CN112729282B - Indoor positioning method integrating single anchor point ranging and pedestrian track calculation - Google Patents

Indoor positioning method integrating single anchor point ranging and pedestrian track calculation Download PDF

Info

Publication number
CN112729282B
CN112729282B CN202011520238.9A CN202011520238A CN112729282B CN 112729282 B CN112729282 B CN 112729282B CN 202011520238 A CN202011520238 A CN 202011520238A CN 112729282 B CN112729282 B CN 112729282B
Authority
CN
China
Prior art keywords
pedestrian
point
turning point
anchor point
distance
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.)
Active
Application number
CN202011520238.9A
Other languages
Chinese (zh)
Other versions
CN112729282A (en
Inventor
姚英彪
姚遥
许晓荣
徐欣
刘兆霆
李沛
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.)
Hangzhou Dianzi University
Original Assignee
Hangzhou Dianzi University
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 Hangzhou Dianzi University filed Critical Hangzhou Dianzi University
Priority to CN202011520238.9A priority Critical patent/CN112729282B/en
Publication of CN112729282A publication Critical patent/CN112729282A/en
Application granted granted Critical
Publication of CN112729282B publication Critical patent/CN112729282B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
    • G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
    • G01C21/10—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration
    • G01C21/12—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning
    • G01C21/16—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning by integrating acceleration or speed, i.e. inertial navigation

Landscapes

  • Engineering & Computer Science (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Remote Sensing (AREA)
  • Automation & Control Theory (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Navigation (AREA)

Abstract

The invention discloses an indoor positioning method integrating single anchor point ranging and pedestrian track calculation, and the method comprises the following steps of S1, setting an anchor point; s2, measuring the distance between the pedestrian starting point and the anchor point, and initializing the position coordinates of the pedestrian starting point; s3, estimating the step length and the heading mean value of each step of the pedestrian by using a PDR algorithm; s4, detecting whether the pedestrian has steering action, if so, executing a step S5, otherwise, continuing to execute a step S3; s5, measuring the distance between the steering point and the anchor point, and obtaining the coordinate of the ith steering point by a PDR algorithm according to the coordinate of the last steering point; s6, if i <2, returning to the step S3; if i is 2, go to step S7; if i >2, go to step S8; s7, calculating the coordinates of the pedestrian starting point according to a least square algorithm, calculating the coordinates of the next two steering points by using a PDR algorithm, and returning to the step S3; and S8, optimizing the coordinates of the pedestrian at the ith steering point according to a gradient descent algorithm. The invention uses an anchor node to combine PDR and a distance measuring and positioning method to realize accurate positioning.

Description

Indoor positioning method integrating single anchor point ranging and pedestrian track calculation
Technical Field
The invention belongs to the technical field of pedestrian positioning, and particularly relates to an indoor positioning method integrating single anchor point ranging and pedestrian track calculation.
Background
Indoor positioning is an important research area in the world of the internet of things at present, has attracted great attention, and has rapidly developed in recent years. Navigation and common day-to-day services (e.g., restaurant search and marketing) are typical location-based application services that rely on obtaining a user's real-time location at low cost. Therefore, an indoor positioning technology which combines accuracy and low cost becomes the first choice.
Currently, ranging-based positioning and Pedestrian Dead Reckoning (PDR) are two of the more popular indoor positioning technologies. The indoor positioning technology based on the ranging is based on the ranging of the intelligent device and the beacon anchor point, and then a trilateral or multilateral positioning algorithm is adopted to obtain the target position. Specifically, the distance between the device and the anchor point is obtained according to different communication technologies, such as signal field intensity ranging, reaching time difference ranging, reaching angle measurement and the like, then a geometric position constraint equation is established according to the known coordinates of the anchor point and the distance to the anchor point, and finally the position of the intelligent device is calculated. The PDR technology provides an indoor positioning technical scheme different from ranging. The PDR technology mainly comprises three steps of gait analysis, course estimation and position calculation. The gait analysis divides the acceleration data into groups according to steps, and estimates the step length of each step of the user; the course estimation calculates course angle by analyzing data of the gyroscope and the magnetometer; and finally, calculating the current coordinate by combining the coordinate of the pedestrian in the previous step.
However, both of the above-mentioned methods have certain limitations in practice. For example, the traditional ranging-based positioning is limited by the arrangement of anchor points, once the anchor points are sparsely distributed, the number of effective anchor points cannot reach the required minimum number (usually at least 3 anchor points are required), and the traditional ranging-based positioning will fail. Although the PDR does not need to arrange any basic equipment in advance, the PDR belongs to relative positioning and has the problems of initial position and positioning error accumulation, wherein the target cannot be positioned. Therefore, under sparse anchor points, how to locate the target remains a challenging problem.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides an indoor positioning method integrating single anchor point ranging and pedestrian track calculation, which not only solves the problem that more than 3 anchor points are needed in the traditional ranging-based positioning method, but also solves the problems that the initial position cannot be given, the positioning error is accumulated and the like in the traditional PDR, and thus, the high-precision indoor target positioning is realized. The invention can be applied to intelligent terminal equipment with built-in accelerometers and heading instruments, such as intelligent mobile phones, palm computers, personal digital equipment, intelligent wearing equipment and the like, and has the advantages of simple technical principle and easy popularization and use.
The invention adopts the following technical scheme: an indoor positioning method integrating single anchor point ranging and pedestrian track calculation comprises the following steps:
s1, setting an anchor point, and obtaining the position coordinate of the anchor point as (x)a,ya)；
S2, measuring distance d between pedestrian starting point and anchor point0And using the unknown coordinates (x)0,y0) Initializing a pedestrian starting point position;
s3, estimating the step length and the course mean value of each step of the pedestrian by using a PDR algorithm in the walking process of the pedestrian;
s4, detecting whether the pedestrian has steering action when walking one step, if so, executing a step S5, otherwise, continuing executing the step S3;
s5, ranging at the steering point to obtain the ranging distance d between the steering point and the anchor pointiI represents the serial number of the turning point and is calculated by the PDR algorithm according to the coordinate (x) of the last turning pointi-1,yi-1) The calculated coordinate of the ith steering point is (x)i,yi)；
S6, if i <2, returning to step S3; if i is 2, performing step S7; if i >2, perform step S8;
s7, calculating the coordinates (x) of the pedestrian starting point according to the least square algorithm0,y0) Then, the coordinates (x) of the next two turning points are calculated by using the PDR algorithm1,y1) And (x)2,y2) After that, the process returns to step S3;
s8, optimizing the coordinate (x) of the pedestrian at the ith steering point according to the gradient descent algorithmi,yi)；
And S9, detecting whether the pedestrian continues to walk, if so, returning to the step S3, and if not, ending the positioning.
Preferably, in step S3, the specific step of estimating the step size of each step of the pedestrian by using the PDR algorithm is as follows:
s3.1.1 triaxial acceleration data a collected by an accelerometerx，ay，azTo calculate the average acceleration atotal；
S3.1.2, removing average acceleration atotalA gravitational acceleration component g;
s3.1.3, carrying out low-pass filtering on the acceleration data obtained in the step S3.1.2 to eliminate high-frequency noise;
s3.1.4, converting the acceleration data obtained in step S3.1.3 into step size according to a non-linear step size estimation algorithm.
Preferably, in step S3.1.1, the average acceleration a is calculatedtotalThe concrete formula is as follows:
Figure GDA0003493254580000032
in step S3.1.2, the average acceleration a is removedtotalThe specific formula of the gravity acceleration component g is as follows:
a′total＝atotal-g；
in step S3.1.3, low-pass filtering is performed on the data to remove high-frequency noise, and the specific formula is as follows:
a＝filter(a′total)，
wherein, filter represents a low-pass filter, a represents the acceleration data after filtering noise;
in step S3.1.4, the acceleration data is converted into step sizes according to a non-linear step size estimation algorithm
Figure GDA0003493254580000031
The specific estimation formula is as follows:
Figure GDA0003493254580000041
wherein the content of the first and second substances,
Figure GDA0003493254580000042
the step length of the pedestrian from the i-1 th turning point to the k-th turning point is represented, and if i is 1, the pedestrian is from the starting point to the 1 st turning point; a and b are constants; a ismax,k,amin,kRespectively representing the maximum value and the minimum value of the acceleration in the k step; f. ofkThe walking frequency of the pedestrian at the k step is shown.
Preferably, in step S3, the specific step of estimating the mean heading value of each step of the pedestrian by using the PDR algorithm includes:
s3.2.1, collecting the heading angle of the pedestrian at the step k from the i-1 st turning point to the i-th turning point by an electronic compass
Figure GDA0003493254580000043
If i is 1, the vehicle moves to the 1 st turning point as a starting point;
s3.2.2 calculating the mean heading value of the step
Figure GDA0003493254580000044
The calculation formula is as follows:
Figure GDA0003493254580000045
wherein the content of the first and second substances,
Figure GDA0003493254580000046
representing the heading mean of the previous step.
Preferably, in step S4, the specific steps of detecting whether there is a steering action for each step of the pedestrian are as follows:
s4.1, calculating newly collected course angle data
Figure GDA0003493254580000047
And the average course of the previous step
Figure GDA0003493254580000048
Deviation delta ofθThe calculation formula is as follows:
Figure GDA0003493254580000049
s4.2, according to the deviation deltaθJudging whether the pedestrian turns, if deltaθ>δThrIf the steering is performed, deltaThrIndicates a pedestrian steering determination threshold value.
Preferably, in step S5, the PDR algorithm calculates the coordinate (x) of the last turning point according to the previous turning pointi-1,yi-1) The calculated coordinate of the ith steering point is (x)i,yi) The calculation formula is as follows:
Figure GDA0003493254580000051
wherein, Δ xi,ΔyiRespectively, represent the X, Y-axis variation amount converted from the i-1 st turning point to the i-th turning point, and Δ xi,ΔyiThe calculation formula of (a) is as follows:
Figure GDA0003493254580000052
wherein N isiRepresenting the number of steps the pedestrian takes between the i-1 st turning point and the i-th turning point.
Preferably, the step S7 specifically includes the following steps:
s7.1, initial point coordinate (x) according to hypothesis0,y0) The coordinates (x) of the next two turning points obtained by the PDR algorithm1,y1) And (x)2,y2) Distance d between the three points and the anchor point0、d1And d2And coordinates (x) of anchor pointsa,ya) The following equation is established:
Figure GDA0003493254580000053
s7.2, mixing
Figure GDA0003493254580000054
And
Figure GDA0003493254580000055
the following equation is obtained:
Figure GDA0003493254580000056
s7.3, calculating the coordinate (x) in the step S5i,yi) Is substituted into the equation of step S7.2 and converted into matrix form:
D＝CB0，
wherein:
Figure GDA0003493254580000057
B0＝[x0 y0]T，
Figure GDA0003493254580000058
s7.3, solving the equation B according to the least square method0，B0I.e. vector representation of the initial point coordinates:
B0＝(CTC)-1CTD；
s7.4, calculating B according to the PDR algorithm1、B2Coordinate vector, B1、B2Respectively, a vector representation of the coordinates of the first turning point and the second turning point.
Preferably, in step S8, the coordinate B of the pedestrian at the ith turning point is optimized according to a gradient descent algorithmi＝(xi,yi) The method comprises the following specific steps:
s8.1, establishing an optimized objective function according to the actually measured distance between each turning point and an anchor point, the distance between each turning point and the anchor point estimated by a PDR algorithm, the distance between each turning point estimated by a PDR algorithm and the distance between each turning point after gradient descent correction, wherein the method specifically comprises the following steps:
s8.1.1, calculating the distance measurement error epsilon of each turning point according to the following formula based on the actually measured distance measurement between each turning point and the anchor point and the distance between each turning point and the anchor point estimated by the PDR algorithmd,i：
εd,i＝||Bi-A||2-di，
Wherein A represents a vector representation of anchor point coordinates;
s8.1.2, calculating the step length estimation error epsilon according to the following formula based on the distance between each steering point after gradient descent correction and the distance between each steering point estimated by the PDR algorithmL,i：
εL,i＝||Bi-Bi-1||2-Li，
Wherein, | | · | | marks the euclidean distance of the calculated vector, LiRepresenting the distance the pedestrian travels from the i-1 turning point to the ith turning point,
Figure GDA0003493254580000061
s8.1.3, establishing an objective function:
Figure GDA0003493254580000062
wherein B isnCoordinate vector, alpha, representing the nth turning pointi,βiThe weights of the distance measurement error and the step error are respectively, and the calculation formulas are respectively as follows:
Figure GDA0003493254580000071
Figure GDA0003493254580000072
wherein e is a very small positive nonzero number;
s8.2, calculating the gradient of the objective function according to the following formula:
Figure GDA0003493254580000073
s8.3, updating the current coordinate vector according to the gradient descending direction, wherein the updating formula is as follows:
Figure GDA0003493254580000074
wherein, Bn,oldRepresenting the coordinate vector before updating, wherein lambda is the adjustment weight;
s8.4, updating the value of lambda, wherein the updating formula is as follows:
λ＝μ·λold，
where μ is the proportionality constant at [0,1 ], λoldThe value of the adjustment weight before the update is represented;
s8.5, judging whether any one of the following two conditions is met, if so, finishing the optimization, if not, returning to the step S8.2,
(1) adjusting the weight λ to be less than or equal to a threshold value, namely:
λ≤λThr，
(2) after the target function is iterated, the value of the target function is greater than or equal to the result of the previous round, that is:
f(Bn)≥f(Bn,old)。
preferably, e is 0.0001.
Preferably, λ isThrWhen the value is 0.0001, mu is in the range of [0.85,0.95 ]]。
The invention has the beneficial effects that: under the condition that only one anchor node exists, the PDR and the ranging positioning method are combined, and accurate target positioning can be achieved. Specifically, a positioning initial value is obtained through a least square method, and the gradient descent is used for quickly refining the positioning, so that the position of the pedestrian is positioned with high precision. The method has simple technical principle and good practicability and application prospect in reality.
Drawings
In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, it is obvious that the drawings in the following description are only some embodiments of the present invention, and for those skilled in the art, other drawings can be obtained according to the drawings without creative efforts.
FIG. 1 is a flow chart of an indoor positioning method incorporating single anchor ranging and pedestrian dead reckoning;
FIG. 2 is a diagram showing an example of a position satisfying a least squares calculation condition;
FIG. 3 is a schematic diagram of gradient descent correction coordinates.
Detailed Description
The following description of the embodiments of the present invention is provided by way of specific examples, and other advantages and effects of the present invention will be readily apparent to those skilled in the art from the disclosure herein. The invention is capable of other and different embodiments and of being practiced or of being carried out in various ways, and its several details are capable of modification in various respects, all without departing from the spirit and scope of the present invention. It is to be noted that the features in the following embodiments and examples may be combined with each other without conflict.
Assume that point-to-point ranging between devices employs Received Signal Strength (RSSI) ranging.
Referring to fig. 1, the present embodiment provides an indoor positioning method integrating single anchor point ranging and pedestrian dead reckoning, including the steps of:
s1, setting an anchor point, and obtaining the position coordinate of the anchor point as (x)a,ya)；
S2, measuring distance d between pedestrian starting point and anchor point0And using the unknown coordinates (x)0,y0) Initializing a pedestrian starting point position;
s3, estimating the step length and the course mean value of each step of the pedestrian by using a PDR algorithm in the walking process of the pedestrian;
s4, detecting whether the pedestrian has steering action when walking one step, if so, executing a step S5, otherwise, continuing executing the step S3;
s5, ranging at the steering point to obtain the ranging distance d between the steering point and the anchor pointiI represents the serial number of the turning point and is calculated by the PDR algorithm according to the coordinate (x) of the last turning pointi-1,yi-1) The calculated coordinate of the ith steering point is (x)i,yi)；
S6, if i <2, returning to step S3; if i is 2, performing step S7; if i >2, perform step S8;
s7, calculating the coordinates (x) of the pedestrian starting point according to the least square algorithm0,y0) Then, the coordinates (x) of the next two turning points are calculated by using the PDR algorithm1,y1) And (x)2,y2) (refer to fig. 2), return to step S3 after completion;
s8, optimizing the coordinate (x) of the pedestrian at the ith steering point according to the gradient descent algorithmi,yi) (refer to fig. 3);
and S9, detecting whether the pedestrian continues to walk, if so, returning to the step S3, and if not, ending the positioning.
Wherein, the RSSI ranging method is adopted for ranging in steps S2 and S5, and the main process is as follows:
the RSSI signal strength model is:
Figure GDA0003493254580000091
wherein the RSSIiIndicating the distance d of the terminal device from the anchor pointiValue of time, signal attenuation, drFor reference distances, δ is a gaussian distribution following a mean of 0, and n is an attenuation factor under different circumstances.
For example, in a city region, n takes a value of 3.7 and refers to a distance dr1 m when the hand is usedThe signal strength received by the machine was-59.8 dBm. Converting the signal intensity of the anchor point received by the terminal equipment into the distance di：
Figure GDA0003493254580000101
Specifically, the method comprises the following steps:
in step S3, the specific steps of estimating the step size of each step of the pedestrian using the PDR algorithm are as follows:
s3.1.1 triaxial acceleration data a collected by an accelerometerx，ay，azTo calculate the average acceleration atotal；
S3.1.2, removing average acceleration atotalA gravitational acceleration component g;
s3.1.3, carrying out low-pass filtering on the acceleration data obtained in the step S3.1.2 to eliminate high-frequency noise;
s3.1.4, converting the acceleration data obtained in step S3.1.3 into step size according to a non-linear step size estimation algorithm.
And the number of the first and second electrodes,
in step S3.1.1, the average acceleration a is calculatedtotalThe concrete formula is as follows:
Figure GDA0003493254580000102
in step S3.1.2, the average acceleration a is removedtotalThe specific formula of the gravity acceleration component g is as follows:
a′total＝atotal-g；
in step S3.1.3, low-pass filtering is performed on the data to remove high-frequency noise, and the specific formula is as follows:
a＝filter(a′total)，
wherein, filter represents a low-pass filter, a represents the acceleration data after filtering noise;
in step S3.1.4, the acceleration data is converted into step sizes according to a non-linear step size estimation algorithm
Figure GDA0003493254580000105
The specific estimation formula is as follows:
Figure GDA0003493254580000103
wherein the content of the first and second substances,
Figure GDA0003493254580000104
the step length of the pedestrian from the i-1 th turning point to the k-th turning point is represented, and if i is 1, the pedestrian is from the starting point to the 1 st turning point; a and b are constants determined by actual environment; a ismax,k,amin,kRespectively representing the maximum value and the minimum value of the acceleration in the k step; f. ofkThe walking frequency of the pedestrian at the k step is shown.
In step S3, the specific steps of estimating the mean heading value of each step of the pedestrian by using the PDR algorithm are as follows:
s3.2.1, collecting the heading angle of the pedestrian at the step k from the i-1 st turning point to the i-th turning point by an electronic compass
Figure GDA0003493254580000111
If i is 1, the vehicle moves to the 1 st turning point as a starting point;
s3.2.2 calculating the mean heading value of the step
Figure GDA0003493254580000112
The calculation formula is as follows:
Figure GDA0003493254580000113
wherein the content of the first and second substances,
Figure GDA0003493254580000114
representing the heading mean of the previous step.
In step S4, the specific steps of detecting whether there is a steering action for each step of the pedestrian are as follows:
s4.1, calculatingNewly collected course angle data
Figure GDA0003493254580000115
And the average course of the previous step
Figure GDA0003493254580000116
Deviation delta ofθThe calculation formula is as follows:
Figure GDA0003493254580000117
s4.2, according to the deviation deltaθJudging whether the pedestrian turns, if deltaθ>δThrIf the steering is performed, deltaThrIndicates a pedestrian steering determination threshold value.
In step S5, the PDR algorithm calculates the coordinate (x) of the last turning pointi-1,yi-1) The calculated coordinate of the ith steering point is (x)i,yi) When i is equal to 1, that is, the coordinates of the 1 st turning point are calculated according to the coordinates of the starting point, the starting point can also be understood as the 0 th turning point, and the calculation formula is as follows:
Figure GDA0003493254580000118
wherein, Δ xi,ΔyiRespectively, represent the X, Y-axis variation amount converted from the i-1 st turning point to the i-th turning point, and Δ xi,ΔyiThe calculation formula of (a) is as follows:
Figure GDA0003493254580000119
wherein N isiRepresenting the number of steps the pedestrian takes between the i-1 st turning point and the i-th turning point.
Referring to fig. 2, step S7 specifically includes the following steps:
s7.1, initial point coordinate (x) according to hypothesis0,y0) By PDR algorithmThe coordinates (x) of the next two turning points are obtained1,y1) And (x)2,y2) Distance d between the three points and the anchor point0、d1And d2And coordinates (x) of anchor pointsa,ya) The following equation is established:
Figure GDA0003493254580000121
s7.2, mixing
Figure GDA0003493254580000122
And
Figure GDA0003493254580000123
the following equation is obtained:
Figure GDA0003493254580000124
s7.3, calculating the coordinate (x) in the step S5i,yi) Is substituted into the equation of step S7.2 and converted into matrix form:
D＝CB0，
wherein:
Figure GDA0003493254580000125
B0＝[x0 y0]T，
Figure GDA0003493254580000126
s7.3, solving the equation B according to the least square method0，B0I.e. vector representation of the initial point coordinates:
B0＝(CTC)-1CTD；
s7.4, according toCalculation of B by PDR algorithm1、B2Coordinate vector, B1、B2Respectively, a vector representation of the coordinates of the first turning point and the second turning point.
In step S8, the coordinate B of the pedestrian at the ith turning point is optimized according to the gradient descent algorithmi＝(xi,yi) The method comprises the following specific steps:
s8.1, establishing an optimized objective function according to the actually measured distance between each turning point and an anchor point, the distance between each turning point and the anchor point estimated by a PDR algorithm, the distance between each turning point estimated by a PDR algorithm and the distance between each turning point after gradient descent correction, wherein the method specifically comprises the following steps:
s8.1.1, calculating the distance measurement error epsilon of each turning point according to the following formula based on the actually measured distance measurement between each turning point and the anchor point and the distance between each turning point and the anchor point estimated by the PDR algorithmd,i：
εd,i＝||Bi-A||2-di，
Wherein A represents a vector representation of anchor point coordinates;
s8.1.2, calculating the step length estimation error epsilon according to the following formula based on the distance between each steering point after gradient descent correction and the distance between each steering point estimated by the PDR algorithmL,i：
εL,i＝||Bi-Bi-1||2-Li，
Wherein, | | · | | marks the euclidean distance of the calculated vector, LiRepresenting the distance the pedestrian travels from the i-1 turning point to the ith turning point,
Figure GDA0003493254580000131
s8.1.3, establishing an objective function:
Figure GDA0003493254580000132
whereinBnCoordinate vector, alpha, representing the nth turning pointi,βiThe weights of the distance measurement error and the step error are respectively, and the calculation formulas are respectively as follows:
Figure GDA0003493254580000133
Figure GDA0003493254580000134
wherein, epsilon is a very small positive nonzero number, and the value in the embodiment is 0.0001;
s8.2, calculating the gradient of the objective function according to the following formula:
Figure GDA0003493254580000141
s8.3, referring to fig. 3, updating the current coordinate vector according to the gradient descent direction, where the updating formula is:
Figure GDA0003493254580000142
wherein, Bn,oldAnd representing a coordinate vector before updating, wherein lambda is an adjusting weight value, the gradient adjusting efficiency is faster when the lambda value is larger, but the gradient is invalid due to the excessively large lambda, and therefore the lambda is not excessively large or small. For example, when measuring distance diHas a value in the range of [10,20 ]]The initial value of λ may be set to 5 and scaled down;
s8.4, updating the value of lambda, wherein the updating formula is as follows:
λ＝μ·λold，
wherein mu is a proportionality constant in [0,1), mu influences the adjustment efficiency of lambda, the larger mu is, the more gradient descending times are increased, but the higher precision is, otherwise, the smaller gradient descending times are decreased, and the precision is reduced, generally, the value range of mu is suggested to be [0.85, 0.95%]，λoldThe value of the adjustment weight before the update is represented;
s8.5, judging whether any one of the following two conditions is met, if so, finishing the optimization, if not, returning to the step S8.2,
(1) adjusting the weight λ to be less than or equal to a threshold value, namely:
λ≤λThr，
in this example λThrTaking 0.0001;
(2) after the target function is iterated, the value of the target function is greater than or equal to the result of the previous round, that is:
f(Bn)≥f(Bn,old)。
the objective function is minimized after the updating step is completed.
The above-mentioned embodiments are merely illustrative of the preferred embodiments of the present invention, and do not limit the scope of the present invention, and various modifications and improvements of the technical solutions of the present invention by those skilled in the art should fall within the protection scope of the present invention without departing from the design spirit of the present invention.

Claims (10)

1. An indoor positioning method integrating single anchor point ranging and pedestrian dead reckoning is characterized by comprising the following steps:
s1, setting an anchor point, and obtaining the position coordinate of the anchor point as (x)a,ya)；
S2, measuring distance d between pedestrian starting point and anchor point0And using the unknown coordinates (x)0,y0) Initializing a pedestrian starting point position;
s3, estimating the step length and the course mean value of each step of the pedestrian by using a PDR algorithm in the walking process of the pedestrian;
s4, detecting whether the pedestrian has steering action when walking one step, if so, executing a step S5, otherwise, continuing executing the step S3;
s5, ranging at the steering point to obtain the ranging distance d between the steering point and the anchor pointiI represents the serial number of the turning point and is calculated by the PDR algorithm according to the coordinate (x) of the last turning pointi-1,yi-1) The ith calculatedThe coordinate of the turning point is (x)i,yi)；
S6, if i <2, returning to step S3; if i is 2, performing step S7; if i >2, perform step S8;
s7, calculating the coordinates (x) of the pedestrian starting point according to the least square algorithm0,y0) Then, the coordinates (x) of the next two turning points are calculated by using the PDR algorithm1,y1) And (x)2,y2) After that, the process returns to step S3;
s8, optimizing the coordinate (x) of the pedestrian at the ith steering point according to the gradient descent algorithmi,yi)；
And S9, detecting whether the pedestrian continues to walk, if so, returning to the step S3, and if not, ending the positioning.
2. The indoor positioning method combining single anchor point ranging and pedestrian dead reckoning as claimed in claim 1, wherein in step S3, the specific step of estimating the step length of each step of the pedestrian by using the PDR algorithm comprises:
s3.1.1 triaxial acceleration data a collected by an accelerometerx，ay，azTo calculate the average acceleration atotal；
S3.1.2, removing average acceleration atotalA gravitational acceleration component g;
s3.1.3, carrying out low-pass filtering on the acceleration data obtained in the step S3.1.2 to eliminate high-frequency noise;
s3.1.4, converting the acceleration data obtained in step S3.1.3 into step size according to a non-linear step size estimation algorithm.
3. The indoor positioning method integrating single anchor point ranging and pedestrian dead reckoning as claimed in claim 2, wherein:
in step S3.1.1, the average acceleration a is calculatedtotalThe concrete formula is as follows:
Figure FDA0003493254570000021
in step S3.1.2, the average acceleration a is removedtotalThe specific formula of the gravity acceleration component g is as follows:
a′total＝atotal-g；
in step S3.1.3, low-pass filtering is performed on the data to remove high-frequency noise, and the specific formula is as follows:
a＝filter(a′total)，
wherein, filter represents a low-pass filter, a represents the acceleration data after filtering noise;
in step S3.1.4, the acceleration data is converted into step sizes according to a non-linear step size estimation algorithm
Figure FDA0003493254570000024
The specific estimation formula is as follows:
Figure FDA0003493254570000022
wherein the content of the first and second substances,
Figure FDA0003493254570000023
the step length of the pedestrian from the i-1 th turning point to the k-th turning point is represented, and if i is 1, the pedestrian is from the starting point to the 1 st turning point; a and b are constants; a ismax,k,amin,kRespectively representing the maximum value and the minimum value of the acceleration in the k step; f. ofkThe walking frequency of the pedestrian at the k step is shown.
4. The indoor positioning method integrating single anchor point ranging and pedestrian dead reckoning as claimed in claim 3, wherein in step S3, the specific step of estimating the mean heading value of each step of the pedestrian by using the PDR algorithm comprises:
s3.2.1, collecting the heading angle of the pedestrian at the step k from the i-1 st turning point to the i-th turning point by an electronic compass
Figure FDA0003493254570000031
If i is 1, the vehicle moves to the 1 st turning point as a starting point;
s3.2.2 calculating the mean heading value of the step
Figure FDA0003493254570000032
The calculation formula is as follows:
Figure FDA0003493254570000033
wherein the content of the first and second substances,
Figure FDA0003493254570000034
representing the heading mean of the previous step.
5. The indoor positioning method integrating single anchor point ranging and pedestrian dead reckoning as claimed in claim 4, wherein in step S4, the specific steps of detecting whether the pedestrian turns or not at each step are as follows:
s4.1, calculating newly collected course angle data
Figure FDA0003493254570000035
And the average course of the previous step
Figure FDA0003493254570000036
Deviation delta ofθThe calculation formula is as follows:
Figure FDA0003493254570000037
s4.2, according to the deviation deltaθJudging whether the pedestrian turns, if deltaθ>δThrIf the steering is performed, deltaThrIndicates a pedestrian steering determination threshold value.
6. The method of claim 5The indoor positioning method integrating single anchor point ranging and pedestrian dead reckoning is characterized in that in step S5, a PDR algorithm is used for calculating coordinates (x) of a last turning point according to coordinates (x) of a last turning pointi-1,yi-1) The calculated coordinate of the ith steering point is (x)i,yi) The calculation formula is as follows:
Figure FDA0003493254570000038
wherein, Δ xi,ΔyiRespectively, represent the X, Y-axis variation amount converted from the i-1 st turning point to the i-th turning point, and Δ xi,ΔyiThe calculation formula of (a) is as follows:
Figure FDA0003493254570000039
wherein N isiRepresenting the number of steps the pedestrian takes between the i-1 st turning point and the i-th turning point.
7. The indoor positioning method combining single anchor point ranging and pedestrian dead reckoning as claimed in claim 6, wherein step S7 specifically includes the following steps:
s7.1, initial point coordinate (x) according to hypothesis0,y0) The coordinates (x) of the next two turning points obtained by the PDR algorithm1,y1) And (x)2,y2) Distance d between the three points and the anchor point0、d1And d2And coordinates (x) of anchor pointsa,ya) The following equation is established:
Figure FDA0003493254570000041
s7.2, mixing
Figure FDA0003493254570000042
And
Figure FDA0003493254570000043
the following equation is obtained:
Figure FDA0003493254570000044
s7.3, calculating the coordinate (x) in the step S5i,yi) Is substituted into the equation of step S7.2 and converted into matrix form:
D＝CB0，
wherein:
Figure FDA0003493254570000045
B0＝[x0 y0]T，
Figure FDA0003493254570000046
s7.3, solving the equation B according to the least square method0，B0I.e. vector representation of the initial point coordinates:
B0＝(CTC)-1CTD；
s7.4, calculating B according to the PDR algorithm1、B2Coordinate vector, B1、B2Respectively, a vector representation of the coordinates of the first turning point and the second turning point.
8. The indoor positioning method combining single anchor point ranging and pedestrian dead reckoning as claimed in claim 7, wherein in step S8, the coordinate B of the pedestrian at the ith turning point is optimized according to a gradient descent algorithmi＝(xi,yi) The method comprises the following specific steps:
s8.1, establishing an optimized objective function according to the actually measured distance between each turning point and an anchor point, the distance between each turning point and the anchor point estimated by a PDR algorithm, the distance between each turning point estimated by a PDR algorithm and the distance between each turning point after gradient descent correction, wherein the method specifically comprises the following steps:
s8.1.1, calculating the distance measurement error epsilon of each turning point according to the following formula based on the actually measured distance measurement between each turning point and the anchor point and the distance between each turning point and the anchor point estimated by the PDR algorithmd,i：
εd,i＝||Bi-A||2-di，
Wherein A represents a vector representation of anchor point coordinates;
s8.1.2, calculating the step length estimation error epsilon according to the following formula based on the distance between each steering point after gradient descent correction and the distance between each steering point estimated by the PDR algorithmL,i：
εL,i＝||Bi-Bi-1||2-Li，
Wherein, | | · | | marks the euclidean distance of the calculated vector, LiRepresenting the distance the pedestrian travels from the i-1 turning point to the ith turning point,
Figure FDA0003493254570000051
s8.1.3, establishing an objective function:
Figure FDA0003493254570000052
wherein B isnCoordinate vector, alpha, representing the nth turning pointi,βiThe weights of the distance measurement error and the step error are respectively, and the calculation formulas are respectively as follows:
Figure FDA0003493254570000053
Figure FDA0003493254570000054
wherein e is a very small positive nonzero number;
s8.2, calculating the gradient of the objective function according to the following formula:
Figure FDA0003493254570000061
s8.3, updating the current coordinate vector according to the gradient descending direction, wherein the updating formula is as follows:
Figure FDA0003493254570000062
wherein, Bn,oldRepresenting the coordinate vector before updating, wherein lambda is the adjustment weight;
s8.4, updating the value of lambda, wherein the updating formula is as follows:
λ＝μ·λold，
where μ is the proportionality constant at [0,1 ], λoldThe value of the adjustment weight before the update is represented;
s8.5, judging whether any one of the following two conditions is met, if so, finishing the optimization, if not, returning to the step S8.2,
(1) adjusting the weight λ to be less than or equal to a threshold value, namely:
λ≤λThr，
(2) after the target function is iterated, the value of the target function is greater than or equal to the result of the previous round, that is:
f(Bn)≥f(Bn,old)。
9. the indoor positioning method integrating single anchor point ranging and pedestrian dead reckoning as claimed in claim 8, wherein e is 0.0001.
10. The indoor positioning method integrating single anchor point ranging and pedestrian dead reckoning as claimed in claim 8, wherein λ isThrWhen the value is 0.0001, mu is in the range of [0.85,0.95 ]]。
CN202011520238.9A 2020-12-21 2020-12-21 Indoor positioning method integrating single anchor point ranging and pedestrian track calculation Active CN112729282B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202011520238.9A CN112729282B (en) 2020-12-21 2020-12-21 Indoor positioning method integrating single anchor point ranging and pedestrian track calculation

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202011520238.9A CN112729282B (en) 2020-12-21 2020-12-21 Indoor positioning method integrating single anchor point ranging and pedestrian track calculation

Publications (2)

Publication Number Publication Date
CN112729282A CN112729282A (en) 2021-04-30
CN112729282B true CN112729282B (en) 2022-03-25

Family

ID=75604345

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202011520238.9A Active CN112729282B (en) 2020-12-21 2020-12-21 Indoor positioning method integrating single anchor point ranging and pedestrian track calculation

Country Status (1)

Country Link
CN (1) CN112729282B (en)

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113777557B (en) * 2021-09-26 2023-09-15 北方工业大学 UWB indoor positioning method and system based on redundant distance screening

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108489489A (en) * 2018-01-23 2018-09-04 杭州电子科技大学 A kind of bluetooth auxiliary corrects the indoor orientation method and system of PDR
CN109405829A (en) * 2018-08-28 2019-03-01 桂林电子科技大学 Pedestrian's method for self-locating based on smart phone audio-video Multi-source Information Fusion
CN109855620A (en) * 2018-12-26 2019-06-07 北京壹氢科技有限公司 A kind of indoor pedestrian navigation method based on from backtracking algorithm
CN110345939A (en) * 2019-07-02 2019-10-18 山东科技大学 A kind of indoor orientation method merging fuzzy logic judgement and cartographic information
US10757539B1 (en) * 2019-07-16 2020-08-25 Eagle Technology, Llc System for mapping building interior with PDR and ranging and related methods

Family Cites Families (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US9188659B2 (en) * 2012-10-10 2015-11-17 Telefonaktiebolaget L M Ericsson (Publ) Methods and network nodes for positioning based on displacement data

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108489489A (en) * 2018-01-23 2018-09-04 杭州电子科技大学 A kind of bluetooth auxiliary corrects the indoor orientation method and system of PDR
CN109405829A (en) * 2018-08-28 2019-03-01 桂林电子科技大学 Pedestrian's method for self-locating based on smart phone audio-video Multi-source Information Fusion
CN109855620A (en) * 2018-12-26 2019-06-07 北京壹氢科技有限公司 A kind of indoor pedestrian navigation method based on from backtracking algorithm
CN110345939A (en) * 2019-07-02 2019-10-18 山东科技大学 A kind of indoor orientation method merging fuzzy logic judgement and cartographic information
US10757539B1 (en) * 2019-07-16 2020-08-25 Eagle Technology, Llc System for mapping building interior with PDR and ranging and related methods

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
A Low-Cost Single-Anchor Solution for Indoor Positioning Using BLE and Inertial Sensor Data;Ye Feng et al.;《IEEE ACCESS》;20191231;第7卷;162439-162453 *
PDR辅助UWB的室内非视距定位方法;孙建强等;《传感技术学报》;20200531;第33卷(第5期);711-717 *
基于PDR和RSSI的室内定位算法研究;郑学理等;《仪器仪表学报》;20150531;第36卷(第5期);1177-1185 *

Also Published As

Publication number Publication date
CN112729282A (en) 2021-04-30

Similar Documents

Publication Publication Date Title
Bai et al. A high-precision and low-cost IMU-based indoor pedestrian positioning technique
CN109413578B (en) Indoor positioning method based on fusion of WIFI and PDR
Zhou et al. Activity sequence-based indoor pedestrian localization using smartphones
CN103968827B (en) A kind of autonomic positioning method of wearable body gait detection
CN106441302B (en) Indoor positioning method in large-scale open area
CN110207704B (en) Pedestrian navigation method based on intelligent identification of building stair scene
CN105589064A (en) Rapid establishing and dynamic updating system and method for WLAN position fingerprint database
CN107830858B (en) Gravity-assisted mobile phone heading estimation method
CN109186601A (en) A kind of laser SLAM algorithm based on adaptive Unscented kalman filtering
CN111024075B (en) Pedestrian navigation error correction filtering method combining Bluetooth beacon and map
CN104599286B (en) A kind of characteristic tracking method and device based on light stream
CN111829516B (en) Autonomous pedestrian positioning method based on smart phone
CN108537101B (en) Pedestrian positioning method based on state recognition
CN106610292A (en) Method of indoor positioning through combination of WIFI and pedestrian dead reckoning (PDR)
CN110346821B (en) SINS/GPS combined attitude-determining and positioning method and system for solving long-time GPS unlocking problem
CN106840163A (en) A kind of indoor orientation method and system
CN111970633A (en) Indoor positioning method based on WiFi, Bluetooth and pedestrian dead reckoning fusion
CA2615211A1 (en) Method and device for measuring the progress of a moving person
CN109459028A (en) A kind of adaptive step estimation method based on gradient decline
CN109916396B (en) Indoor positioning method based on multidimensional geomagnetic information
CN112729282B (en) Indoor positioning method integrating single anchor point ranging and pedestrian track calculation
CN111964667B (en) geomagnetic-INS (inertial navigation System) integrated navigation method based on particle filter algorithm
CN106871894B (en) Map matching method based on conditional random field
CN112362057A (en) Inertial pedestrian navigation algorithm based on zero-speed correction and attitude self-observation
Chen et al. Pedestrian positioning with physical activity classification for indoors

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