CN110487273B - Indoor pedestrian trajectory calculation method assisted by level gauge - Google Patents
Indoor pedestrian trajectory calculation method assisted by level gauge Download PDFInfo
- Publication number
- CN110487273B CN110487273B CN201910634626.0A CN201910634626A CN110487273B CN 110487273 B CN110487273 B CN 110487273B CN 201910634626 A CN201910634626 A CN 201910634626A CN 110487273 B CN110487273 B CN 110487273B
- Authority
- CN
- China
- Prior art keywords
- pedestrian
- mobile phone
- acceleration
- axis
- course
- 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
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
- G01C21/165—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 combined with non-inertial navigation instruments
-
- 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
- G01C21/20—Instruments for performing navigational calculations
- G01C21/206—Instruments for performing navigational calculations specially adapted for indoor 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 belongs to the field of indoor positioning, relates to technologies such as sensor signal processing, particle filter algorithm, pedestrian dead reckoning and the like, and particularly relates to a level-meter-assisted indoor pedestrian trajectory reckoning method which is used for overcoming the problem that course information cannot be used due to unstable mobile phone posture in a traditional PDR method. Firstly, detecting the gesture of a mobile phone by using a built-in level gauge of the mobile phone, acquiring reliable pedestrian course information, analyzing the Wi-Fi received signal strength of each AP to acquire the rough position of a pedestrian, detecting whether the pedestrian turns or not by using the geomagnetic information of the mobile phone, updating the pedestrian course by combining with actual map information, and finally calculating the movement track of the pedestrian by combining three kinds of information, namely pedestrian step frequency, step length and course; compared with general mobile phone inertial navigation, the method does not need to keep the posture of the mobile phone unchanged relative to the posture of a human body in the navigation process, updates the course of the pedestrian in real time, has simple technology and strong practicability, and is suitable for the field of indoor crowdsourcing fingerprint database establishment.
Description
Technical Field
The invention belongs to the field of indoor positioning, relates to the technologies of sensor signal processing, particle filter algorithm, pedestrian Dead Reckoning (PDR) and the like, and particularly relates to a gradienter-assisted indoor pedestrian trajectory Reckoning method.
Background
In the PDR technology, three kinds of information, namely step frequency, step length and course, are important; the research on the step frequency and the step length is mature, and how to improve the accuracy of course information acquisition is still a research hotspot. At present, the mainstream thought is to solve the course by a quaternion method and by combining gyroscope information, and the method requires that a mobile phone always keeps a flat-end attitude fixed relative to a human body, so that the calculation amount is large, the hardware requirement is high, errors are easily accumulated, the course information is unavailable, and the failure of a PDR algorithm is caused.
Currently, the mainstream heading estimation method generally requires that a mobile phone always keeps a fixed flat-end attitude to ensure that geomagnetic data and gyroscope data are accurate, and is suitable for calculation of a quaternion method. However, in a practical situation, the mobile phone has various gestures, and can make a call, place a bag and horizontally place the mobile phone to watch a video, and the user experience is seriously influenced by the limiting condition, so that the current mobile phone PDR algorithm has no practical application value.
In an indoor environment, the pedestrian motion state is generally divided into long-time straight walking and short-time quarter turn. Researches find that the mobile phone is most frequently in a posture that the body of the mobile phone is flat and the head of the mobile phone is forward, and at the moment, geomagnetic information obtained by the mobile phone can well represent the heading of a pedestrian. The documents "a.rai, k.k.chintalapudi, v.n.padmanahan, and r.sen," ze: zero-effort crowdsourceing for inductor localization, "in proc.acm mobilcom, 2012, pp.293-304" indicate that in indoor environments, walls, corridors can assist in positioning. According to the background, the invention provides a pedestrian motion trajectory calculation method integrating actual environment information.
Disclosure of Invention
The invention aims to provide a gradienter-assisted indoor pedestrian trajectory calculation method aiming at the problem that course information is unavailable due to unstable mobile phone posture in the traditional PDR method; the method comprises the steps of firstly detecting the posture of a mobile phone by using a built-in level gauge of the mobile phone, obtaining reliable pedestrian course information, then analyzing the Wi-Fi Received Signal Strength (RSS) of each AP to obtain the rough position of a pedestrian, then detecting whether the pedestrian turns or not by using geomagnetic information of the mobile phone, updating the course of the pedestrian by combining with actual map information, and finally calculating the movement track of the pedestrian by combining three information of pedestrian step frequency, step length and course.
In order to realize the purpose, the invention adopts the technical scheme that:
a leveling instrument assisted indoor pedestrian trajectory estimation method comprises the following specific steps:
step 1. Data preprocessing
Acquiring data through an acceleration sensor, a gyroscope sensor, a geomagnetic sensor and a gravity sensor, and performing filtering smoothing on the acquired data;
step 2, calculating step frequency
2-1, preprocessing acceleration data, setting a sliding time window delta T and a sampling frequency f;
for each sample point, calculate:
wherein, a x 、a y 、a z The accelerations of the X axis, the Y axis and the Z axis are measured by an acceleration sensor;
2-2, searching all wave crests in a sliding time window delta T, and marking in sequence;
and 2-3, removing the pseudo wave peaks to obtain all effective wave peaks in the sliding time window delta T: location = [ l 1 ,l 2 ,l 3 ,...,l i ,...,l n ,] T If n is the total number of effective wave crests, the step number in the sliding time window delta T is n; removing the pseudo wave crest by adopting a three-threshold method, which specifically comprises the following steps:
2-3-1, setting the threshold value of the resultant acceleration to be 1.2g: when the resultant acceleration a is less than 1.2g, determining the resultant acceleration as a pseudo peak;
2-3-2, setting a peak threshold value: calculating the average A of the three maxima within the time window max Average of the three minima A min Taking A max And A min The average value average of is a peak threshold value, and when the average value average is smaller than the peak threshold value, the peak is determined as a pseudo peak;
2-3-3, setting a sampling point threshold value M: m =0.2f; when the number of sampling points between two adjacent wave crests is less than M, determining that the second wave crest is a pseudo wave crest; otherwise, both wave crests are effective wave crests;
step 3, calculating step length
3-1. Calculation
Obtaining the starting point of the ith step and simultaneously taking the starting point as the end point of the (i-1) th step; and further obtaining a set of all starting points and all end points:
RANGE＝[range_1,range_2,...,range_i,...,range_n,range_L] T ；
wherein, range _1 represents the 1 st sampling point in the sliding time window delta T as the starting point of the 1 st step, and range _ L represents the last 1 sampling point in the sliding time window delta T as the ending point of the nth step;
3-2, calculating the step length of each step:
wherein Q i Total number of samples, a, of step i q To representThe combined acceleration of the q sampling point in the ith step, k and b are preset constants;
step 4, auxiliary calculation of pedestrian course angle based on level meter
4-1, corresponding time stamp t, azimuth angle alpha, pitch angle beta, roll angle chi and X-axis gravity component G at each moment x Y-axis gravity component G y Z-axis gravity component G z And (3) forming a vector: GRAND = [ t, alpha, beta, chi, G x ,G y ,G z ] T (ii) a Wherein, the azimuth angle alpha is measured by a geomagnetic sensor, the pitch angle beta and the roll angle chi are measured by a gyro sensor, and the gravity component G of the X axis x Y-axis gravity component G y Z-axis gravity component G z Measured by a gravity sensor;
for time t, if the condition is satisfied: the pitch angle beta and the roll angle chi are both smaller than a preset threshold value beta th And the component G of the X-axis of gravity x Less than a predetermined threshold G th If the mobile phone is in the horizontal posture at the moment t, determining that the mobile phone is in the horizontal posture; starting to detect from zero time, and when the mobile phone is continuously detected to be in a flat posture and exceeds a time threshold t th Effective sampling is realized, and the average value of the azimuth angles at all the moments in the time period is used as the predicted initial pedestrian course angle;
4-2, combining the indoor map and the AP layout to obtain all possible headings; comparing the predicted initial pedestrian course angle with all possible courses to obtain the possible course closest to the predicted initial pedestrian course angle, and taking the possible course as the initial pedestrian direction;
5. turn detection
5-1, setting the collection window delta T orien Calculating Δ T for each acquisition window orien The average value orien of the azimuthal angles α within the range;
5-2, calculating the variation amplitudes of the front orien and the rear orien, and if the variation amplitudes exceed a threshold value of 80 degrees, judging that a turning action occurs;
5-3, detecting the signal intensity of all APs at the pedestrian position in real time, selecting the inflection point where the AP with the maximum signal intensity is located as the closest inflection point of the pedestrian at the turning moment, steering to the AP with the maximum signal intensity at the closest inflection point, and updating the course angle of the pedestrian;
and 6, obtaining pedestrian track calculation by adopting a particle filter algorithm according to the step frequency, the step length and the pedestrian course angle information obtained in the steps 2 to 5.
The invention has the beneficial effects that:
the invention provides a leveling instrument assisted indoor pedestrian track calculation method, which is characterized in that a built-in leveling instrument of a mobile phone is used for detecting the posture of the mobile phone, when the mobile phone is detected to be in a flat posture, geomagnetic data at the moment is recorded as the heading of a pedestrian, and the geomagnetic data is continuously used until the pedestrian is detected to turn; after the turning action of the pedestrian is detected, roughly acquiring the approximate position of the pedestrian by using the AP signal strength information, matching the approximate position to the corner where the pedestrian is located, automatically matching the current pedestrian course to the turning direction of the corner where the pedestrian is located by combining map information, and iterating in the way until the next turning action is taken, so as to update the course information of the pedestrian in real time. Compared with general mobile phone inertial navigation, the method does not need to keep the posture of the mobile phone unchanged relative to the posture of a human body in the navigation process, updates the heading of the pedestrian in real time by fusing RSS signal intensity and a particle filter algorithm, has simple technology and strong practicability, and is suitable for the field of indoor crowdsourcing fingerprint database establishment.
Drawings
Fig. 1 is a flow chart of an indoor pedestrian trajectory estimation method assisted by a level gauge.
FIG. 2 is a flow chart of a heading update algorithm based on a level of the present invention.
FIG. 3 is a graph of the level detection performance in an embodiment of the present invention.
Fig. 4 is a schematic diagram of an experimental scenario in an embodiment of the present invention.
FIG. 5 is a schematic diagram of a mobile phone posture in an embodiment of the invention.
Detailed Description
The technical scheme of the invention is further described in detail by combining the drawings and the examples.
The present embodiment provides a method for calculating an indoor pedestrian trajectory assisted by a level, the flow of which is shown in fig. 1, and the method specifically includes the following steps:
1. data pre-processing
Registering monitoring of an acceleration sensor, a gyroscope sensor, a geomagnetic sensor and a gravity sensor by using an interface provided by an android system, monitoring data changes of the three sensors in real time, and performing filtering smoothing processing on the obtained data;
when the user moves, the step frequency, the step length and the direction angle information of the user are obtained by utilizing an acceleration sensor, a gyroscope sensor and a geomagnetic sensor, so that a PDR positioning result is obtained, and finally, the optimal positioning result of the user is obtained by fusing all data through a particle filter algorithm;
2. step frequency measurement by three-threshold method
2-1, preprocessing acceleration data, setting delta T as a sliding time window, and analyzing step number, step length and course information in a delta T time range, wherein in the embodiment, the delta T =3 seconds;
get the
Is processed acceleration data, wherein a x 、a y 、a z The acceleration of the X axis, the acceleration of the Y axis and the acceleration of the Z axis are measured by an acceleration sensor; the time range of the normal person going forward is 0.2s to 1s, and a large number of experiments show that the sampling frequency is set to be 50Hz, the pedestrian step frequency state can be effectively described through acceleration information, and the problem of missing detection of steps is avoided, so that in the embodiment, the sampling frequency is set to be 50Hz;
2-2, analyzing each acceleration sampling point data in the delta T range in sequence, if the current acceleration sampling point is larger than the previous sampling point and smaller than the next sampling point, recording the position of the current sampling point as a peak, and obtaining all peak positions in the delta T range in sequence;
2-3, analyzing all peak data by a three-threshold method, and removing pseudo peaks; the method comprises the following specific steps:
2-3-1, setting the threshold value of the resultant acceleration to be 1.2g: when the resultant acceleration a is less than 1.2g, determining the resultant acceleration as a pseudo peak;
2-3-2, setting a peak threshold value: avoid jitter interference and reduce the timeThe average of the three maxima in the window is set as A max The average of the three minima is set to A min Taking A max And A min The average value of (2) is a peak threshold value, and when the average value is smaller than the peak threshold value, the average value is determined as a pseudo peak;
2-3-3, setting a sampling point threshold value M: m =0.2f, f is the sampling frequency; at least 0.2 seconds are spaced from the previous peak to the next peak, namely 2 steps, and 10 sampling points are corresponding to the condition of 50Hz sampling rate; when the number of sampling points between two adjacent wave crests is less than M, determining that the second wave crest is a pseudo wave crest; otherwise, both wave crests are effective wave crests;
2-4, obtaining all effective wave peaks in the delta T time period through a three-threshold method: location = [ l = 1 ,l 2 ,l 3 ,...,l i ,...,l n ,] T If n is the total number of the effective wave crests, the step number in the delta T time period is n;
3. measuring the step length by a formula method, and analyzing the starting time and the ending time of each step according to the effective peak position;
3-1. Calculating
Obtaining the starting point of the ith step and taking the starting point as the end point of the (i-1) th step;
RANGE＝[range_1,range_2,...,range_i,...,range_n,range_L] T ；
it should be noted that range _1 represents the 1 st sampling point in the Δ T time period and serves as the starting point of the 1 st step, range _2 is the ending point of the 1 st step and also serves as the starting point of the 2 nd step, and so on, range _ n serves as the starting point of the nth step, and range _ L represents the last 1 sampling point in the Δ T time period and serves as the ending point of the nth step;
3-2, calculating the step length of each step, wherein the step length measurement mainly utilizes the linear relation between the step length and the acceleration:
wherein Q is i The total number of sampling points (including two end points) a in the ith step q The total acceleration of the q sampling point in the ith step is represented, k and b are preset constants and are obtained through least square fitting;
4. pedestrian course angle algorithm based on assistance of gradienter
In the moving process of the pedestrian, the change range of the posture of the mobile phone is large, the mobile phone can be placed in a pocket, the mobile phone can be used for receiving a call at the ear, the mobile phone can swing along with the arm, and the geomagnetic data can not accurately represent the actual advancing direction of the pedestrian in real time; as shown in fig. 5-1, when the pedestrian reads the screen with the mobile phone, the mobile phone is usually in a flat posture, the machine head faces forward, the geomagnetic data can provide an absolute heading value at the moment, and the actual heading direction of the pedestrian can be well represented without integral operation; in an indoor environment, the pedestrian motion state can be divided into two main states of long-distance straight walking and short-time right-angle turning, if the flat posture of the mobile phone can be detected, the direction data can be used until the turning action of the pedestrian is detected, and accordingly, the pedestrian motion track can be better tracked by combining objective condition limitation of indoor walls, corridors and the like and through a particle filter algorithm;
the spirit level is used for detecting whether the mobile phone is horizontally placed in the movement, if so, the geomagnetic direction data can better represent the movement direction of the pedestrian, and the specific process is as follows:
the Android can directly call a geomagnetic sensor to measure an azimuth angle, and a gyro sensor measures a pitch angle and a roll angle, wherein the heading angle is the advancing direction (-180 degrees to 180 degrees) of geomagnetic data, the pitch angle represents the angle (90 degrees to 90 degrees) for lifting or putting down the narrow side of the mobile phone, and the roll angle represents the angle (90 degrees to 90 degrees) for lifting or putting down the wide side of the mobile phone;
the three-axis gravity component of the mobile phone on the terrestrial coordinate system can also be obtained by directly calling a gravity sensor for measurement according to the Android regulation, if the mobile phone is in a completely horizontally placed state, the gravity components of the X axis and the Y axis are zero, and the gravity component of the Z axis is g (9.8); the narrow side of the mobile phone is upright, the machine head is upward, the weight component of the Y axis is g, and the other two direction components are zero; similarly, the wide side of the mobile phone is set on the side, the weight component of the X axis is g, and the weight components of the other two axes are zero (the opposite direction is-g);
therefore, in order to detect the flat posture of the mobile phone, the invention provides two standards of a data threshold and a time threshold:
data threshold value: when the mobile phone is laid horizontally, the roll angle and the pitch angle should be less than beta th In the aspect of triaxial gravity components, when a user uses the mobile phone, the user cannot be completely horizontal, as shown in fig. 5-1 and 5-3, but mostly the narrow side of the mobile phone generates an angle error with the horizontal direction, and the wide side of the mobile phone generates an angle error with the horizontal direction, as shown in fig. 5-2 and 5-4, so that the X axis of the gravity components, namely the influence of the up-and-down swing angle of the narrow side of the mobile phone on gravity, is sensitive to the judgment of the flat posture of the mobile phone; therefore, the X-axis gravity component should be less than the threshold G th (ii) a When the roll angle and the pitch angle are both less than beta th Should be less than a threshold G with respect to the X-axis gravity component th When the mobile phone is in a flat posture, the mobile phone is considered to be in a flat posture; in this example,. Beta. th At 10-20 degrees, G th Is 3;
time threshold value: when the mobile phone swings with the arm in a pocket, a short flat posture may occur, the duration is short, and the geomagnetic azimuth data at the moment is not credible due to the influence of sampling errors; therefore, only after the mobile phone is detected to be in a flat posture and continues for a period of time, the geomagnetic azimuth data at the moment can well represent the moving direction of the pedestrian; thus, the duration threshold t is set th (ii) a In this example, t th Is 3 seconds; as shown in fig. 3, in this embodiment, in the time period from the sampling point 210 to the sampling point 500, the posture of the mobile phone is detected as being placed at the horizontal flat end, and the current geomagnetic heading information is also kept stable and available, which proves that the detection is effective;
4-1, collecting corresponding time stamp t, azimuth angle alpha, pitch angle beta, roll angle chi and gravity X-axis component G at each moment x Y-axis component G y Z-axis component G z (ii) a Composition vector GRAND = [ t, alpha, beta, chi, G x ,G y ,G z ] T ；
4-2 depression at time tElevation angle beta and rolling angle chi are smaller than threshold value beta th Component G of the X-axis of gravity x Less than threshold G th Then the flag of time t is marked t ＝1；
4-3, if the flag of continuous 150 (1000 t/delta t) sampling points is detected at the sampling frequency of 50Hz times t If the direction angle is not less than 1, recording as effective sampling, and calculating the azimuth angle average value delta alpha of the period of time as a predicted initial pedestrian direction angle;
4-4, combining the indoor map and the AP layout to obtain all possible headings, and caching;
in this embodiment, the setting environment is as shown in fig. 4, and four APs are arranged at four inflection points of a rectangle; each side corresponds to the positive direction and the negative direction, namely, the pedestrian can have four headings, and the four headings are cached; when the fact that the information of the user level meter is available is detected, the predicted initial pedestrian direction angle is compared with the four cached headings to obtain the closest possible heading, and the possible heading is used as the initial pedestrian direction;
5. turn detection
5-1, setting a time window delta T orien Real-time acquisition window Δ T orien Taking the mean value of azimuth angles alpha within the range to obtain orien;
5-2, calculating the variation amplitude of the front orien and the rear orien, and if the variation amplitude exceeds a threshold value of 80 degrees, judging that a turning action occurs; as shown in fig. 3, near the sampling point 900, the geomagnetic compass data has obvious step changes, and it is determined that a turning action occurs;
5-3, detecting the signal intensity of the four inflection points AP at the position of the pedestrian in real time, wherein the inflection point where the AP with the maximum signal intensity is located is the closest inflection point of the pedestrian at the moment, and steering to the AP with the maximum signal intensity at the closest inflection point; and updating the pedestrian direction;
6. PDR dead reckoning based on a particle filter algorithm, and acquiring step frequency, step length and pedestrian direction information which are complete and correspond to the moment in the user movement process through the collected data, and drawing a user movement track by utilizing the particle filter algorithm and a map fusion algorithm; the specific process is as follows:
the particle carries 5 parameters, p = (x, y, t, w, L, θ), wherein x and y represent the abscissa and ordinate of the particle in the map respectively; t represents a time; w represents the weight of the particle, and the larger the weight is, the higher the possibility of representing the pedestrian at the position is; l represents the moving step length of the pedestrian at the moment; theta represents the pedestrian direction at the moment;
6-1, setting the weights of all the particles to be the same at the initial moment; after the pedestrian enters the indoor space, the particle filtering starts to iterate, and the actual positions and weights of the particles at different positions are iterated continuously according to the following algorithm;
6-2, determining the position change of the particles according to the step frequency, the step number and the pedestrian direction information:
wherein L is k Denotes the step size of the k step, θ k Indicating course value, x, in the k-th step k 、y k Representing the displacement of the kth step in the X-axis and Y-axis directions of the map;
6-3. Due to the indoor space limitation, when the front and back variation tracks of the particle positions have intersection points with the wall lines, the weight value is set to zero immediately:
6-4, after each iteration, resampling the particles, wherein the particles with low weight are considered to be far away from the position of a real user, redistributing the particles can concentrate limited particles in a region with higher confidence coefficient to obtain new weights of all the particles, and calculating
Obtaining the horizontal coordinate after iteration, and the vertical coordinate is the same; the places with larger weights are the places where pedestrians are more likely to appear;
and 6-5, combining the time information to obtain the moving track of the pedestrian.
The steps 3 and 4 may be performed in parallel.
While the invention has been described with reference to specific embodiments, any feature disclosed in this specification may be replaced by alternative features serving the same, equivalent or similar purpose, unless expressly stated otherwise; all of the disclosed features, or all of the method or process steps, may be combined in any combination, except mutually exclusive features and/or steps.
Claims (1)
1. A level instrument assisted indoor pedestrian trajectory estimation method comprises the following specific steps:
step 1. Data preprocessing
Acquiring data through an acceleration sensor, a gyroscope sensor, a geomagnetic sensor and a gravity sensor, and performing filtering smoothing on the acquired data;
step 2, calculating step frequency
2-1, preprocessing acceleration data, and setting a sliding time window delta T and a sampling frequency f;
for each sample point, calculate:
wherein, a x 、a y 、a z The acceleration of the X axis, the acceleration of the Y axis and the acceleration of the Z axis are measured by an acceleration sensor;
2-2, searching all wave crests in a sliding time window delta T, and marking in sequence;
and 2-3, removing the pseudo wave peaks to obtain all effective wave peaks in the sliding time window delta T: location = [ l 1 ,l 2 ,l 3 ,...,l i ,...,l n ,] T If n is the total number of the effective wave crests, the step number in the sliding time window delta T is n; removing the pseudo wave crest by adopting a three-threshold method, which specifically comprises the following steps:
2-3-1, setting the threshold value of the resultant acceleration to be 1.2g: when the resultant acceleration a is less than 1.2g, determining the resultant acceleration as a pseudo peak;
2-3-2, setting a peak threshold value: calculating the average A of the three maxima within the sliding time window DeltaT max Average of three minima, A min Taking A max And A min Average of (2) is waveA peak threshold value, which is determined as a pseudo peak when being smaller than the peak threshold value;
2-3-3, setting a sampling point threshold value M: m =0.2f; when the number of sampling points between two adjacent wave crests is less than M, determining that the second wave crest is a pseudo wave crest; otherwise, the two wave crests are both effective wave crests;
step 3, calculating step length
3-1. Calculation
Obtaining the starting point of the ith step and taking the starting point as the end point of the (i-1) th step; and further obtaining a set of all starting points and all end points:
RANGE＝[range_1,range_2,...,range_i,...,range_n,range_L] T ；
wherein l i Representing the ith effective peak in the sliding time window delta T, representing the 1 st sampling point in the sliding time window delta T as the starting point of the 1 st step by range _1, representing the last 1 sampling point in the sliding time window delta T as the ending point of the nth step by range _ L;
3-2, calculating the step length of each step:
wherein Q is i Total number of samples, a, of step i q The total acceleration of the q sampling point in the ith step is represented, and k and b are preset constants;
step 4, auxiliary calculation of pedestrian course angle based on level meter
4-1, corresponding time stamp t, azimuth angle alpha, pitch angle beta, roll angle chi and X-axis gravity component G at each moment x Y-axis gravity component G y Z-axis gravity component G z Forming a vector: GRAND = [ t, alpha, beta, chi, G x ,G y ,G z ] T (ii) a Wherein, the azimuth angle alpha is measured by a geomagnetic sensor, the pitch angle beta and the roll angle chi are measured by a gyro sensor, and the gravity component G of the X axis x Y-axis gravity component G y Z-axis gravity component G z By weightThe force sensor measures the force;
for time t, if the condition is satisfied: the pitch angle beta and the roll angle chi are both smaller than a preset threshold value beta th And the component G of the X-axis of gravity x Less than a predetermined threshold G th If the mobile phone is in the flat-laying posture at the moment t; starting to detect from zero time, and when the mobile phone is continuously detected to be in a flat posture and exceeds a time threshold t th If so, effective sampling is realized, and the average value of azimuth angles at all times in the time period is used as the predicted initial pedestrian course angle;
4-2, combining the indoor map and the AP layout to obtain all possible headings; comparing the predicted initial pedestrian course angle with all possible courses to obtain the possible course closest to the predicted initial pedestrian course angle, and taking the possible course as the initial pedestrian direction;
step 5. Turning detection
5-1, setting the collection window delta T orien Calculating Δ T for each acquisition window orien The average value orien of the azimuthal angles α within the range;
5-2, calculating the variation amplitudes of the front orien and the rear orien, and if the variation amplitudes exceed a threshold value of 80 degrees, judging that a turning action occurs;
5-3, detecting the signal intensity of all APs at the pedestrian position in real time, selecting the inflection point where the AP with the maximum signal intensity is located as the closest inflection point of the pedestrian at the turning moment, steering to the AP with the maximum signal intensity at the closest inflection point, and updating the heading angle of the pedestrian;
and 6, obtaining pedestrian track calculation by adopting a particle filter algorithm according to the step frequency, the step length and the pedestrian course angle information obtained in the steps 2 to 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910634626.0A CN110487273B (en) | 2019-07-15 | 2019-07-15 | Indoor pedestrian trajectory calculation method assisted by level gauge |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910634626.0A CN110487273B (en) | 2019-07-15 | 2019-07-15 | Indoor pedestrian trajectory calculation method assisted by level gauge |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110487273A CN110487273A (en) | 2019-11-22 |
| CN110487273B true CN110487273B (en) | 2022-11-08 |
Family
ID=68546966
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910634626.0A Active CN110487273B (en) | 2019-07-15 | 2019-07-15 | Indoor pedestrian trajectory calculation method assisted by level gauge |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110487273B (en) |
Families Citing this family (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111189443B (en) * | 2020-01-14 | 2022-11-11 | 电子科技大学 | Pedestrian navigation method for online step length calibration, motion deviation angle correction and adaptive energy management |
| CN111595344B (en) * | 2020-06-01 | 2023-02-24 | 中国矿业大学 | Multi-posture downlink pedestrian dead reckoning method based on map information assistance |
| CN113114850B (en) * | 2021-03-18 | 2021-09-21 | 电子科技大学 | Online fusion positioning method based on surveillance video and PDR |
Family Cites Families (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7747409B2 (en) * | 2004-03-12 | 2010-06-29 | Vectronix Ag | Pedestrian navigation apparatus and method |
| JP4552658B2 (en) * | 2005-01-13 | 2010-09-29 | 日立金属株式会社 | Attitude detection method using two-axis magnetic sensor and program thereof |
| KR100947046B1 (en) * | 2007-11-19 | 2010-03-10 | 황진상 | Apparatus of chasing posture of moving material object, method of chasing posture of moving material object, apparatus of chasing posture of toothbrush and method of chasing posture of toothbrush using the same |
| CN104266646B (en) * | 2014-09-11 | 2017-01-25 | 北京邮电大学 | Method and device for determining movement direction of terminal |
| CN106370178B (en) * | 2015-07-21 | 2020-09-22 | 阿里巴巴集团控股有限公司 | Attitude measurement method and device of mobile terminal equipment |
| CN105241478B (en) * | 2015-10-13 | 2018-02-16 | 中国船舶重工集团公司第七一七研究所 | A kind of single axis modulation Laser-gym Inertial Navigation System multi-coordinate scaling method |
| CN106643711A (en) * | 2016-11-25 | 2017-05-10 | 北京邮电大学 | Indoor positioning method and indoor positioning system based on hand-held device |
| CN106705968B (en) * | 2016-12-09 | 2020-11-27 | 北京工业大学 | Indoor inertial navigation algorithm based on attitude identification and step size model |
| CN107084718A (en) * | 2017-04-14 | 2017-08-22 | 桂林电子科技大学 | Indoor orientation method based on pedestrian's reckoning |
| CN108225324B (en) * | 2017-12-27 | 2021-08-17 | 中国矿业大学 | Indoor positioning method based on intelligent terminal and integrating geomagnetic matching and PDR |
| CN108632761B (en) * | 2018-04-20 | 2020-03-17 | 西安交通大学 | Indoor positioning method based on particle filter algorithm |
| CN108917760B (en) * | 2018-04-25 | 2021-01-19 | 西安交通大学 | Magnetic field correction-based pedestrian dead reckoning course angle correction method |
| CN109211229A (en) * | 2018-08-26 | 2019-01-15 | 桂林电子科技大学 | A kind of personnel's indoor orientation method based on mobile phone sensor and WiFi feature |
| CN109460040A (en) * | 2018-12-28 | 2019-03-12 | 珠海凯浩电子有限公司 | It is a kind of that map system and method are established by mobile phone shooting photo array floor |
-
2019
- 2019-07-15 CN CN201910634626.0A patent/CN110487273B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN110487273A (en) | 2019-11-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110487273B (en) | Indoor pedestrian trajectory calculation method assisted by level gauge | |
| CN108844533B (en) | Free attitude PDR positioning method based on multi-sensor fusion and attitude calculation | |
| CN104121905B (en) | Course angle obtaining method based on inertial sensor | |
| CN106705968B (en) | Indoor inertial navigation algorithm based on attitude identification and step size model | |
| CN106908060A (en) | A kind of high accuracy indoor orientation method based on MEMS inertial sensor | |
| WO2017088196A1 (en) | Fusion navigation device and method based on wireless fingerprints and mems sensor | |
| CN105674984B (en) | A method of acquisition for mobile terminal pedestrian advancing direction is used without constraint | |
| CN105865450A (en) | Zero-speed update method and system based on gait | |
| CN110579207B (en) | Indoor positioning system and method based on combination of geomagnetic signals and computer vision | |
| CN108537101B (en) | Pedestrian positioning method based on state recognition | |
| CN102128625A (en) | Initial matching method for use in gravimetric map matching in gravity-aided inertial navigation system | |
| CN110986941B (en) | Method for estimating installation angle of mobile phone | |
| CN109612463B (en) | Pedestrian navigation positioning method based on lateral speed constraint optimization | |
| CN110412596A (en) | A kind of robot localization method based on image information and laser point cloud | |
| CN109459028A (en) | A kind of adaptive step estimation method based on gradient decline | |
| CN109629379A (en) | A kind of pavement detection method based on mobile phone sensor data | |
| Cho et al. | Threshold-less zero-velocity detection algorithm for pedestrian dead reckoning | |
| US10466054B2 (en) | Method and system for estimating relative angle between headings | |
| CN109708631A (en) | A kind of pedestrian navigation navigational calibration method based on mobile phone modal idenlification | |
| CN104546391B (en) | Gyro stabilizer for tactile sticks and complementary filtering method thereof | |
| CN112362057A (en) | Inertial pedestrian navigation algorithm based on zero-speed correction and attitude self-observation | |
| CN113325455B (en) | Method and system for tracking and determining indoor position of object | |
| CN204788411U (en) | Wall climbing robot navigational positioning system based on sound localization , INS | |
| CN111435083A (en) | Pedestrian track calculation method, navigation method and device, handheld terminal and medium | |
| 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 |