CN104748735B - Indoor orientation method and equipment based on intelligent terminal - Google Patents
Indoor orientation method and equipment based on intelligent terminal Download PDFInfo
- Publication number
- CN104748735B CN104748735B CN201310727255.3A CN201310727255A CN104748735B CN 104748735 B CN104748735 B CN 104748735B CN 201310727255 A CN201310727255 A CN 201310727255A CN 104748735 B CN104748735 B CN 104748735B
- Authority
- CN
- China
- Prior art keywords
- value
- paces
- indoor positioning
- neuron
- walking
- 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
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)
Abstract
The invention discloses a kind of indoor orientation method based on intelligent terminal and equipment, it is related to mobile communication technology field.Indoor positioning neutral net is trained using indoor traveling data sample；When indoor positioning need to be carried out：Judge whether that row is further according to the acceleration information obtained from the positioner of intelligent terminal from initial position；When completing a step when judging to advance, this walking is obtained through indoor positioning neural computing according to acceleration information in this step and paces frequency and enters step value and deviation angle；Enter step value according to described walking to position current location with the deviation angle.Because the embodiment of the present invention uses intelligent Neural Network algorithm, the average reference step-length that motion unit difference will be embodied is combined with using sensor acceleration analysis data, it is achieved thereby that more science accurately paces displacement calculating in real time, then can realize accurate indoor positioning and course more new solution by the positioner of intelligent terminal.
Description
Technical field
The present invention relates to mobile communication technology field, more particularly to a kind of indoor orientation method based on intelligent terminal and set
It is standby.
Background technology
At present, for the place that cannot effectively receive gps signal, such as indoor environment, the interior based on intelligent terminal is realized
Location solution is also little., nearly all there is positioning precision not enough in only several solutions, the degree of accuracy is limited, positioning
Coarse problem.In the urgent need to a kind of new solution, using the teaching of the invention it is possible to provide it is a kind of it is new based on intelligent terminal realize it is accurate
Indoor positioning solution.
The content of the invention
In view of the above problems, the embodiment of the present invention provides a kind of indoor orientation method and equipment based on intelligent terminal, energy
Enough rational adjustment algorithm is carried out by the positioner acquisition data of intelligent terminal realize accurate indoor positioning and course more
New solution.
The embodiment of the present invention employs following technical scheme：
One embodiment of the invention provides a kind of indoor orientation method based on intelligent terminal, and methods described includes：
Indoor positioning neutral net is trained using indoor traveling data sample；
When needing to carry out indoor positioning：
From initial position, judge whether that row is further according to the acceleration information obtained from the positioner of intelligent terminal；
When completing a step when judging to advance, according to acceleration information in this step and paces frequency through indoor positioning neutral net
It is calculated this walking and enters step value；Also, obtain this walking from the positioner of intelligent terminal and enter rear course with respect to this walking
Deviation angle before entering；
Enter step value according to described walking to position current location with the deviation angle；
Using current location as initial position, the acceleration that the basis is obtained from the positioner of intelligent terminal is continued executing with
Degrees of data judges whether that row is further, terminates to this positioning.
The indoor traveling data sample includes paces frequency, direct of travel acceleration variance, vertically travels direction acceleration
Degree variance and reference step value；
It is described using indoor traveling data sample indoor positioning neutral net is trained including：
For further data are often gone in multiple samples as a training data, perform respectively：By this training number
Paces frequency, direct of travel acceleration variance in, vertically travel directional acceleration variance and reference step value respectively as
The value of the input neuron of indoor positioning neutral net；Using the actual step size value in this training data as indoor positioning nerve
The value of the output neuron of network；
The value of neuron and the value of output neuron are respectively input into according to the indoor positioning neutral net, calculate each defeated respectively
Enter the connection weight between each hidden layer neuron in neuron and indoor positioning neutral net, and each hidden layer neuron with
Connection weight between output neuron；
The connection weight is modified according to multiple training data.
It is described this walking is obtained through indoor positioning neural computing according to acceleration information in this step and paces frequency to enter
Step value includes：
Paces frequency, direct of travel acceleration variance during this is walked, vertically travel directional acceleration variance and reference step
It is worth the value as each input neuron of the indoor positioning neutral net；
According to it is each input neuron and each hidden layer neuron between connection weight, and each hidden layer neuron with it is defeated
Go out connection weight between neuron, calculate the value of output neuron, step value is entered as this walking.
The acceleration information that the basis is obtained from the positioner of intelligent terminal judges whether that row is further included：
Judge whether the acceleration information obtained from the positioner of intelligent terminal meets paces Rule of judgment；
The paces Rule of judgment is：Vertically travelling the acceleration information in direction includes that one is negative oblique more than the first preset value
Rate, and direct of travel acceleration information includes a negative slope for being more than the second preset value；
If meeting the paces Rule of judgment, it is determined that advance and complete a step.
It is described this walking is obtained through indoor positioning neural computing according to acceleration information in this step and paces frequency to enter
Also include after step value：
Judge that whether described walking enters step value less than the 3rd preset value；
If being more than the 3rd preset value, perform and described step value is entered with the deviation angle to current according to described walking
Position is positioned；
If being less than the 3rd preset value, re-execute and sentenced according to the acceleration information obtained from the positioner of intelligent terminal
It is disconnected whether to go further.
In addition, the embodiment of the present invention additionally provides a kind of indoor positioning device based on intelligent terminal, the equipment includes：
Training module, for being trained to indoor positioning neutral net using indoor traveling data sample；
Indoor positioning module, for when needing to carry out indoor positioning, carrying out indoor positioning；
The indoor positioning module includes：
Paces judging unit, for from initial position, according to the acceleration information obtained from the positioner of intelligent terminal
Judge whether that row is further；
Acquiring unit, for when the paces judging unit judge to advance complete a step when, according to accelerating the number of degrees in this step
Enter step value according to this walking is obtained through indoor positioning neural computing with paces frequency；Also, filled from the positioning of intelligent terminal
Put and obtain the deviation angle that this walking is entered before rear course is entered with respect to this walking；
Positioning unit, the described walking for being obtained according to acquiring unit enters step value with the deviation angle to current
Position is positioned；
Cycling element, it is fixed to this for using current location as initial position, continuing to start the paces judging unit
Terminate position.
The indoor traveling data sample includes paces frequency, direct of travel acceleration variance, vertically travels direction acceleration
Degree variance and reference step value；
The training module includes：
Assignment unit, for that for further data are often gone in multiple samples as a training data, will instruct every time
Practice the paces frequency in data, direct of travel acceleration variance, vertically travel directional acceleration variance and reference step value difference
As the value of the input neuron of indoor positioning neutral net；Using the actual step size value in each training data as indoor positioning
The value of the output neuron of neutral net；
Connection weight computing unit, value and output god for being respectively input into neuron according to the indoor positioning neutral net
Through the value of unit, the connection weight between each hidden layer neuron in each input neuron and indoor positioning neutral net is calculated respectively
Value, and connection weight between each hidden layer neuron and output neuron；
Amending unit, is modified for the connection weight according to multiple training data.
The acquiring unit includes step-length acquiring unit and deviation angle acquiring unit：
The step-length acquiring unit includes,
Input determination subelement, for when the paces judging unit judge advance complete a step when, paces during this is walked
Frequency, direct of travel acceleration variance, vertically travel directional acceleration variance and reference step value as the indoor positioning god
Through the value of each input neuron of network；
Step size computation subelement, for according to it is each input neuron and each hidden layer neuron between connection weight, with
And connection weight between each hidden layer neuron and output neuron, the value of output neuron is calculated, enter step-length as this walking
Value；
The deviation angle acquiring unit, for when the paces judging unit judges to advance one step of completion, from intelligence
The positioner of terminal obtains the deviation angle that this walking is entered before rear course is entered with respect to this walking.
The paces judging unit includes：
Whether condition judgment subelement, the acceleration information that the positioner for judging from intelligent terminal is obtained meets step
Cut down Rule of judgment；
The paces Rule of judgment is：Vertically travelling the acceleration information in direction includes that one is negative oblique more than the first preset value
Rate, and direct of travel acceleration information includes a negative slope for being more than the second preset value；
Determination subelement, if being to meet the paces Rule of judgment for the judged result of the condition judgment subelement,
Then determine to advance to complete a step.
The positioning unit also includes：
Location determination subelement, for according to acceleration information in this step and paces frequency through indoor positioning neutral net
It is calculated this walking to enter after step value, judges that whether described walking enters step value less than the 3rd preset value；
Continue locator unit, for being more than the 3rd preset value, then when the judged result of the location determination subelement
Enter step value according to described walking to position current location with the deviation angle；
Again paces judgment sub-unit, for being preset less than the 3rd when the judged result of the location determination subelement
Value, then restart the paces judging unit.
It can be seen that, the embodiment of the present invention provides a kind of indoor orientation method and equipment based on intelligent terminal, using indoor row
Enter data sample to be trained indoor positioning neutral net；When needing to carry out indoor positioning：From initial position, according to from intelligence
The acceleration information that the positioner of energy terminal is obtained judges whether that row is further；When completing a step when judging to advance, according to this
Acceleration information and paces frequency obtain this walking and enter step value through indoor positioning neural computing in step；Also, from intelligence
The positioner of terminal obtains the deviation angle that this walking is entered before rear course is entered with respect to this walking；Step-length is entered according to described walking
Value is positioned with the deviation angle to current location；Using current location as initial position, continue executing with the basis from
The acceleration information that the positioner of intelligent terminal is obtained judges whether that row is further, terminates to this positioning.Due to the present invention
Embodiment use intelligent Neural Network algorithm, and will embody motion unit difference average reference step-length with use sensor add
Speed measurement data is combined, it is achieved thereby that more science accurately paces displacement calculating in real time, then can be by intelligence
The positioner of terminal is combined realizes accurate indoor positioning and course more new solution to the accurate adjustment algorithm of step-length.
Brief description of the drawings
Fig. 1 is a kind of indoor orientation method flow chart based on intelligent terminal provided in an embodiment of the present invention；
Fig. 2 is provided in an embodiment of the present invention a kind of using intelligent terminal direct of travel schematic diagram；
Fig. 3 is a kind of indoor positioning neutral net internal arithmetic schematic diagram provided in an embodiment of the present invention；
Fig. 4 is a kind of indoor positioning device structured flowchart based on intelligent terminal provided in an embodiment of the present invention.
Specific embodiment
To make the object, technical solutions and advantages of the present invention clearer, below in conjunction with accompanying drawing to embodiment party of the present invention
Formula is described in further detail.
Referring to Fig. 1, the embodiment of the present invention provides a kind of indoor orientation method based on intelligent terminal, specifically includes：
S101：Indoor positioning neutral net is trained using indoor traveling data sample.
It should be noted that indoor traveling data sample is multiple motion units（That is multidigit actual test personnel）By hand
Hold terminal device（Intelligent terminal）The data sample that assignment test process is obtained is carried out indoors.
It should be noted that in the embodiment of the present invention, advanced indoors by hand-held intelligent terminal, by intelligent terminal
Positioner（Such as gyroscope etc.）Location information is obtained, with reference to actual test data, data sample is obtained jointly.
In the embodiment of the present invention, intelligent terminal refers to that can obtain location information and hand-holdable equipment, as long as can obtain
Take location information and can move all in the range of intelligent terminal described in the embodiment of the present invention, such as can be smart mobile phone,
PAD etc..
Preferably, the indoor traveling data sample includes paces frequency, direct of travel acceleration variance, the side of vertically travelling
To acceleration variance and reference step value.
The present embodiments relate to application Portable intelligent terminal（That is mobile terminal）In MEMS（Micro-Electro-
Mechanical Systems MEMSs）In positioner（Such as position sensor）Carry out indoor location positioning.
The place of gps signal, such as indoor environment cannot be effectively received, can be using compass, gravitometer, the top in MEMS
The sensors such as spiral shell instrument, accelerometer are positioned.Common position sensing location technology uses magnetometer, gyro in intelligent terminal
Instrument differentiates course（Obtain relative preceding deviation angle of advancing after advancing）, carry out speed with acceleration transducer and displacement judge.
It is simple to carry out speed and during displacement judges using the positioner in intelligent terminal, usual only consideration and speed measurement number in itself
According to carrying out speed calculating and offset estimation, thus, the embodiment of the present invention is additionally contemplates that the difference of motion unit simultaneously：Such as individual
Height, body weight, fitness can bring the difference of motion unit average reference step-length, i.e., by reference to step value（Each motion
The average reference step value of body, can actually measure in practice and obtain）.With the further accuracy for improving positioning.
It is further to note that referring to Fig. 2, direct of travel vertically travels direction such as Fig. 2 as shown in y-axis direction in Fig. 2
Shown in middle z-axis direction.Direct of travel is y-axis direction, before producing in the traveling（Backward acceleration）；X-axis direction becomes in position
Level left rightward acceleration can be produced during change；Z-axis direction can produce the acceleration of vertical direction in traveling.
Preferably, it is described using indoor traveling data sample indoor positioning neutral net is trained including：
For further data are often gone in multiple samples as a training data, perform respectively：By this training number
Paces frequency, direct of travel acceleration variance in, vertically travel directional acceleration variance and reference step value respectively as
The value of the input neuron of indoor positioning neutral net；Using the actual step size value in this training data as indoor positioning nerve
The value of the output neuron of network.
The value of neuron and the value of output neuron are respectively input into according to the indoor positioning neutral net, calculate each defeated respectively
Enter the connection weight between each hidden layer neuron in neuron and indoor positioning neutral net, and each hidden layer neuron with
Connection weight between output neuron.
Then, above-mentioned connection weight is modified according to multiple training data.
Specifically, as shown in figure 3, indoor positioning neutral net input neuron be 4, respectively correspond to cut down frequency, advance
Directional acceleration variance, vertically travel directional acceleration variance and reference step value.Hidden layer neuron number is also 4, defeated
Go out neuron, it is step value.
Wherein, the algorithm of paces frequency is：
fstep(the t of (n)=1/n-tn-1)
tk-tk-1It it is a step holding time cycle, that is to say, that often go further data as a training data, tk-
tk-1It is the time cycle of each training data.
The computing formula of direct of travel acceleration variance is：
The computing formula for vertically travelling directional acceleration variance is：
A kind of preferred embodiment, tk-1To tkIn one step in the time cycle,（The such as positioner of I9023 embedded in mobile phone）
The number of samples of accelerometer output may at least only have two.In practice, it may be considered that to obtain traveling acceleration and hang down
Keep straight on and carry out appropriate interpolation processing into directional acceleration, further to improve accuracy.Linear compensation algorithm is considered first（If
Put the axially different acceleration bound threshold values of Y and Z, it may be considered that be axially different | a |max）.
The input of each neuron of hidden layer is as follows：
Indoor positioning neutral net in the embodiment of the present invention specifically uses Sigmoid excitation functions：
That is, each neuronal layers output is implied to be specially：
Network output function：
Output Shaping equation：
Wherein ldownIt is step-length lower limit（Preferably take ldown=0.3m, can adjust）, lupIt is the step-length upper limit（Preferably take
lup=1.5m, can adjust）.
In addition, in the embodiment of the present invention, further, also including to each connection weight amendment the step of.Describe in detail such as
Under：
By the connection weight and output layer and the connection of hidden layer neuron of hidden layer neuron and input layer
Weights composition of vector W=[vki,θi,wij], or
Backward learning process：
Learning sample is provided with for { in1(n),in2(n),in3(n),x4(n);lstep_ref(n) }, n=1,2 ..., P；P is
Learning sample sum.For certain sample { in1(n),in2(n),in3(n),x4(n);lstep_ref(n) }, in given network connection power
After value vector W, the output valve l of network can be calculatedstepN (), the error for defining output is d (n)=lstep_ref(n)-lstep
(n), and define the error function of sample p and be：
Wherein, lstep_refN () is n-th sample reference value, lstepN () judges output valve for n-th step-length.It is implicit
Layer neuron i and input layer xjConnection weight, θiIt is the threshold values of hidden layer neuron i, hidden layer neuron number is M, this calculation
M=4 in method.Output layer neuron number is L, L=1 in this algorithm.
By adjusting the value of W, error d (n) is gradually reduced, to improve the computational accuracy of network.It is propagated through in opposite direction
Cheng Zhong, is that W is modified with the negative gradient direction that W changes along error function e (n).Takeα in formula
It is learning rate,（0~1）Between value.
Wherein,When sample is n, the element in Δ W is：
Preferably, fed back by difference referring to Fig. 3 and be adjusted, in each step length data, judge neural network algorithm base
This parameter is as follows：
M=4, α=0.01, γ=0.02, ρ=0.993, numite=200
Wherein it is hidden layer neuron number, α is learning rate, and γ is factor of momentum, and ρ is the Learning Process Control factor（
Reduce learning rate in learning process）, numiteIt is iterations.
Weights learning adjustment algorithm of k-th hidden layer to output layer：
J-th input layer is calculated to i-th hidden layer weights learning adjustment
Method:
Hidden layer threshold values learns adjustment algorithm：
Need to carry out the collection of various sample datas, to software in each layer weights and threshold parameter optimize adjustment,
A set of parameter for Practical Project test is obtained, for accurately judging paces and single step displacement.Reference step in algorithm
Vary with each individual, be the personal parameter of Test Engineer's input when test starts every time, to improve the accuracy of step-length judgement.
When needing to carry out indoor positioning, following operation is performed：
S102：From initial position, judge whether to advance according to the acceleration information obtained from the positioner of intelligent terminal
One step.
S103：When completing a step when judging to advance, according to acceleration information in this step and paces frequency through indoor positioning god
This walking is obtained through network calculations enter step value；Also, obtaining this walking from the positioner of intelligent terminal, to enter rear course relative
This walking enter before deviation angle.
Wherein, this walking is obtained through indoor positioning neural computing according to acceleration information in this step and paces frequency to enter
Step value includes：
Paces frequency, direct of travel acceleration variance during this is walked, vertically travel directional acceleration variance and reference step
It is worth the value as each input neuron of the indoor positioning neutral net；
According to it is each input neuron and each hidden layer neuron between connection weight, and each hidden layer neuron with it is defeated
Go out connection weight between neuron, calculate the value of output neuron, step value is entered as this walking.
Preferably, judge whether that row is further included according to the acceleration information obtained from the positioner of intelligent terminal：
Judge whether the acceleration information obtained from the positioner of intelligent terminal meets paces Rule of judgment；
The paces Rule of judgment is：Vertically travelling the acceleration information in direction includes that one is negative oblique more than the first preset value
Rate, and direct of travel acceleration information includes a negative slope for being more than the second preset value；
If meeting the paces Rule of judgment, it is determined that advance and complete a step.
It should be noted that preferred, the first preset value is 1, and the acceleration information for vertically travelling direction is more than including one
The negative slope of the first preset value, illustrates to vertically travel the behavior of taking a step that above-below direction of tester is collected on direction；It is excellent
Choosing, the second preset value is 0.2, because under normal circumstances, between [- 0.1 ,+0.1], the overwhelming majority falls acceleration straw rope
Between [- 0.08 ,+0.08], here, the second preset value is preferably chosen for 0.2, is in order to avoid due to being judged by accident caused by noise
It is disconnected, direct of travel acceleration information include one more than the second preset value negative slope, illustrate to collect direct of travel front and rear once
Acceleration and deceleration behavior.
In practical operation, the embodiment of the present invention can also be in the acceleration information obtained from the positioner of intelligent terminal
Afterwards, effective filter operation of noise is carried out to acceleration information, further to improve the degree of accuracy.
As preferred, this is obtained through indoor positioning neural computing according to acceleration information in this step and paces frequency
Walking also includes after entering step value：
Judge that whether described walking enters step value less than the 3rd preset value；
If being more than the 3rd preset value, perform and described step value is entered with the deviation angle to current according to described walking
Position is positioned；
If being less than the 3rd preset value, re-execute and sentenced according to the acceleration information obtained from the positioner of intelligent terminal
It is disconnected whether to go further.
3rd preset value can be chosen to be 0.3 meter, that is to say, that if the current step-length for judging is actually smaller than 0.3 meter, recognize
It is, because noise causes, to abandon this result, positioning is not done and is used.
Preferably, after noise filtering operation being carried out to the acceleration information that is obtained from the positioner of intelligent terminal,
Re-execute and row further step is judged whether according to the acceleration information obtained from the positioner of intelligent terminal.
S104：Enter step value according to described walking to position current location with the deviation angle.
In actual applications, the embodiment of the present invention can also be implemented to render according to positioning result is often walked, and constantly update boat
To realization is based on indoor positioning.
S105：Using current location as initial position, continue executing with the basis and obtained from the positioner of intelligent terminal
Acceleration information judge whether that row is further, terminate to this positioning.
It can be seen that, the embodiment of the present invention provides a kind of indoor orientation method based on intelligent terminal, due to the embodiment of the present invention
Using intelligent Neural Network algorithm, and the average reference step-length of motion unit difference will be embodied and the survey of sensor acceleration will be used
Amount data are combined, it is achieved thereby that more science accurately paces displacement calculating in real time, then can be by intelligent terminal
Positioner is combined realizes accurate indoor positioning and course more new solution to the accurate adjustment algorithm of step-length.
It should be noted that the indoor orientation method based on intelligent terminal provided in an embodiment of the present invention, can apply
In plurality of application scenes, the various applications of field of locating technology can be greatly expanded, further promote correlation technique development, such as
Indoor orientation method provided in an embodiment of the present invention, in can apply to indoor measurement scene, it is also possible to be applied to location-based service
In scene, can also be applied in wireless network test scene, etc..
Referring to Fig. 4, the embodiment of the present invention also provides a kind of indoor positioning device based on intelligent terminal, the equipment bag
Include：Training module 400 and indoor positioning module 500.
Training module 400, for being trained to indoor positioning neutral net using indoor traveling data sample.
Indoor positioning module 500, for when needing to carry out indoor positioning, carrying out indoor positioning.
Specifically, the indoor positioning module 500 includes：
Paces judging unit 501, for from initial position, according to the acceleration number of degrees obtained from the positioner of intelligent terminal
It is judged that whether going further；
Acquiring unit 502, for when the paces judging unit judges to advance one step of completion, according to acceleration in this step
Data and paces frequency obtain this walking and enter step value through indoor positioning neural computing；Also, from the positioning of intelligent terminal
Device obtains the deviation angle that this walking is entered before rear course is entered with respect to this walking；
Positioning unit 503, the described walking for being obtained according to acquiring unit enters step value with the deviation angle pair
Current location is positioned；
Cycling element 504, for using current location as initial position, continuing to start the paces judging unit, to this
Secondary positioning terminates.
Preferably, the indoor traveling data sample includes paces frequency, direct of travel acceleration variance, the side of vertically travelling
To acceleration variance and reference step value.
Accordingly, the training module includes：
Assignment unit, for that for further data are often gone in multiple samples as a training data, will instruct every time
Practice the paces frequency in data, direct of travel acceleration variance, vertically travel directional acceleration variance and reference step value difference
As the value of the input neuron of indoor positioning neutral net；Using the actual step size value in each training data as indoor positioning
The value of the output neuron of neutral net.
Connection weight computing unit, value and output god for being respectively input into neuron according to the indoor positioning neutral net
Through the value of unit, the connection weight between each hidden layer neuron in each input neuron and indoor positioning neutral net is calculated respectively
Value, and connection weight between each hidden layer neuron and output neuron.
With amending unit is modified for the connection weight according to multiple training data.
The acquiring unit includes step-length acquiring unit and deviation angle acquiring unit：
Wherein, the step-length acquiring unit includes,
Input determination subelement, for when the paces judging unit judge advance complete a step when, paces during this is walked
Frequency, direct of travel acceleration variance, vertically travel directional acceleration variance and reference step value as the indoor positioning god
Through the value of each input neuron of network.
With, step size computation subelement, for according to the connection weight between each input neuron and each hidden layer neuron,
And connection weight between each hidden layer neuron and output neuron, the value of output neuron is calculated, as this walking progress
Long value.
The deviation angle acquiring unit, for when the paces judging unit judges to advance one step of completion, from intelligence
The positioner of terminal obtains the deviation angle that this walking is entered before rear course is entered with respect to this walking.
Preferably, the paces judging unit includes：
Whether condition judgment subelement, the acceleration information that the positioner for judging from intelligent terminal is obtained meets step
Cut down Rule of judgment；
The paces Rule of judgment is：Vertically travelling the acceleration information in direction includes that one is negative oblique more than the first preset value
Rate, and direct of travel acceleration information includes a negative slope for being more than the second preset value；
Determination subelement, if being to meet the paces Rule of judgment for the judged result of the condition judgment subelement,
Then determine to advance to complete a step.
Further, the positioning unit also includes：
Location determination subelement, for according to acceleration information in this step and paces frequency through indoor positioning neutral net
It is calculated this walking to enter after step value, judges that whether described walking enters step value less than the 3rd preset value；
Continue locator unit, for being more than the 3rd preset value, then when the judged result of the location determination subelement
Enter step value according to described walking to position current location with the deviation angle；
Again paces judgment sub-unit, for being preset less than the 3rd when the judged result of the location determination subelement
Value, then restart the paces judging unit.
It should be noted that the indoor positioning device based on intelligent terminal provided in an embodiment of the present invention, can be built-in
In intelligent terminal, or separately from intelligent terminal, can be communicated with intelligent terminal, be filled from the positioning of intelligent terminal
Put the information such as acquisition acceleration information.
It should be noted that the operation principle of modules or submodule in present system embodiment and treated
Journey may refer to the associated description in embodiment of the method shown in above-mentioned Fig. 1-Fig. 3, and here is omitted.
It can be seen that, the embodiment of the present invention provides a kind of indoor positioning device based on intelligent terminal, due to the embodiment of the present invention
Using intelligent Neural Network algorithm, and the average reference step-length of motion unit difference will be embodied and the survey of sensor acceleration will be used
Amount data are combined, it is achieved thereby that more science accurately paces displacement calculating in real time, then can be by intelligent terminal
Positioner is combined realizes accurate indoor positioning and course more new solution to the accurate adjustment algorithm of step-length.
For the ease of clearly describing the technical scheme of the embodiment of the present invention, in inventive embodiment, employ " first ",
Printed words such as " second " make a distinction to function and the essentially identical identical entry of effect or similar item, and those skilled in the art can manage
The printed words such as solution " first ", " second " are not defined to quantity and execution order.
It will appreciated by the skilled person that all or part of step in realizing above-described embodiment method can be
The hardware of correlation is instructed to complete by program, described program can be stored in a computer read/write memory medium,
The program upon execution, comprises the following steps：（The step of method）, described storage medium, such as：ROM/RAM, magnetic disc, CD
Deng.
Presently preferred embodiments of the present invention is the foregoing is only, is not intended to limit the scope of the present invention.It is all
Any modification, equivalent substitution and improvements made within the spirit and principles in the present invention etc., are all contained in protection scope of the present invention
It is interior.
Claims (8)
1. a kind of indoor orientation method based on intelligent terminal, it is characterised in that methods described includes：
Indoor positioning neutral net is trained using indoor traveling data sample；
When needing to carry out indoor positioning：
From initial position, judge whether that row is further according to the acceleration information obtained from the positioner of intelligent terminal；
When completing a step when judging to advance, according to acceleration information in this step and paces frequency through indoor positioning neural computing
Obtain this walking and enter step value；Also, from the positioner of intelligent terminal this walking is obtained to enter before rear course enters with respect to this walking
Deviation angle；
Enter step value according to described walking to position current location with the deviation angle；
Using current location as initial position, the acceleration number of degrees that the basis is obtained from the positioner of intelligent terminal are continued executing with
It is judged that whether going further, terminate to this positioning；
The acceleration information that the basis is obtained from the positioner of intelligent terminal judges whether that row is further included：
Judge whether the acceleration information obtained from the positioner of intelligent terminal meets paces Rule of judgment；
The paces Rule of judgment is：Vertically travelling the acceleration information in direction includes a negative slope for being more than the first preset value,
And direct of travel acceleration information includes a negative slope for being more than the second preset value；
If meeting the paces Rule of judgment, it is determined that advance and complete a step.
2. method according to claim 1, it is characterised in that the indoor traveling data sample includes paces frequency, OK
Enter directional acceleration variance, vertically travel directional acceleration variance and reference step value；
It is described using indoor traveling data sample indoor positioning neutral net is trained including：
For further data are often gone in multiple samples as a training data, perform respectively：By in this training data
Paces frequency, direct of travel acceleration variance, vertically travel directional acceleration variance and reference step value respectively as interior
The value of the input neuron of positioning neutral net；Using the actual step size value in this training data as indoor positioning neutral net
Output neuron value；
The value of neuron and the value of output neuron are respectively input into according to the indoor positioning neutral net, each input god is calculated respectively
Through the connection weight between each hidden layer neuron in unit and indoor positioning neutral net, and each hidden layer neuron and output
Connection weight between neuron；
The connection weight is modified according to multiple training data.
3. method according to claim 2, it is characterised in that described to be passed through according to acceleration information in this step and paces frequency
Indoor positioning neural computing obtains this walking and enters step value including：
Paces frequency, direct of travel acceleration variance during this is walked, vertically travel directional acceleration variance and reference step value is made
It is the value of each input neuron of the indoor positioning neutral net；
According to the connection weight between each input neuron and each hidden layer neuron, and each hidden layer neuron and output god
Through connection weight between unit, the value of output neuron is calculated, step value is entered as this walking.
4. method according to claim 1, it is characterised in that described to be passed through according to acceleration information in this step and paces frequency
Indoor positioning neural computing obtains also including after step value is entered in this walking：
Judge that whether described walking enters step value less than the 3rd preset value；
If being more than the 3rd preset value, perform and described step value is entered with the deviation angle to current location according to described walking
Positioned；
If being less than the 3rd preset value, re-execute is according to the acceleration information judgement obtained from the positioner of intelligent terminal
No row is further.
5. a kind of indoor positioning device based on intelligent terminal, it is characterised in that the equipment includes：
Training module, for being trained to indoor positioning neutral net using indoor traveling data sample；
Indoor positioning module, for when needing to carry out indoor positioning, carrying out indoor positioning；
The indoor positioning module includes：
Paces judging unit, for from initial position, being judged according to the acceleration information obtained from the positioner of intelligent terminal
Whether go further；
Acquiring unit, for when the paces judging unit judge advance complete a step when, according to acceleration information in this step with
Paces frequency obtains this walking and enters step value through indoor positioning neural computing；Also, obtained from the positioner of intelligent terminal
Take the deviation angle that this walking is entered before rear course is entered with respect to this walking；
Positioning unit, the described walking for being obtained according to acquiring unit enters step value with the deviation angle to current location
Positioned；
Cycling element, for using current location as initial position, continuing to start the paces judging unit, to this positioning knot
Beam；
The paces judging unit includes：
Condition judgment subelement, whether the acceleration information that the positioner for judging from intelligent terminal is obtained meets paces is sentenced
Broken strip part；
The paces Rule of judgment is：Vertically travelling the acceleration information in direction includes a negative slope for being more than the first preset value,
And direct of travel acceleration information includes a negative slope for being more than the second preset value；
Determination subelement, if being to meet the paces Rule of judgment for the judged result of the condition judgment subelement, really
Fixed traveling completes a step.
6. equipment according to claim 5, it is characterised in that the indoor traveling data sample includes paces frequency, OK
Enter directional acceleration variance, vertically travel directional acceleration variance and reference step value；
The training module includes：
Assignment unit, for for further data are often gone in multiple samples as a training data, will every time train number
Paces frequency, direct of travel acceleration variance in, vertically travel directional acceleration variance and reference step value respectively as
The value of the input neuron of indoor positioning neutral net；Using the actual step size value in each training data as indoor positioning nerve
The value of the output neuron of network；
Connection weight computing unit, value and output neuron for being respectively input into neuron according to the indoor positioning neutral net
Value, the connection weight between each hidden layer neuron in each input neuron and indoor positioning neutral net is calculated respectively, with
And connection weight between each hidden layer neuron and output neuron；
Amending unit, is modified for the connection weight according to multiple training data.
7. equipment according to claim 6, it is characterised in that the acquiring unit includes step-length acquiring unit and deviation angle
Degree acquiring unit：
The step-length acquiring unit includes,
Input determination subelement, for when the paces judging unit judge to advance complete a step when, paces frequency during this is walked,
Direct of travel acceleration variance, directional acceleration variance and reference step value are vertically travelled as the indoor positioning nerve net
The value of each input neuron of network；
Step size computation subelement, for according to the connection weight between each input neuron and each hidden layer neuron, and respectively
Connection weight between hidden layer neuron and output neuron, calculates the value of output neuron, and step value is entered as this walking；
The deviation angle acquiring unit, for when the paces judging unit judges to advance one step of completion, from intelligent terminal
Positioner obtain the deviation angle that this walking is entered before rear course is entered with respect to this walking.
8. equipment according to claim 6, it is characterised in that the positioning unit also includes：
Location determination subelement, for according to acceleration information in this step and paces frequency through indoor positioning neural computing
Obtain this walking to enter after step value, judge that whether described walking enters step value less than the 3rd preset value；
Continue locator unit, for being the then basis more than the 3rd preset value when the judged result of the location determination subelement
Described walking is entered step value current location is positioned with the deviation angle；
Again paces judgment sub-unit, for being less than the 3rd preset value, then when the judged result of the location determination subelement
Restart the paces judging unit.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310727255.3A CN104748735B (en) | 2013-12-25 | 2013-12-25 | Indoor orientation method and equipment based on intelligent terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310727255.3A CN104748735B (en) | 2013-12-25 | 2013-12-25 | Indoor orientation method and equipment based on intelligent terminal |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN104748735A CN104748735A (en) | 2015-07-01 |
| CN104748735B true CN104748735B (en) | 2017-06-27 |
Family
ID=53588780
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310727255.3A Active CN104748735B (en) | 2013-12-25 | 2013-12-25 | Indoor orientation method and equipment based on intelligent terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN104748735B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105223593B (en) * | 2015-09-17 | 2018-01-23 | 北京奇虎科技有限公司 | Terminal positioning frequency adjustment method, system and positioning interval adjusting method, system |
| CN105938659A (en) * | 2016-06-20 | 2016-09-14 | 中山大学 | Parking lot vehicle positioning method and system based on audio address coding |
| CN106643715A (en) * | 2016-11-17 | 2017-05-10 | 天津大学 | Indoor inertial navigation method based on bp neural network improvement |
| CN107515004B (en) * | 2017-07-27 | 2020-12-15 | 台州市吉吉知识产权运营有限公司 | Step length calculation device and method |
| CN107566978A (en) * | 2017-08-22 | 2018-01-09 | 上海爱优威软件开发有限公司 | A kind of tracking terminal method and system based on intelligent Neural Network |
| 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 |
| CN110866419A (en) * | 2018-08-28 | 2020-03-06 | 北京嘀嘀无限科技发展有限公司 | Step length determination method, system and computer readable storage medium |
| CN109883416A (en) * | 2019-01-23 | 2019-06-14 | 中国科学院遥感与数字地球研究所 | A kind of localization method and device of the positioning of combination visible light communication and inertial navigation positioning |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103411607A (en) * | 2013-08-30 | 2013-11-27 | 华中师范大学 | Method for pedestrian step size estimation and dead reckoning |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8688375B2 (en) * | 2006-05-31 | 2014-04-01 | Trx Systems, Inc. | Method and system for locating and monitoring first responders |
-
2013
- 2013-12-25 CN CN201310727255.3A patent/CN104748735B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103411607A (en) * | 2013-08-30 | 2013-11-27 | 华中师范大学 | Method for pedestrian step size estimation and dead reckoning |
Non-Patent Citations (3)
| Title |
|---|
| MEMS based pedestrian navigation system;Cho S Y,Park C G;《Journal of Navigation》;20061231;第1-2节、第4节第1段 * |
| 基于GPSOne和航位推算的定位算法研究;陈娇等;《计算机与数字工程》;20120131;第14卷(第1期);第16-18页 * |
| 基于iPhone OS架构的计步器应用软件设计;刘倩邓;《计算机工程与应用》;20111231;第182-184页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN104748735A (en) | 2015-07-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN104748735B (en) | Indoor orientation method and equipment based on intelligent terminal | |
| CN104061934B (en) | Pedestrian indoor position tracking method based on inertial sensor | |
| CN104406586B (en) | Pedestrian navigation apparatus and method based on inertial sensor | |
| CN105180937B (en) | A kind of MEMS IMU Initial Alignment Methods | |
| CN103558617B (en) | Positioning method and device | |
| CN106979780B (en) | A kind of unmanned vehicle real-time attitude measurement method | |
| CN106168485B (en) | Walking track data projectional technique and device | |
| CN103389506B (en) | A kind of adaptive filtering method for a strapdown inertia/Beidou satellite integrated navigation system | |
| CN105723240B (en) | The method and apparatus of the dislocation between equipment and ship are determined with acceleration/deceleration | |
| CN112334737A (en) | Positioning apparatus, positioning method, and program | |
| CN108844533A (en) | A kind of free posture PDR localization method based on Multi-sensor Fusion and attitude algorithm | |
| CN106643715A (en) | Indoor inertial navigation method based on bp neural network improvement | |
| CN107255474B (en) | PDR course angle determination method integrating electronic compass and gyroscope | |
| CN105122006A (en) | Method and system for varying step length estimation using nonlinear system identification | |
| CN110118560A (en) | A kind of indoor orientation method based on LSTM and Multi-sensor Fusion | |
| CN108362289B (en) | Mobile intelligent terminal PDR positioning method based on multi-sensor fusion | |
| CN103323012A (en) | Vehicular Ad hoc Network vehicle positioning method based on in-vehicle inertia device | |
| CN110184885A (en) | A method of surface evenness is tested based on smart phone | |
| Wahdan et al. | Three-dimensional magnetometer calibration with small space coverage for pedestrians | |
| CN106123900A (en) | Indoor pedestrian navigation magnetic heading calculation method based on modified model complementary filter | |
| CN104266648A (en) | Indoor location system based on Android platform MARG sensor | |
| CN105022055A (en) | IMU indoor positioning method | |
| CN105444763A (en) | IMU indoor positioning method | |
| CN106153069A (en) | Attitude rectification apparatus and method in autonomous navigation system | |
| CN103644910A (en) | Personal autonomous navigation system positioning method based on segment RTS smoothing algorithm |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| CP02 | Change in the address of a patent holder |
Address after: Room 818, 8 / F, 34 Haidian Street, Haidian District, Beijing 100080 Patentee after: BEIJING ULTRAPOWER SOFTWARE Co.,Ltd. Address before: 100089 Beijing city Haidian District wanquanzhuang Road No. 28 Wanliu new building 6 storey block A Room 601 Patentee before: BEIJING ULTRAPOWER SOFTWARE Co.,Ltd. |
|
| CP02 | Change in the address of a patent holder |