Z)k+1Is an observed value, is a Bluetooth RSSI ranging value dBLE，PBLEIs a bluetooth coordinate, R is the observation noise;
s303, in order to make the particle filter algorithm more robust, the invention provides a novel measurement error model which can adaptively adjust the variance of observation noise according to an observation value, so that the robustness and the precision of the algorithm are improved, and the probability density of the observation noise is shown as the following formula:
Figure BDA0003591076830000093
wherein Z is an independent variable of the probability density of the observed noise, and is a distance measurement quantity, ZmaxIs the maximum Bluetooth RSSI ranging value, μRIn order to observe the mean value of the noise,
Figure BDA0003591076830000094
to observe the variance of the noise.
In this embodiment, the conventional bluetooth triangulation algorithm can perform positioning only after receiving at least 3 bluetooth for each positioning, and cannot perform positioning if the number of the received bluetooth is less than 3. According to the indoor positioning method with the Bluetooth and PDR information deeply fused, the Bluetooth RSSI ranging amount and the PDR algorithm are fused by using the double particle filter algorithm and the novel measurement error model provided by the invention, so that the positioning accuracy and the algorithm robustness are improved, the number of the Bluetooth required by positioning is reduced, only one Bluetooth is needed during each positioning, the implementation cost is reduced, and the positioning problem of a scene with low Bluetooth coverage density is effectively solved.
S4, taking the estimated particles of the particle filter at the previous moment as input, carrying out first heavy particle filter to generate M position particles, wherein the expression is as follows:
Figure BDA0003591076830000095
wherein, the first and the second end of the pipe are connected with each other,
Figure BDA0003591076830000096
the method comprises the steps of generating estimated particles by a dual particle filter algorithm at the kth moment, then sorting according to particle weights, and selecting the first M particles as position particles of the first heavy particle filter at the kth +1 moment.
In this embodiment, in order to improve the lack of the prior particle information in the particle filter algorithm, a dual particle filter algorithm is proposed, in which the first M particles updated in the last particle filter are used as the first heavy particles of the current particle filter. On the basis, N step particles are generated by using a PDR algorithm, and M x N position particles are generated in total, so that the prior particle information of the particles estimated at the last moment is fully utilized, and the precision and the robustness of the traditional particle filter algorithm are improved.
S5, calculating the advancing direction and the step length through the sensor data of the mobile terminal, taking the first heavy position particles, the advancing direction and the step length as input, generating N step particles by the second heavy particle filter by using a PDR algorithm, and totally predicting the number of particles N M, wherein the implementation method comprises the following steps:
s501, calculating the advancing direction yaw of the pedestrian at the k-th moment by using the sensor information of the mobile terminalkAnd step size SLk；
S502, particles in advancing direction
Figure BDA0003591076830000101
And step size particles
Figure BDA0003591076830000102
Are respectively obeys
Figure BDA0003591076830000103
And
Figure BDA0003591076830000104
is generated from a normal distribution random number, generates second-order particle-filtered step particles
Figure BDA0003591076830000105
As shown in the following formula:
Figure BDA0003591076830000106
s503, generating the predicted particle at the k +1 th time
Figure BDA0003591076830000107
Figure BDA0003591076830000108
For the M position particles of the first heavy particle filter, N step particles are fused, respectively, to generate L new predicted particles, as shown in the following formula:
Figure BDA0003591076830000109
s504, updating the particle weight
Figure BDA00035910768300001010
Estimating weights of the particles for the first M times at the k-th time, and normalizing the weights of the particles, wherein
Figure BDA00035910768300001011
To observe the probability density function of noise R, as shown below:
Figure BDA00035910768300001012
s6, randomly resampling all the predicted particles to generate estimated particles, weighting and calculating the positions of the estimated particles to obtain the positioning positions, wherein the implementation method comprises the following steps:
s601, in order to prevent particle degradation, all predicted particles at the k +1 th time are randomly re-sampled to generate estimated particles at the k +1 th time
Figure BDA0003591076830000111
Generating a section { [ cdf (i-1), cdf (i) } according to the following formula]l1, 2, …, L, and each time a uniformly distributed random number between (0, 1) is generated, the particles of the corresponding interval are copied according to the interval in which the random number is located, and finally, all the particle weights are set to be 1, 2, …, L
Figure BDA0003591076830000112
Figure BDA0003591076830000113
S602, counting the number of times each particle is copied, and recording the number of times of copying as { clAnd L is 1, 2 … L, the first M estimated particles at the k +1 th moment are selected as position particles of the first weight particle filter at the next moment in descending order, and the weights of the first M estimated particles are stored as the weight of the current estimated particle multiplied by the copy times of the particlesAs shown in the following formula:
Figure BDA0003591076830000114
Figure BDA0003591076830000115
s603, calculating the final estimation Position of the particle Position at the k +1 th momentk+1And also the estimated position of the pedestrian at the k +1 th time, as shown in the following formula:
Figure BDA0003591076830000116
in this embodiment, a specific flow of an indoor positioning method with deep integration of bluetooth and PDR information is shown in fig. 2. The method comprises the steps of taking the position of a particle as a state quantity and taking a Bluetooth ranging value as an observed quantity, generating predicted particles through a PDR algorithm, correcting an accumulated error of the PDR through the Bluetooth ranging value, and meanwhile, in order to enable the algorithm to be more robust, providing a novel measurement error model, adaptively adjusting the variance of observation noise according to the Bluetooth ranging value, improving the accuracy and robustness of the algorithm, and solving the problem of the accumulated error of the PDR. In addition, the positioning algorithm can utilize single or multiple blueteeth to position each time, thereby solving the problem that the traditional triangulation positioning algorithm cannot effectively position when the blueteeth are less and also reducing the positioning cost.
In order to verify the performance of an indoor positioning method with deep integration of Bluetooth and PDR information, a square field with the length of 25 meters is adopted as an experimental scene for positioning to carry out experiments. The performance of the bluetooth triangulation location algorithm, the PDR algorithm, and the indoor location method based on the bluetooth and PDR information deep fusion provided by the present invention was tested experimentally, and a schematic diagram of a location result using the fusion algorithm of the present application in this embodiment is shown in fig. 3. The positioning precision of the Bluetooth triangulation algorithm is highly related to the deployment density of Bluetooth, and 11 Bluetooth test Bluetooth devices are used in the experimentThe performance of the triangulation algorithm, but the fusion algorithm experiment provided by the invention only uses 4 blueteeth, which is less than that of the triangulation algorithm, but the positioning accuracy is obviously higher than that of the triangulation algorithm. In the first heavy particle filtering of the indoor positioning method with the Bluetooth and PDR information deeply fused, the number M of position particles is 3, and in the second heavy particle filtering, the number of step particles is 30, and the total number of the particles is 90. By passing
Figure BDA0003591076830000121
The positioning error of the positioning algorithm is calculated,
Figure BDA0003591076830000122
is the estimated position of the positioning algorithm, (X)real，Yreal) For the true position, a positioning error map using the fusion algorithm of the present application in this embodiment is shown in fig. 4. The algorithm provided by the invention only uses 4 Bluetooth for positioning in a test scene, and the probability of positioning error less than 2 meters is 72.4%. The PDR algorithm completely depends on an inertial sensor built in the mobile terminal to realize autonomous positioning, so that accumulated errors can occur, the positioning effect is poor, and the maximum positioning error is 7 meters. The number of the used Bluetooth is 7 more than that of the algorithm provided by the invention in the Bluetooth triangulation positioning algorithm, but the positioning effect is still unsatisfactory, and the positioning effect is obviously worse than that of the fusion algorithm provided by the invention. Therefore, the indoor positioning method with the Bluetooth and PDR information deeply integrated provided by the invention not only reduces the requirement on the Bluetooth deployment density in a positioning scene, reduces the number of the Bluetooth used for positioning and reduces the implementation cost, but also has higher positioning accuracy compared with a Bluetooth triangulation algorithm and a PDR algorithm, solves the problem that the Bluetooth triangulation algorithm cannot position in a Bluetooth deployment sparse scene, and realizes high-accuracy positioning in the Bluetooth deployment sparse scene.

Claims (7)

1. An indoor positioning method for deep fusion of Bluetooth and PDR information is characterized by comprising the following steps:
s1, positioning Bluetooth data in a scene by collecting;
s2, processing the collected RSSI data of the Bluetooth, sequencing the RSSIs and measuring the distance to obtain the distance measurement amount;
s3, taking the distance measurement obtained by calculation as input, and generating a novel measurement error model, a state equation and an observation equation of particle filtering;
s4, taking the estimated particles of the particle filter at the previous moment as input, and carrying out first heavy particle filter to generate M position particles;
s5, calculating the advancing direction and the step length through sensor data, taking the first-weight position particles, the advancing direction and the step length as input, generating N step particles by the second-weight particle filter through a PDR algorithm, and completely predicting the particle number N x M;
and S6, randomly resampling all the predicted particles to generate estimated particles, and weighting and calculating the positions of the estimated particles to obtain the positioning positions.
2. The method for deep-blending Bluetooth and PDR information indoor positioning according to claim 1, wherein the data of Bluetooth collected in step S1 is as follows:
in the off-line stage, the information of the deployed Bluetooth in the positioning scene and the information in the environment are collected, (UUID, Major, Minor, alpha, P)cali，η，PBLE(x, y)), wherein UUID, Major and Minor are Bluetooth IDs for distinguishing different Bluetooth, alpha is local declination, and P is local declinationcaliIs the received power measured 1 meter from the Bluetooth, η is the path loss factor in the positioning scenario, PBLE(x, y) are the position coordinates of Bluetooth.
3. The method for indoor positioning with bluetooth and PDR information depth fusion as claimed in claim 2, wherein the step S2 is as follows:
s201, in an online stage, scanning Bluetooth in a positioning scene, and acquiring data of the corresponding Bluetooth and the positioning scene from a database;
s202, filtering Received RSSI (Received Signal Strength) data of the bluetooth, where a size of a filtering window is n, and is shown as follows:
Figure FDA0003591076820000021
s203, sorting the filtered RSSI values, selecting the Bluetooth with the maximum RSSI value, and calculating the distance d between the pedestrian and the Bluetooth by using an RSSI ranging formulaBLEThe Bluetooth coordinate is PBLE(x，y)：
Figure FDA0003591076820000022
PcaliA received power at a distance of 1 m from the transmitting antenna isBLEHas a received power of PrAnd η is the path loss factor.
4. The method for indoor positioning with bluetooth and PDR information depth fusion as claimed in claim 3, wherein said step S3 includes the following steps:
s301, resolving the advancing direction and the step length of the Pedestrian by using sensor data, and carrying out PDR (Pedestrian Dead Reckoning) estimation to obtain the relative position of the Pedestrian, wherein the relative position is shown as the following formula:
Figure FDA0003591076820000023
Figure FDA0003591076820000024
wherein SLkIs the step size, yaw, of the pedestrian at the k-th momentkIs the advancing direction of the pedestrian at the k-th time, F (S)k) Is the displacement of the pedestrian at the k-th moment;
s302, a state equation and an observation equation of Particle Filter (PF) are shown as follows:
Figure FDA0003591076820000031
wherein, Pk+1＝[Xk+1 Yk+1]TIs a priori estimate of the pedestrian location, i.e. the predicted particle,
Figure FDA0003591076820000032
is the estimated particle of the dual particle filter at time k, Q is the process noise, obeys zero mean, and has a variance of
Figure FDA0003591076820000033
Gaussian distribution of (Z)k+1Is an observed value, is a Bluetooth RSSI ranging value dBLE，PBLEIs a bluetooth coordinate, R is the observation noise;
s303, in order to make the particle filter algorithm more robust, the invention provides a novel measurement error model which can adaptively adjust the variance of observation noise according to an observation value, so that the robustness and the precision of the algorithm are improved, and the probability density of the observation noise is shown as the following formula:
Figure FDA0003591076820000034
wherein Z is an independent variable of the probability density of the observed noise, and is a distance measurement quantity, ZmaxIs the maximum Bluetooth RSSI ranging value, μRIn order to observe the mean value of the noise,
Figure FDA0003591076820000035
to observe the variance of the noise.
5. The method for indoor positioning with bluetooth and PDR information depth fusion as claimed in claim 1, wherein the expression of the step S4 that the first heavy particle filter generates M position particles is as follows:
Figure FDA0003591076820000036
wherein the content of the first and second substances,
Figure FDA0003591076820000037
the estimated particles are generated by a double particle filter algorithm at the kth moment, then, sorting is carried out according to the weight of the particles, and the first M particles are selected as position particles of the first heavy particle filter at the kth +1 moment.
6. The method for indoor positioning with bluetooth and PDR information depth fusion as claimed in claim 5, wherein said step S5 includes the following steps:
s501, calculating the advancing direction yaw of the k-th time of the pedestrian by using the sensor informationkSum step SLk；
S502, particles in advancing direction
Figure FDA0003591076820000041
And step size particle
Figure FDA0003591076820000042
Are respectively obeys
Figure FDA0003591076820000043
And
Figure FDA0003591076820000044
is generated from the normal distribution random number to generate second-order particle-filtered step particles
Figure FDA0003591076820000045
As shown in the following formula:
Figure FDA0003591076820000046
s503, generating the predicted particle at the k +1 th time
Figure FDA0003591076820000047
Figure FDA0003591076820000048
For the M position particles of the first heavy particle filter, N step particles are fused, respectively, to generate L new predicted particles, as shown in the following formula:
Figure FDA0003591076820000049
s504, updating the particle weight
Figure FDA00035910768200000410
Figure FDA00035910768200000411
Estimating weights of the particles for the first M times at the k-th time, and normalizing the weights of the particles, wherein
Figure FDA00035910768200000412
To observe the probability density function of the noise R, the following equation is shown:
Figure FDA00035910768200000413
7. the method for indoor positioning with deep fusion of bluetooth and PDR information according to claim 6, wherein the step S6 comprises the following steps:
s601, in order to prevent the particle degradation, randomly resampling all the predicted particles at the k +1 th time to generate estimated particles at the k +1 th time
Figure FDA00035910768200000414
Generating a section { [ cdf (i-1), cdf (i) } according to the following formula]lAnd L is 1, 2, …, L, and each time a uniformly distributed random number between (0, 1) is generated, the particles of the corresponding section are copied according to the section where the random number is located, and finally, all the particle weights are set to be 1, 2, …, L
Figure FDA00035910768200000415
Figure FDA00035910768200000416
S602, counting the copied times of each particle, and marking the copied times as { c }lAnd L is 1, 2 … L, which is sorted from large to small, the first M estimated particles at the k +1 th time are selected as the position particles of the first-weight particle filter at the next time, and the weights of the first M particles are stored as the current estimated particle weight multiplied by the number of times of particle copy, as shown in the following formula:
Figure FDA0003591076820000051
Figure FDA0003591076820000052
s603, calculating the final estimation Position of the particle Position at the k +1 th momentk+1And also the estimated position of the pedestrian at the k +1 th time, as shown in the following formula:
Figure FDA0003591076820000053
CN202210378318.8A 2022-04-12 2022-04-12 Indoor positioning method with Bluetooth and PDR information deeply fused Active CN114710745B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202210378318.8A CN114710745B (en) 2022-04-12 2022-04-12 Indoor positioning method with Bluetooth and PDR information deeply fused

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202210378318.8A CN114710745B (en) 2022-04-12 2022-04-12 Indoor positioning method with Bluetooth and PDR information deeply fused

Publications (2)

Publication Number Publication Date
CN114710745A true CN114710745A (en) 2022-07-05
CN114710745B CN114710745B (en) 2023-04-18

Family

ID=82172498

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202210378318.8A Active CN114710745B (en) 2022-04-12 2022-04-12 Indoor positioning method with Bluetooth and PDR information deeply fused

Country Status (1)

Country Link
CN (1) CN114710745B (en)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2017173001A (en) * 2016-03-18 2017-09-28 国立大学法人東北大学 Information terminal, position estimation method, and position estimation program
CN108632761A (en) * 2018-04-20 2018-10-09 西安交通大学 A kind of indoor orientation method based on particle filter algorithm
CN109298389A (en) * 2018-08-29 2019-02-01 东南大学 Indoor pedestrian based on multiparticle group optimization combines position and orientation estimation method
CN112333818A (en) * 2020-10-27 2021-02-05 中南民族大学 Multi-source fusion indoor positioning system and method based on self-adaptive periodic particle filtering
CN113155131A (en) * 2021-04-14 2021-07-23 浙江工业大学 Particle filter-based iBeacon and PDR fusion indoor positioning method
CN113566820A (en) * 2021-06-17 2021-10-29 电子科技大学 Fusion pedestrian positioning method based on position fingerprint and PDR algorithm

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2017173001A (en) * 2016-03-18 2017-09-28 国立大学法人東北大学 Information terminal, position estimation method, and position estimation program
CN108632761A (en) * 2018-04-20 2018-10-09 西安交通大学 A kind of indoor orientation method based on particle filter algorithm
CN109298389A (en) * 2018-08-29 2019-02-01 东南大学 Indoor pedestrian based on multiparticle group optimization combines position and orientation estimation method
CN112333818A (en) * 2020-10-27 2021-02-05 中南民族大学 Multi-source fusion indoor positioning system and method based on self-adaptive periodic particle filtering
CN113155131A (en) * 2021-04-14 2021-07-23 浙江工业大学 Particle filter-based iBeacon and PDR fusion indoor positioning method
CN113566820A (en) * 2021-06-17 2021-10-29 电子科技大学 Fusion pedestrian positioning method based on position fingerprint and PDR algorithm

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
YIJIE ZHU: "Indoor Positioning Method Based on WiFi/Bluetooth and PDR Fusion Positioning" *
刘雯: "基于EKF/PF的蓝牙/PDR/地图的融合定位算法研究" *
王晓甜;毛永毅;: "蓝牙测距与PDR测角融合的室内定位方法" *
郑晨辉;陈?;张雨婷;薛伟;: "融合粒子滤波与蓝牙地标矫正的定位算法" *

Also Published As

Publication number Publication date
CN114710745B (en) 2023-04-18

Similar Documents

Publication Publication Date Title
CN107038717B (en) A method of 3D point cloud registration error is automatically analyzed based on three-dimensional grid
CN111368881B (en) Low-frequency GPS track road network matching method based on multidimensional data fusion analysis
CN108307301B (en) Indoor positioning method based on RSSI ranging and track similarity
CN107621263B (en) Geomagnetic positioning method based on road magnetic field characteristics
CN104703143A (en) Indoor positioning method based on WIFI signal strength
CN104599286B (en) A kind of characteristic tracking method and device based on light stream
CN112881979B (en) Initial state self-adaptive fusion positioning method based on EKF filtering
CN113344956B (en) Ground feature contour extraction and classification method based on unmanned aerial vehicle aerial photography three-dimensional modeling
CN112050821B (en) Lane line polymerization method
CN111970633A (en) Indoor positioning method based on WiFi, Bluetooth and pedestrian dead reckoning fusion
CN109932711B (en) Atmospheric refraction correction method for radar measurement
CN111901749A (en) High-precision three-dimensional indoor positioning method based on multi-source fusion
CN110047133A (en) A kind of train boundary extraction method towards point cloud data
CN102506812B (en) VT checking method for stability judgment of reference points in deformation monitoring
CN115100376A (en) Electromagnetic spectrum map implementation method based on improved inverse distance interpolation method
CN108834047A (en) A kind of AP selection indoor orientation method of path loss model
CN114710745B (en) Indoor positioning method with Bluetooth and PDR information deeply fused
CN114710744B (en) Indoor positioning method integrating WiFi ranging and PDR calculation in depth
CN110927765B (en) Laser radar and satellite navigation fused target online positioning method
CN110888142A (en) Spacecraft hidden target point measuring method based on MEMS laser radar measuring technology
CN113810846B (en) Indoor positioning method based on WiFi and IMU fusion
CN108093364B (en) Improved weighted positioning method based on RSSI non-uniform spatial resolution
CN107451992B (en) Method and device for detecting SAR image change
CN111735449B (en) Geomagnetic matching positioning method and device
CN112578369B (en) Uncertainty estimation method and device, electronic equipment and storage medium

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
GR01 Patent grant
GR01 Patent grant