CN109099922A - A kind of indoor orientation method based on physical field - Google Patents
A kind of indoor orientation method based on physical field Download PDFInfo
- Publication number
- CN109099922A CN109099922A CN201810919750.7A CN201810919750A CN109099922A CN 109099922 A CN109099922 A CN 109099922A CN 201810919750 A CN201810919750 A CN 201810919750A CN 109099922 A CN109099922 A CN 109099922A
- Authority
- CN
- China
- Prior art keywords
- state
- pedestrian
- thr
- particle
- population
- 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.)
- Withdrawn
Links
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
- G01C21/20—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)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a kind of indoor orientation method based on physical field, the method is a kind of indoor orientation method that improved pedestrian's reckoning is merged with earth magnetism finger print matching method.Wherein, the cadence detection algorithm False Rate based on finite state machine is low and calculation amount is smaller；Adaptive step model can replace step-length model parameter according to pedestrian movement's state, realize adaptive step estimation, and the pedestrian estimated is more accurate per length step by step；The course calculated after slope compensation and declination compensation is more accurate；Judge whether that resampling can obtain balance between weight degeneration and diversity scarcity according to effective particle scale, carrying out particle coarsening after resampling can increase the diversity of particle, and then improve the robustness of algorithm.Inventive algorithm complexity is low, and accuracy is high, and does not depend on any nodal information, and locating desired information is provided by physical field completely, has important practical value.
Description
Technical field
The present invention relates to field of locating technology, and in particular to a kind of indoor positioning side merged based on PDR and earth magnetism fingerprint
Method.
Background technique
Location technology is divided into outdoor positioning technology and two kinds of indoor positioning technologies, and common indoor orientation method has very much
Kind, having for most common of them is following several: (1) pedestrian's reckoning (PDR): PDR is a kind of relative positioning method, it passes through reality
When measure cadence, step-length and the course information of pedestrian, not against external information autonomous positioning, precision is high in a short time, stability
It is good.But PDR will receive the influence of sensor characteristics, error can accumulate at any time.(2) WIFI is positioned: WIFI positioning is to pass through
The signal strength and prior established fingerprint database for acquiring target point compare to obtain the location information of target point.WIFI
Error hiding rate is low but positioning result fluctuation is significant, and need to dispose enough nodes just and can guarantee the accuracy of positioning.
(3) ultra wide band (UWB): super-broadband tech is to transmit data by sending nanosecond or nanosecond Ultra-short pulse below, can
To obtain Ghz grades of data bandwidth, indoor positioning is realized.But UWB is difficult to realize a wide range of in-door covering, and system Construction
Cost is high.(4) geomagnetic matching (MM): geomagnetic matching is the reference map that will be stored in real-time Geomagnetism Information and geomagnetic database
It is matched, positioning can be realized by finding out most like point according to the fitting degree of both certain criterion judgements.However it is local
When scheming excessive, geomagnetic matching can have fuzzy solution.
Currently, having certain limitation using the indoor orientation method of monotechnics, and part indoor orientation method needs
Arrange that great deal of nodes could be completed to position, being able to satisfy current all indoor positioning services there are no a kind of generalization technology needs
It asks, is badly in need of a kind of indoor orientation method that a variety of method characteristics have complementary advantages.
Summary of the invention
The present invention for earth's magnetic field position indoors in there are fuzzy solution and pedestrian's reckoning (PDR) there are accumulated errors
The problem of, a kind of indoor orientation method that improved pedestrian's reckoning is merged with earth magnetism finger print matching method is provided, is not depended on
Any node reduces geomagnetic matching region centered on PDR positioning result, solves earth magnetism using improved particle filter algorithm
The fuzzy solution problem of fingerprint reaches amendment PDR accumulated error in real time, the effective purpose for promoting indoor position accuracy.
It mainly include improved to pedestrian's cadence detecting step, step-length detecting step and course in technical solution of the present invention
Process of solution.In the cadence detecting step, when a (the three axis resultant accelerations for representing pedestrian) have passed through following whole states
After be calculated as pedestrian and stepped a step:
(1) original state is S0, and when a is greater than Thr, state is transferred to S1 by S0, otherwise in S0 cycle of states.
(2) in S1 state, when a is greater than Peak_Thr, state is transferred to S2 by S1, when a is less than Thr, is jumped back to by S1
S0, otherwise in S1 cycle of states.
(3) in S2 state, when a is greater than Neg_Thr and is less than Peak_Thr, state is transferred to S3 by S2, otherwise in S2
Cycle of states.
(4) in S3 state, when a is less than Neg_Thr, state is transferred to S4 by S3, when a is greater than Peak_Thr, state
S2 is jumped back to by S3, is otherwise recycled in state S3.
(5) in S4 state, when a is greater than Neg_Thr and is less than Thr, state is transferred to S5 by S4, otherwise follows in state S4
Ring.
(6) in S5 state, when a is greater than Thr, state is transferred to S0 by S5, and when a is less than Neg_Thr, state is jumped by S5
S4 is returned, is otherwise recycled in state S5.
In above-mentioned steps: Thr represents the detection threshold value of a；Peak_Thr represents the peak threshold of a；Neg_Thr represents a's
Valley threshold；S0 represents pedestrian's stationary state；S1 represents a rising；S2 represents a and ascends into wave crest state；S3 represent a decline from
It opens wave crest and enters trough state；S4 represents a in trough cycle of states；S5 represents a and leaves trough recurrence stationary state.
Step-length measurement in the step-length detecting step differentiates that pedestrian is in usual walking level state using acceleration variance
Or high speed ambulatory status, when pedestrian is in usual walking level state, step size computation formula are as follows: li=0.35fi+0.48vi+
0.22.When pedestrian is in high speed ambulatory status, step size computation formula are as follows: li=0.30fi+0.05vi+0.72.Wherein: liIt is
The step-length of pedestrian；fiIt is the cadence of pedestrian；viIt is the three axis resultant acceleration variances of pedestrian.
Course angle calculation formula in the course process of solution are as follows: θi=atan (Hy/Hx)+Mag_dec；In formula: θi
It is the course angle of pedestrian, Hx、HyIt is X, Y direction absolute force respectively；Mag_dec is local magnetic declination.
In technical solution of the present invention, a kind of improved cadence, step-length, course information according to pedestrian is additionally provided
And the particle filter algorithm that earth magnetism reference map carries out, key step include the following:
(1) population is initialized.
(2) population state is updated.
(3) whether detection particle is through walls and removes particle through walls.
(4) weight of each particle is calculated.
(5) judge whether to need to carry out resampling to population according to effective particle scale Ness, its calculation formula is:Given thresholdWork as Ness< NessWhen _ Thr, then resampling is carried out；If Ness≥
NessWhen _ Thr, then without resampling；Wherein: n represents n-th of particle, and N indicates the number of particles of population；ωnExpression is returned
Weight shared by n-th of particle after one change.
(6) if having carried out resampling to population, population is roughened, coarsening process includes: first, is adopted again
A noise is increased to each particle after sample, it is made to scatter in state space；Second, population state is updated in next step
When, increase the random error of step-length and course.
Detailed description of the invention
Fig. 1 is the indoor orientation method flow chart merged based on PDR and earth magnetism fingerprint.
Fig. 2 is the finite state machine flow chart of cadence detection.
Fig. 3 is particle filter algorithm flow chart.
In Fig. 2: Input represents three axis resultant acceleration values, and what Thr was represented is detection threshold value, and what Peak_Thr was represented is peak
It is worth threshold value, what Valley_Thr was represented is valley threshold.
Specific embodiment
Design scheme of the invention is elaborated with embodiment below.
Indoor locating system based on physical field is relied on comprising three axis accelerometer, three-axis gyroscope, three axle magnetometer
Smart phone.Earth gravitational field, these included physical field information of earth's magnetic field are obtained by the sensor in smart phone, it is right
Pedestrian carries out reckoning, constructs earth magnetism reference map using geomagnetic field information, in each gait week of each pedestrian's reckoning
The accurate location for obtaining pedestrian when geomagnetic matching using particle filter is carried out in phase.
Specific mainly includes following content:
(1) cadence detects:
The pedestrian acceleration's data collected according to three axis accelerometer in smart phone, it is complete using finite state machine method
It is detected at effective cadence of pedestrian, and records each initial time stamp effectively walked.
(2) adaptive step is estimated:
The each step-length effectively walked of pedestrian is estimated according to adaptive step model.
(3) course resolves:
Slope compensation is carried out to magnetometer data according to the angular velocity data that gyroscope collects, after slope compensation
Magnetometer data calculate behind absolute course along with local magnetic declination calculates the direction of motion of current gait.
(4) particle filter:
Particle filter is carried out according to obtained cadence, step-length, course information and earth magnetism reference map, it is accurate to estimate pedestrian
Position.
Cadence described in one, detection the following steps are included:
S11: the acceleration information that three axis accelerometer collects in smart phone is denoted as A={ ax、ay、az}。
S12: the 3-axis acceleration data A that handheld device is collected, smothing filtering is carried out, and calculate three axis and close and add
Speed a:The purpose of smothing filtering is smoothed data, filters out noise.The reason of using three axis resultant accelerations
It is that can be avoided equipment installation bring extra error, without considering the specific direction of accelerometer, the waveform energy of resultant acceleration
It is enough to stablize the variation in [- g, g].
S13: being detected using cadence of the finite state machine to pedestrian, obtains effective pedestrian's step number.Finite state machine
In each state respectively represent: S0: user's stationary state；S1: acceleration rises, and pedestrian is possibly into motion state；S2: acceleration
Ascend into wave crest state；S3: acceleration is lowered away from wave crest and enters trough state；S4:a is in trough cycle of states；S5: accelerate
Degree leaves trough and returns stationary state.
As shown in Fig. 2, the original state of the cadence detection algorithm based on finite state machine is S0, and when a is greater than Thr, shape
State is transferred to S1 by S0, otherwise in S0 cycle of states.
In S1 state, when a is greater than Peak_Thr, state is transferred to S2 by S1, when a is less than Thr, jumps back to S0 by S1,
Otherwise in S1 cycle of states.
In S2 state, when a is greater than Valley_Thr and is less than Peak_Thr, state is transferred to S3 by S2, otherwise in S2
Cycle of states.
In S3 state, when a is less than Valley_Thr, state is transferred to S4 by S3, when a is greater than Peak_Thr, state
S2 is jumped back to by S3, is otherwise recycled in state S3.
In S4 state, when a is greater than Valley_Thr and is less than Thr, state is transferred to S5 by S4, otherwise follows in state S4
Ring.
In S5 state, when a is greater than Thr, state is transferred to S0 by S5, and when a is less than Valley_Thr, state is jumped by S5
S4 is returned, is otherwise recycled in state S5.
Only when three axis resultant acceleration a after S0~S5 state, are just thinking that pedestrian has stepped a step.It is converted in state
In figure, what Thr was represented is detection threshold value, and what Peak_Thr was represented is peak threshold, and what Valley_Thr was represented is valley threshold.
Specific Thr, Peak_Thr and Valley_Thr can make change for different pedestrians.In embodiments of the present invention, excellent
Thr is selected to be set as 9.8, Peak_Thr and be set as 10.5, Valley_Thr being set as 9.1.Finite state machine method overcomes peak value
The shortcomings that peak value and threshold value relationship are only focused in detection method, is set as a kind of state for peak value and valley, focuses more on acceleration
The situation of change of data.This method has preferable processing capacity to the good acceleration information of sine wave rule.
S14: each initial time stamp effectively walked is recordedIn formulaRepresent the i-th effectively walked
Begin the time.
Adaptive step described in two, estimation the following steps are included:
S21: the paces frequency f each effectively walked is calculatediWith acceleration variance vi, paces frequency fiCalculation formula are as follows:In formula: tsIt is stabbed at the beginning of representing the currently active step, teRepresent the ending time stamp of the currently active step.
Acceleration variance viCalculation formula are as follows:In formula: atIt is acceleration transducer t moment
Acceleration value,It is the three axis resultant acceleration average values that pedestrian currently walks, K is the sampled point of the currently active step acceleration transducer
Quantity.
S22: judge the acceleration variance v of the currently active stepiWhether be greater than threshold value 3.8, if it is greater than or be equal to 3.8, that
Determine that pedestrian is high speed ambulatory status, if it is less than 3.8, then determining that pedestrian is in usual walking level state.
S23: adaptive step model can indicate are as follows: li=astep·fi+bstep·vi+cstep；In formula: astep、bstep、
cstepFor the parameter of adaptive step model, when determining that current pedestrian is in usual walking level state, astep、bstep、cstepValue
0.35,0.48 and 0.22 is taken respectively, when determining that current pedestrian is in high speed ambulatory status, astep、bstep、cstepValue difference
Take 0.3,0.05 and 0.72.
The present invention differentiates that pedestrian is in usual walking level state or high speed ambulatory status using acceleration variance, adaptively
Step-length model can adaptive adjustment pedestrian be in the step-length model parameter under different motion state, obtained estimation step-length is more
It is accurate to add, while algorithm is more stable, is suitble to prolonged indoor positioning.
Course described in three, resolve the following steps are included:
S31: the three-axis gyroscope data of recorder's institute's holding equipment in the process of walking are denoted as B={ ωx、ωy、ωz, three
Axis magnetometer data is denoted as C={ mx、my、mz, using low-pass filter to three-axis gyroscope data and three axle magnetometer data
Low-pass filtering is carried out, can effectively remove and shake caused noise in pedestrian's walking process.
S32: the pitch angle (pitch) and cross within each effectively step are calculated according to the gyro data after low-pass filtering
Roll angle (roll)；The calculation formula of pitch angle (pitch) are as follows:The calculation formula of roll angle (roll) are as follows:ω in formulax、ωyThe data for being three-axis gyroscope after low-pass filtering.
S33: slope compensation is carried out to magnetometer data according to the calculated attitude angle of acceleration information, obtains magnetic force counting
According to projection in the horizontal direction: Hy=-my·cos(roll)+mz·sin(roll)；Hz=mx·cos(pitch)+my·
sin(roll)·sin(pitch)+mz·cos(roll)·sin(pitch)；M in formulax、my、mzFor three axle magnetometer process
Data after low-pass filtering.
S34: course angle resolves: θi=atan (Hy/Hx)+Mag_dec；Mag_dec in formula is local magnetic declination, can
By tabling look-up to obtain；It, can be effective after gyro data calculates attitude angle to magnetometer data progress pose compensation
Elimination pedestrian's handheld device when brought heeling error further plus local magnetic declination can make the boat calculated
It is more accurate to angle.
Particle filter described in four, the following steps are included:
S41: initialization population: initializing the number of particles of population, and N=5000, the expression pair of each particle is arranged
One of current state assumes and has weights omega, and all particles are uniformly distributed in space, the weight of initial each particle
For
S42: update population state: the step-length and course information obtained according to abovementioned steps updates the attribute of population,
X in formulai、yiIndicate the position coordinates of the i-th step, li+1And θi+1Respectively indicate the step weighted by particle weights
Long and direction value, Δ l and Δ θ indicate the random error of step-length and direction, equal Gaussian distributed.
S43: whether detection particle is through walls and removes particle through walls: in practical applications, understanding the position of some particle
It appears on the position that user can not reach, this phenomenon is referred to as " through walls ", is removed according to earth magnetism reference map information through walls
Particle.
S44: the weights omega of each particle is calculatedn:
In formulaIndicate that n-th of particle obtains the corresponding earth magnetism fingerprint of state S, z in the i-th step from fingerprint basei
Indicate the absolute force value measured in the i-th step, σ is that the offset criteria of magnetometer is poor.
S45: judge whether to need resampling, the calculation formula of Ness according to effective particle scale Ness are as follows:
The value of Ness is smaller to show that sample degeneracy is more serious, given thresholdWhen
Ness< NessWhen _ Thr, then resampling is carried out, if Ness≥Ness_ Thr, then without resampling.
S46: if having carried out resampling, being roughened population, and coarsening process includes: first, right after resampling
Each particle increases a noise, it is made to scatter in state space, and second, when updating population state in next step, increase step
Long and direction random error.
S47: estimation pedestrian's actual position, the true shape for being finally averagely used as user current the state weight of all particles
State
ω in formulanIndicate weight shared by n-th of particle after normalizing.
Beneficial effects of the present invention: in the indoor orientation method that the present invention provides, the cadence detection based on finite state machine
Algorithm False Rate is low and calculation amount is smaller；Adaptive step model can replace step-length model parameter according to pedestrian movement's state,
Realize adaptive step estimation, the pedestrian estimated is more accurate per length step by step；It is solved after slope compensation and declination compensation
The course of calculating is more accurate；Judge whether that resampling can degenerate and diversity scarcity in weight according to effective particle scale
Between obtain balance, particle coarsening is carried out after resampling can increase the diversity of particle, and then improve the robustness of algorithm.This
Invention algorithm complexity is low, and accuracy is high, and does not depend on any nodal information, and locating desired information is provided by physical field completely,
With important practical value.
Claims (5)
1. a kind of indoor orientation method based on physical field, the method includes the cadence detecting steps to pedestrian, adaptive step
Long estimating step and course process of solution, it is characterised in that: in the cadence detecting step, when a have passed through following whole shapes
It is calculated as pedestrian after state and has stepped a step:
(1) original state is S0, and when a is greater than Thr, state is transferred to S1 by S0, otherwise in S0 cycle of states；
(2) in S1 state, when a is greater than Peak_Thr, state is transferred to S2 by S1, when a is less than Thr, jumps back to S0 by S1, no
Then in S1 cycle of states；
(3) in S2 state, when a is greater than Neg_Thr and is less than Peak_Thr, state is transferred to S3 by S2, otherwise in S2 state
Circulation；
(4) in S3 state, when a is less than Neg_Thr, state is transferred to S4 by S3, and when a is greater than Peak_Thr, state is by S3
S2 is jumped back to, is otherwise recycled in state S3；
(5) in S4 state, when a is greater than Neg_Thr and is less than Thr, state is transferred to S5 by S4, otherwise recycles in state S4；
(6) in S5 state, when a is greater than Thr, state is transferred to S0 by S5, and when a is less than Neg_Thr, state is jumped back to by S5
Otherwise S4 is recycled in state S5；
Wherein: a represents the three axis resultant accelerations of pedestrian；Thr represents the detection threshold value of a；Peak_Thr represents the peak threshold of a；
Neg_Thr represents the valley threshold of a；S0 represents pedestrian's stationary state；S1 represents a rising；S2 represents a and ascends into wave crest shape
State；S3, which represents a and is lowered away from wave crest, enters trough state；S4 represents a in trough cycle of states；S5 represents a and leaves trough recurrence
Stationary state.
2. according to the method described in claim 1, it is characterized in that in the adaptive step estimating step, when pedestrian is in general
When logical ambulatory status, step size computation formula are as follows: li=0.35fi+0.48vi+0.22；When pedestrian is in high speed ambulatory status, step
Long calculation formula are as follows: li=0.30fi+0.05vi+0.72；Wherein: liIt is the step-length of pedestrian；fiIt is the cadence of pedestrian；viIt is pedestrian
Three axis resultant acceleration variances.
3. according to the method described in claim 1, it is characterized in that course angle calculation formula in the course process of solution are as follows:
θi=atan (Hy/Hx)+Mag_dec；In formula: θiIt is the course angle of pedestrian, Hx、HyIt is X, Y direction absolute force respectively；Mag_
Dec is local magnetic declination.
4. according to the method described in claim 2, it is characterized by: the method for discrimination of pedestrian's ambulatory status, is that judgement currently has
Imitate the acceleration variance v of stepiWhether threshold value 3.8 is greater than；If it is greater than or be equal to 3.8, then determine pedestrian be high speed walking shape
State, if it is less than 3.8, then determining that pedestrian is in usual walking level state.
5. method according to any of claims 1-4, the method also includes the cadences, step-length, boat according to pedestrian
The particle filter step carried out to information and earth magnetism reference map, the particle filter step includes following content:
(1) population is initialized；
(2) population state is updated；
(3) whether detection particle is through walls and removes particle through walls；
(4) weight of each particle is calculated；
(5) judge whether to need to carry out resampling to population according to effective particle scale Ness, its calculation formula is:Given thresholdWork as Ness< NessWhen _ Thr, then resampling is carried out；If Ness≥
NessWhen _ Thr, then without resampling；Wherein: n represents n-th of particle, and N indicates the number of particles of population；ωnExpression is returned
Weight shared by n-th of particle after one change；
(6) if having carried out resampling to population, population is roughened, coarsening process includes: first, resampling it
A noise is increased to each particle afterwards, it is made to scatter in state space；Second, when updating population state in next step, add
The random error of big step-length and course.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810919750.7A CN109099922A (en) | 2018-08-14 | 2018-08-14 | A kind of indoor orientation method based on physical field |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810919750.7A CN109099922A (en) | 2018-08-14 | 2018-08-14 | A kind of indoor orientation method based on physical field |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109099922A true CN109099922A (en) | 2018-12-28 |
Family
ID=64849448
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810919750.7A Withdrawn CN109099922A (en) | 2018-08-14 | 2018-08-14 | A kind of indoor orientation method based on physical field |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109099922A (en) |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109982263A (en) * | 2019-04-04 | 2019-07-05 | 中国矿业大学 | A kind of WiFi fingerprint base update method based on inertia measurement tracing point |
| CN111358472A (en) * | 2020-03-27 | 2020-07-03 | 山东建筑大学 | Gait detection algorithm, device and system |
| CN114136321A (en) * | 2021-11-19 | 2022-03-04 | 中国船舶重工集团公司第七0七研究所 | Ocean multi-physical field information fusion positioning method based on homologous features |
| CN115235479A (en) * | 2022-09-23 | 2022-10-25 | 江西省智能产业技术创新研究院 | Positioning method and device of automatic guided vehicle, readable storage medium and electronic equipment |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103175529A (en) * | 2013-03-01 | 2013-06-26 | 上海美迪索科电子科技有限公司 | Pedestrian inertial positioning system based on indoor magnetic field feature assistance |
| CN103954295A (en) * | 2014-05-04 | 2014-07-30 | 中国科学院计算技术研究所 | Step-counting method based on acceleration sensor |
| US20140288822A1 (en) * | 2013-03-22 | 2014-09-25 | Qualcomm Incorporated | Controlling position uncertainty in a mobile device |
| CN104215238A (en) * | 2014-08-21 | 2014-12-17 | 北京空间飞行器总体设计部 | Indoor positioning method of intelligent mobile phone |
| CN105865448A (en) * | 2016-03-18 | 2016-08-17 | 常州大学 | Indoor positioning method based on IMU |
-
2018
- 2018-08-14 CN CN201810919750.7A patent/CN109099922A/en not_active Withdrawn
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103175529A (en) * | 2013-03-01 | 2013-06-26 | 上海美迪索科电子科技有限公司 | Pedestrian inertial positioning system based on indoor magnetic field feature assistance |
| US20140288822A1 (en) * | 2013-03-22 | 2014-09-25 | Qualcomm Incorporated | Controlling position uncertainty in a mobile device |
| CN103954295A (en) * | 2014-05-04 | 2014-07-30 | 中国科学院计算技术研究所 | Step-counting method based on acceleration sensor |
| CN104215238A (en) * | 2014-08-21 | 2014-12-17 | 北京空间飞行器总体设计部 | Indoor positioning method of intelligent mobile phone |
| CN105865448A (en) * | 2016-03-18 | 2016-08-17 | 常州大学 | Indoor positioning method based on IMU |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109982263A (en) * | 2019-04-04 | 2019-07-05 | 中国矿业大学 | A kind of WiFi fingerprint base update method based on inertia measurement tracing point |
| CN111358472A (en) * | 2020-03-27 | 2020-07-03 | 山东建筑大学 | Gait detection algorithm, device and system |
| CN111358472B (en) * | 2020-03-27 | 2022-08-23 | 山东建筑大学 | Gait detection algorithm, device and system |
| CN114136321A (en) * | 2021-11-19 | 2022-03-04 | 中国船舶重工集团公司第七0七研究所 | Ocean multi-physical field information fusion positioning method based on homologous features |
| CN114136321B (en) * | 2021-11-19 | 2023-06-20 | 中国船舶重工集团公司第七0七研究所 | Ocean multi-physical-field information fusion positioning method based on homologous characteristics |
| CN115235479A (en) * | 2022-09-23 | 2022-10-25 | 江西省智能产业技术创新研究院 | Positioning method and device of automatic guided vehicle, readable storage medium and electronic equipment |
| CN115235479B (en) * | 2022-09-23 | 2022-12-06 | 江西省智能产业技术创新研究院 | Positioning method and device of automatic guided vehicle, readable storage medium and electronic equipment |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109099922A (en) | A kind of indoor orientation method based on physical field | |
| CN109579853B (en) | Inertial navigation indoor positioning method based on BP neural network | |
| CN103925923B (en) | A kind of earth magnetism indoor locating system based on adaptive particle filter device algorithm | |
| CN104969030B (en) | Inertial device, methods and procedures | |
| US10267646B2 (en) | Method and system for varying step length estimation using nonlinear system identification | |
| AU781848B2 (en) | Pedestrian navigation method and apparatus operative in a dead reckoning mode | |
| CN109827577A (en) | High-precision inertial navigation location algorithm based on motion state detection | |
| CN110553643B (en) | Pedestrian self-adaptive zero-speed updating point selection method based on neural network | |
| CN110579207B (en) | Indoor positioning system and method based on combination of geomagnetic signals and computer vision | |
| CN106643715A (en) | Indoor inertial navigation method based on bp neural network improvement | |
| CN108844533A (en) | A kind of free posture PDR localization method based on Multi-sensor Fusion and attitude algorithm | |
| CN103884337A (en) | Multi-attitude step correction positioning system and method | |
| CN104061934A (en) | Pedestrian indoor position tracking method based on inertial sensor | |
| CN105102928A (en) | Inertial device, method, and program | |
| TWI442019B (en) | Position determination apparatus and system and position determination method thereof | |
| EP3204721A1 (en) | Pedestrian dead reckoning position tracker | |
| CN105865450A (en) | Zero-speed update method and system based on gait | |
| CN109115207A (en) | Pedestrian's foot path detection method, apparatus and system | |
| CN109323695A (en) | A kind of indoor orientation method based on adaptive Unscented kalman filtering | |
| CN108537101B (en) | Pedestrian positioning method based on state recognition | |
| CN109459028A (en) | A kind of adaptive step estimation method based on gradient decline | |
| CN112729301A (en) | Indoor positioning method based on multi-source data fusion | |
| JP6804908B2 (en) | Estimator, estimation method and computer program | |
| CN104546391B (en) | Gyro stabilizer for tactile sticks and complementary filtering method thereof | |
| CN108801267A (en) | It is a kind of to merge the indoor without anchor point localization method of multisensor |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20181228 |