CN106643715A - Indoor inertial navigation method based on bp neural network improvement - Google Patents
Indoor inertial navigation method based on bp neural network improvement Download PDFInfo
- Publication number
- CN106643715A CN106643715A CN201611020582.5A CN201611020582A CN106643715A CN 106643715 A CN106643715 A CN 106643715A CN 201611020582 A CN201611020582 A CN 201611020582A CN 106643715 A CN106643715 A CN 106643715A
- Authority
- CN
- China
- Prior art keywords
- paces
- acceleration
- data
- neural network
- inertial navigation
- 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.)
- Pending
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
- G01C21/10—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
- G01C21/18—Stabilised platforms, e.g. by gyroscope
-
- 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 discloses an indoor inertial navigation method based on bp neural network improvement. The indoor inertial navigation method comprises an offline training phase and an online positioning phase. The offline phase comprises the following steps: establishing a step size estimation model of a bp neural network; setting the number of layers of the neural network and the number of nerve cells and constructing a neural network model with 2k0 inputs and a single output, wherein the inputs are accelerations of k0 moments before a current moment k and angular velocity data, and the output is step size; obtaining and pre-treating original data: a user carries mobile equipment to move in a room, and the original data is obtained by a tri-axial acceleration sensor and a tri-axial angular velocity sensor in the mobile equipment in real time, which respectively contain scalar sizes on three dimensions, are recoded as sequence signal sequences and are subjected to low-pass filtering pretreatment; training a step recognition model to obtain an initial step recognition model. The indoor inertial navigation method disclosed by the invention has the characteristic of higher precision.
Description
Technical field
The present invention relates to a kind of indoor inertial navigation method, more particularly to a kind of indoor inertial navigation method.
Background technology
With the development of modern location and navigation technology, location Based service (LBS) is increasingly becoming in Intelligent life can not
A part for acquisition, Global Navigation System provides the positioning service of globalization, because building is blocked to signal, indoors
Cannot be serviced using GPS, the rapid development that the indoor positioning technologies for thus arising at the historic moment are obtained, people also gradually start to adapt to
Various indoor positioning services.
Inertial navigation system (inertial navigation system, INS) is measured using accelerometer and gyroscope
The acceleration and angular speed of positioning target, using these data position, attitude, the speed of positioning target are calculated, in space flight, boat
Extensively apply in the fields such as sea, guided weapon, vehicle.
With MEMS development and use in a large number, small-sized INS starts to occur, and becomes the focus in indoor positioning field.
Indoor positioning field is commonly referred to as walked using the system of user's positioning that dead reckoning (Dead-reckoning, DR) is walking
Passerby's dead reckoning (PDR) system, in PDR systems in addition to using INS, that relatively common is paces heading system (Step-
And-Heading System, SHS), PDR systems are all using the dead reckoning lower a moment of sectional displacement from eve
Position.INS is to utilize the sensing data at each moment to calculate and track the track of target in complete three-dimensional space, and
SHS (represents the step-length of each step dedicated for the location estimation of pedestrian by a motion vector [distance, heading]
And direction) calculate pedestrian current location.
PDR system costs are low, easy to use, and are independent of external information, also not to environmental effects, but PDR
System is relative positioning, it is desirable to have the position initialization of an outside, or can only start positioning from default origin, and only short
Possess higher precision and reliability in phase, over time with the increase of distance, the cumulative errors of PDR systems can be more and more tighter
Weight.
Paces are detected and the degree of accuracy of step recognition plays conclusive effect to the positioning precision of PDR systems, at present, are utilized
The paces of accelerometer detection people's walking can reach very high accuracy rate, but the evaluated error of paces length is larger.
Traditional strapdown inertia is taken to carry out the method for quadratic integral to acceleration information to estimate step-length, but with it is high-precision specially
Industry inertial navigation equipment ratio, the acceleration datum error that MEMS is measured is larger, therefore the error that inertial navigation positioning is obtained can be rapid
Increase.There are certain methods that sensor is fixed on into the positions such as sole, leg, waist, by zero-velocity curve come to per a bit of position
Shifting is integrated respectively computing, can reach of a relatively high precision, but sensor is fixed on into specific position and attitude exists
It is in many cases inapplicable.
Neutral net (NN) is one of non-linear input-output mappings most efficient method, can be approached arbitrarily complicated non-
Linear relationship, and with powerful learning ability, memory capability and computing capability.Neutral net is by substantial amounts of neuron and god
Connection composition between Jing is first, in the information processing for imitating human brain real system in various degree and on level.
The content of the invention
The present invention provides a kind of indoor inertial navigation method that can improve positioning precision, and technical scheme is as follows：
A kind of indoor inertial navigation method improved based on bp neutral nets, including off-line training step and tuning on-line rank
Section：
1) off-line training step, comprises the following steps
A. the step-size estimation model of a bp neutral net is set up；The neutral net number of plies and neuron number are set, are built
One has 2k0Individual input and the neural network model of single output, are input into as k before current time k0The acceleration at individual moment
With angular velocity data, the length of paces is output as；
B. obtain initial data and pre-processed：User carries mobile device and moves indoors, by mobile device
3-axis acceleration sensor is obtained in real time with three axis angular rate sensors, respectively comprising the scalar size in three dimensions, record
For clock signal sequence, and carry out LPF pretreatment；
C. the training of step recognition model：Collect respectively under a large amount of various states the pretreated acceleration of Jing steps b with
Angular velocity data, using each data corresponding to each section of complete paces as one group of training data, to based on bp neutral nets
Step-size estimation model be trained, determine the initial value of each parameter of model, obtain initial step recognition model；
2) the tuning on-line stage, comprise the following steps
A. obtain initial data and pre-processed：By user carry mobile device in 3-axis acceleration sensor with
Three axis angular rate sensors are obtained in real time, respectively comprising the scalar size in three dimensions, are recorded as clock signal sequence, are gone forward side by side
Row LPF is pre-processed；
B. paces detection；Time window is set, the acceleration and angular speed of actual time window is updated；By accelerating the number of degrees
Value detects whether to produce new paces with the comparison of upper lower threshold value；If paces are detected successfully, into step c, otherwise after
Continuous detection；
C. paces length estimate；The nerve net that acceleration in actual time window and angular velocity data input are trained
Network model, exports current paces length；
D. direction estimation；Using the magnetometer measures absolute force in equipment, estimate the direction of motion, calculate the direction of motion with
The angle in north；
E. position；Coordinate offset is calculated by the displacement of current time k, finally according to the position at upper a moment, in position fixing process
Used in expanded Kalman filtration algorithm calculate current location.
The estimation problem of paces length is processed into the present invention mapping of acceleration, angular velocity data and paces length, is adopted
Collection mass data be based on accordingly bp neutral net paces length estimate models to train, and estimates so as to obtain high-precision step-length
Meter, carries out on this basis indoor inertial navigation, improves its positioning precision.
Description of the drawings
Fig. 1 shows the relation block diagram of holistic approach of the present invention.
Fig. 2 shows the FB(flow block) of the indoor inertial navigation method improved using bp neutral nets.
Specific embodiment
The indoor inertial navigation method improved based on bp neutral nets of the present invention is done further below in conjunction with the accompanying drawings
Description.
Illustrate by instantiation of smart mobile phone, using accelerometer, turn meter, magnetometer record user's walking
Data and carry out indoor navigation.The indoor inertial navigation method improved based on bp neutral nets of the present invention, including offline instruction
Practice stage and tuning on-line stage：
1) off-line training step, comprises the following steps
A. the step-size estimation model of a bp neutral net is set up.The neutral net number of plies and neuron number are set, are built
One has 2k0Individual input and neural network model f () of single output, are represented by
Y (k)=f (A (k-k0),…,A(k),Ω(k-k0),…,Ω(k))
Wherein input is current k0Acceleration information A (k) at moment and angular velocity data Ω (k), export the length of paces.
Selection excitation function is hyperbolic tangent function,
Tanh (x)=(ex-e-x)/(ex+e-x)。
In this example, neutral net includes an input layer, an output layer and a hidden layer, the nodal point number of hidden layer
Mesh is 10.
B. obtain initial data and pre-processed.By user carry mobile device in 3-axis acceleration sensor with
Three axis angular rate sensors are obtained in real time, respectively comprising the scalar size in three dimensions, are recorded as clock signal sequence ax
(t)、ay(t)、az(t)、ωx(t)、ωy(t)、ωzT (), carries out respectively real-time LPF, single order used in the present embodiment
LPF, such as x (t)=0.7x (t)+0.3x (t-1), the data obtained after filtering respectively are a 'x(t)、a′y(t)、a′z
(t)、ω′x(t)、ω′y(t)、ω′z(t),
A (k)={ a 'x(k),a′y(k),a′z(k) },
Ω (k)={ ω 'x(k),ω′y(k),ω′z(k)}。
C. the training of step-size estimation model.The pretreated acceleration of Jing step 2 under a large amount of various states is collected respectively
With angular velocity data, using the time series and its paces state of each data corresponding to each section of complete paces as one group of instruction
Practice data, the step-size estimation model based on bp neutral nets is trained.
Specifically, every time one group of data being selected at random from training data and carrying out training pattern, initialization neutral net is respectively weighed
Value parameter, calculates neural network model and the data is predicted the outcome by propagated forward, is compared with legitimate reading, so
Each neuron node is fed back to backward, the weighting parameter between a node is updated, constantly repeats this process, make result tend to receiving
Hold back.Finally, the initial value of each parameter of model is determined by cross validation, initial step recognition model f () is obtained.
2) the tuning on-line stage, comprise the following steps
A. obtain initial data and pre-processed.By pedestrian carry mobile device in 3-axis acceleration sensor with
Three axis angular rate sensors are obtained in real time, respectively comprising the scalar size in three dimensions, are recorded as clock signal sequence ax
(t)、ay(t)、az(t)、ωx(t)、ωy(t)、ωzT (), carries out respectively real-time LPF, obtain filtered data and be
a′x(t)、a′y(t)、a′z(t)、ω′x(t)、ω′y(t)、ω′z(t),
B. paces detection.The acceleration and angular speed for updating actual time window is respectively
Wa(k)={ Ag(k-k0),Ag(k-k0+1),…,Ag(k)}
Wω(k)={ ω (k-k0),ω(k-k0+1),…,ω(k)}
k0For default time window constant.It is new relatively to detect whether generation with upper lower threshold value by acceleration value
Paces.If paces are detected successfully, into step c, otherwise continue to detect.
C. paces length estimate.The nerve net that acceleration in actual time window and angular velocity data input are trained
Network model f (),
Y (k)=f (A (k-k0),…,A(k),Ω(k-k0),…,Ω(k))
I.e. exportable current paces length.
D. direction estimation.Using the magnetometer measures absolute force in equipment, estimate the direction of motion, calculate the direction of motion with
The angle α in north.
E. position.Calculate relative displacement,
Δx(k)=l (k) cos (α)
Δy(k)=l (k) sin (α)
Finally, according to the position at upper a moment, the EKF used in position fixing process calculates current location.Specifically
, if the position at a upper moment is (x (t-1), y (t-1)), then the position at current time is：
X (k)=(1-Kg)[x(k-1)+Δx(k)]+KgZy(k|k-1)
Y (k)=(1-Kg)[y(k-1)+Δy(k)]+KgZy(k|k-1)
Wherein, KgFor the kalman gain for calculating, ZxAnd ZyIt is this when estimated according to upper moment Position And Velocity
The position at quarter.
Table 1 shows the error tested during one group of straight line moving.
Table 1
Claims (1)
1. a kind of indoor inertial navigation method improved based on bp neutral nets, including off-line training step and tuning on-line rank
Section：
1) off-line training step, comprises the following steps
A. the step-size estimation model of a bp neutral net is set up；The neutral net number of plies and neuron number are set, one is built
With 2k0Individual input and the neural network model of single output, are input into as k before current time k0The acceleration at individual moment and angle
Speed data, is output as the length of paces；
B. obtain initial data and pre-processed：User carry mobile device move indoors, by mobile device in three axles
Acceleration transducer is obtained in real time with three axis angular rate sensors, respectively comprising the scalar size in three dimensions, when being recorded as
Sequential signal sequence, and carry out LPF pretreatment；
C. the training of step recognition model：The pretreated acceleration of Jing steps b and angle speed under a large amount of various states are collected respectively
Degrees of data, using each data corresponding to each section of complete paces as one group of training data, to the step based on bp neutral nets
It is long to estimate that model is trained, determine the initial value of each parameter of model, obtain initial step recognition model；
2) the tuning on-line stage, comprise the following steps
A. obtain initial data and pre-processed：3-axis acceleration sensor and three axles in the mobile device carried by user
Angular-rate sensor is obtained in real time, respectively comprising the scalar size in three dimensions, is recorded as clock signal sequence, and is carried out low
Pass filter is pre-processed；
B. paces detection；Time window is set, the acceleration and angular speed of actual time window is updated；By acceleration value with
The comparison of the upper lower threshold value paces new to detect whether generation；If paces are detected successfully, into step c, otherwise continue to examine
Survey；
C. paces length estimate；The neutral net mould that acceleration in actual time window and angular velocity data input are trained
Type, exports current paces length；
D. direction estimation；Using the magnetometer measures absolute force in equipment, the direction of motion is estimated, calculate the direction of motion with north
Angle；
E. position；Coordinate offset is calculated by the displacement of current time k, finally according to the position at upper a moment, is made in position fixing process
Current location is calculated with expanded Kalman filtration algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611020582.5A CN106643715A (en) | 2016-11-17 | 2016-11-17 | Indoor inertial navigation method based on bp neural network improvement |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611020582.5A CN106643715A (en) | 2016-11-17 | 2016-11-17 | Indoor inertial navigation method based on bp neural network improvement |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106643715A true CN106643715A (en) | 2017-05-10 |
Family
ID=58807702
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611020582.5A Pending CN106643715A (en) | 2016-11-17 | 2016-11-17 | Indoor inertial navigation method based on bp neural network improvement |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106643715A (en) |
Cited By (14)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107907127A (en) * | 2017-09-30 | 2018-04-13 | 天津大学 | A kind of step-size estimation method based on deep learning |
| CN108362289A (en) * | 2018-02-08 | 2018-08-03 | 浙江大学城市学院 | A kind of mobile intelligent terminal PDR localization methods based on Multi-sensor Fusion |
| CN108592908A (en) * | 2018-04-28 | 2018-09-28 | 山东交通学院 | One kind is ridden carrier posture safety monitoring method and device |
| CN109405832A (en) * | 2018-10-18 | 2019-03-01 | 南京理工大学 | A kind of target step estimation method |
| CN109579853A (en) * | 2019-01-24 | 2019-04-05 | 燕山大学 | Inertial navigation indoor orientation method based on BP neural network |
| CN109781094A (en) * | 2018-12-24 | 2019-05-21 | 上海交通大学 | Earth magnetism positioning system based on Recognition with Recurrent Neural Network |
| WO2019138225A1 (en) * | 2018-01-10 | 2019-07-18 | Oxford University Innovation Limited | Determining the location of a mobile device |
| CN110375741A (en) * | 2019-07-09 | 2019-10-25 | 中移（杭州）信息技术有限公司 | Pedestrian's dead reckoning method and terminal |
| CN111148058A (en) * | 2019-12-31 | 2020-05-12 | 武汉工程大学 | Method and system for positioning moving target in indoor environment and mobile robot |
| CN111207739A (en) * | 2018-11-22 | 2020-05-29 | 千寻位置网络有限公司 | Pedestrian walking zero-speed detection method and device based on GRU neural network |
| CN112313478A (en) * | 2018-06-21 | 2021-02-02 | 西斯纳维 | Analysis of the pace of a pedestrian walking |
| CN113686335A (en) * | 2021-06-10 | 2021-11-23 | 上海奥欧智能科技有限公司 | Method for performing accurate indoor positioning through IMU data by one-dimensional convolutional neural network |
| CN113891243A (en) * | 2021-11-10 | 2022-01-04 | 中国电信股份有限公司 | Terminal indoor positioning method, device and system and storage medium |
| CN113686335B (en) * | 2021-06-10 | 2024-05-24 | 上海奥欧智能科技有限公司 | Method for carrying out accurate indoor positioning by using IMU data through one-dimensional convolutional neural network |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102413564A (en) * | 2011-11-25 | 2012-04-11 | 北京工业大学 | Indoor positioning method based on BP neural network and improved centroid algorithm |
| CN104197938A (en) * | 2014-09-03 | 2014-12-10 | 厦门大学 | Indoor localization method based on foot inertia sensor |
| CN104748735A (en) * | 2013-12-25 | 2015-07-01 | 北京神州泰岳软件股份有限公司 | Intelligent terminal-based indoor positioning method and equipment thereof |
-
2016
- 2016-11-17 CN CN201611020582.5A patent/CN106643715A/en active Pending
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102413564A (en) * | 2011-11-25 | 2012-04-11 | 北京工业大学 | Indoor positioning method based on BP neural network and improved centroid algorithm |
| CN104748735A (en) * | 2013-12-25 | 2015-07-01 | 北京神州泰岳软件股份有限公司 | Intelligent terminal-based indoor positioning method and equipment thereof |
| CN104197938A (en) * | 2014-09-03 | 2014-12-10 | 厦门大学 | Indoor localization method based on foot inertia sensor |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107907127A (en) * | 2017-09-30 | 2018-04-13 | 天津大学 | A kind of step-size estimation method based on deep learning |
| WO2019138225A1 (en) * | 2018-01-10 | 2019-07-18 | Oxford University Innovation Limited | Determining the location of a mobile device |
| US11788843B2 (en) | 2018-01-10 | 2023-10-17 | Oxford University Innovation Limited | Determining the location of a mobile device |
| CN108362289A (en) * | 2018-02-08 | 2018-08-03 | 浙江大学城市学院 | A kind of mobile intelligent terminal PDR localization methods based on Multi-sensor Fusion |
| CN108362289B (en) * | 2018-02-08 | 2020-12-08 | 浙江大学城市学院 | Mobile intelligent terminal PDR positioning method based on multi-sensor fusion |
| CN108592908A (en) * | 2018-04-28 | 2018-09-28 | 山东交通学院 | One kind is ridden carrier posture safety monitoring method and device |
| CN108592908B (en) * | 2018-04-28 | 2023-06-02 | 山东交通学院 | Riding carrier posture safety monitoring method and device |
| CN112313478A (en) * | 2018-06-21 | 2021-02-02 | 西斯纳维 | Analysis of the pace of a pedestrian walking |
| CN109405832A (en) * | 2018-10-18 | 2019-03-01 | 南京理工大学 | A kind of target step estimation method |
| CN109405832B (en) * | 2018-10-18 | 2020-06-09 | 南京理工大学 | Target step length estimation method |
| CN111207739A (en) * | 2018-11-22 | 2020-05-29 | 千寻位置网络有限公司 | Pedestrian walking zero-speed detection method and device based on GRU neural network |
| CN109781094A (en) * | 2018-12-24 | 2019-05-21 | 上海交通大学 | Earth magnetism positioning system based on Recognition with Recurrent Neural Network |
| CN109579853A (en) * | 2019-01-24 | 2019-04-05 | 燕山大学 | Inertial navigation indoor orientation method based on BP neural network |
| CN109579853B (en) * | 2019-01-24 | 2021-02-26 | 燕山大学 | Inertial navigation indoor positioning method based on BP neural network |
| CN110375741A (en) * | 2019-07-09 | 2019-10-25 | 中移（杭州）信息技术有限公司 | Pedestrian's dead reckoning method and terminal |
| CN111148058A (en) * | 2019-12-31 | 2020-05-12 | 武汉工程大学 | Method and system for positioning moving target in indoor environment and mobile robot |
| CN113686335A (en) * | 2021-06-10 | 2021-11-23 | 上海奥欧智能科技有限公司 | Method for performing accurate indoor positioning through IMU data by one-dimensional convolutional neural network |
| CN113686335B (en) * | 2021-06-10 | 2024-05-24 | 上海奥欧智能科技有限公司 | Method for carrying out accurate indoor positioning by using IMU data through one-dimensional convolutional neural network |
| CN113891243A (en) * | 2021-11-10 | 2022-01-04 | 中国电信股份有限公司 | Terminal indoor positioning method, device and system and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106643715A (en) | Indoor inertial navigation method based on bp neural network improvement | |
| CN108362282B (en) | Inertial pedestrian positioning method based on self-adaptive zero-speed interval adjustment | |
| CN110118560B (en) | Indoor positioning method based on LSTM and multi-sensor fusion | |
| CN109991636A (en) | Map constructing method and system based on GPS, IMU and binocular vision | |
| CN108731670A (en) | Inertia/visual odometry combined navigation locating method based on measurement model optimization | |
| CN107478223A (en) | A kind of human body attitude calculation method based on quaternary number and Kalman filtering | |
| CN110146909A (en) | A kind of location data processing method | |
| CN106780699A (en) | A kind of vision SLAM methods aided in based on SINS/GPS and odometer | |
| CN105953796A (en) | Stable motion tracking method and stable motion tracking device based on integration of simple camera and IMU (inertial measurement unit) of smart cellphone | |
| CN108387236B (en) | Polarized light SLAM method based on extended Kalman filtering | |
| CN108318038A (en) | A kind of quaternary number Gaussian particle filtering pose of mobile robot calculation method | |
| CN107255474B (en) | PDR course angle determination method integrating electronic compass and gyroscope | |
| CN103822633A (en) | Low-cost attitude estimation method based on second-order measurement update | |
| CN104613965B (en) | A kind of step-by-step movement pedestrian navigation method based on bidirectional filtering smoothing technique | |
| CN108132053B (en) | Pedestrian track construction method and system and inertia measurement device | |
| CN103940442A (en) | Location method and device adopting accelerating convergence algorithm | |
| CN113252033A (en) | Positioning method, positioning system and robot based on multi-sensor fusion | |
| CN109459028A (en) | A kind of adaptive step estimation method based on gradient decline | |
| CN110375731A (en) | A kind of mixing interacting multiple model filters method | |
| CN107677267A (en) | Indoor pedestrian navigation course feedback modifiers method based on MEMS IMU | |
| CN107796387A (en) | Localization method, positioner and electronic equipment | |
| CN103644910A (en) | Personal autonomous navigation system positioning method based on segment RTS smoothing algorithm | |
| CN110440797A (en) | Vehicle attitude estimation method and system | |
| CN109858137A (en) | It is a kind of based on the complicated maneuvering-vehicle track estimation method that can learn Extended Kalman filter | |
| CN106643713A (en) | Zero-velocity correct walker trajectory estimation method and device for smooth and self-adaptive threshold value adjustment |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20170510 |
|
| WD01 | Invention patent application deemed withdrawn after publication |