CN109151714A - A kind of three-dimensional Robust Estimation localization method - Google Patents
A kind of three-dimensional Robust Estimation localization method Download PDFInfo
- Publication number
- CN109151714A CN109151714A CN201810998022.XA CN201810998022A CN109151714A CN 109151714 A CN109151714 A CN 109151714A CN 201810998022 A CN201810998022 A CN 201810998022A CN 109151714 A CN109151714 A CN 109151714A
- Authority
- CN
- China
- Prior art keywords
- weight
- measurement group
- estimated location
- group
- squares
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/02—Services making use of location information
- H04W4/023—Services making use of location information using mutual or relative location information between multiple location based services [LBS] targets or of distance thresholds
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
- G01S5/02—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using radio waves
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W64/00—Locating users or terminals or network equipment for network management purposes, e.g. mobility management
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
Abstract
The present invention provides a kind of three-dimensional Robust Estimation localization method, belongs to wireless location technology field.The method steps are as follows: obtaining multiple TDOA measurement groups about positioning target, selects 2 or more effective measurement groups；Calculate the corresponding initial weight of each effectively measurement group；Solve the corresponding initial estimated location of each effectively measurement group；For each effective measurement group, according to its positioning equation group and its corresponding initial weight and corresponding initial estimated location, update is iterated to estimated location using Robust Estimation, until the estimated location for obtaining meeting iteration stopping condition is as final estimated position；Wherein, each iteration also updates weight；Calculate final positioning result of the average value of all final estimated positions as positioning target.The present invention is not only iterated update to estimated location when solving final estimated position, also update weight, can effectively overcome NLOS error bring to adversely affect, obtain higher positioning accuracy.
Description
Technical field
The present invention relates to a kind of three-dimensional Robust Estimation localization methods, belong to wireless location technology field.
Background technique
At present during wireless location, due to blocking for barrier caused by non line of sight (NLOS) error can be to positioning accurate
Degree has a huge impact.Existing research shows in the ecotopia propagated there is only sighting distance (LOS) or at mobile station (MS)
In the environment that NLOS error between all base stations (BS) can be eliminated, higher positioning accuracy can be realized.However, true
Indoor and outdoor surroundings be often complicated LOS/NLOS hybird environment, since the direct path of MS and BS is always hindered by barrier
Gear, measured observation data include the Errors Catastrophic as caused by exaggerated reflex path, cause to be difficult to obtain preferable positioning accurate
Degree.
Summary of the invention
The object of the present invention is to provide a kind of three-dimensional Robust Estimation localization methods, mixed in complicated LOS/NLOS to solve
Wireless location is carried out under cyclization border, it is difficult to the problem of obtaining preferable positioning accuracy.
To achieve the above object, the present invention provides a kind of three-dimensional Robust Estimation localization method, steps are as follows:
(1) obtain about positioning target multiple TDOA measurement groups, according to the measurement quality of TDOA measurement group select 2 with
On effective measurement group；
(2) the corresponding initial weight of each effectively measurement group is calculated；
(3) each positioning equation group of the effectively measurement group about positioning target of building solves each effectively measurement group and corresponds to
Initial estimated location；
(4) for each effectively measurement group, according to its positioning equation group and its corresponding initial weight and it is corresponding just
Beginning estimated location is iterated update to estimated location using Robust Estimation, until obtaining the estimation for meeting iteration stopping condition
Position is as final estimated position；Wherein, each iteration also updates weight；
(5) final positioning result of the average value of all final estimated positions as positioning target is calculated.
The beneficial effect of this method is: when solving final estimated position, being not only iterated update to estimated location, also
Weight is updated, can effectively overcome NLOS error bring to adversely affect, obtain higher positioning accuracy；In addition, passing through selection
The multiple groups measurement higher TDOA measurement group of quality is positioned, and the influence of NLOS error on the one hand can be reduced from source, separately
On the one hand it takes the average value of all final estimated positions as final positioning result, can also effectively reduce position error, further
Improve positioning accuracy.
In order to select multiple effective measurement groups, as a kind of improvement to above-mentioned three-dimensional Robust Estimation localization method, institute
Stating step (1) includes: to calculate the residual sum of squares (RSS) of each TDOA measurement group；Residual sum of squares (RSS) is selected to be less than the TDOA of given threshold
Measurement group is as effective measurement group.
In order to calculate the corresponding initial weight of each effective measurement group, as to above-mentioned three-dimensional Robust Estimation localization method
Another kind improves, and the step (2) includes: the average value for calculating the residual sum of squares (RSS) of all effective measurement groups, and average value is taken
Inverse is as the corresponding initial weight of each effectively measurement group；Or calculate separately each effective measurement group residual sum of squares (RSS) fall
Number, as the corresponding initial weight of corresponding measurement group.
In order to calculate the corresponding initial estimated location of each effective measurement group, as to above-mentioned three-dimensional Robust Estimation positioning side
Another improvement of method, the step (3) include: initially to estimate using each effective measurement group of least-squares algorithm solution is corresponding
Count position.
In order to realize the precise positioning to positioning target, change as another to above-mentioned three-dimensional Robust Estimation localization method
Into the step (4) includes: to carry out Taylor series expansion to each positioning equation group, ignores the above component of second order, and building is closed
In the matrix equation of positioning target position deviation and residual error；Using least-squares algorithm to the corresponding residual sum of squares (RSS) of the residual error
It carries out differential and seeks extreme value, obtain the position deviation general formula of positioning target；By corresponding initial weight and corresponding initial estimation position
It sets the substitution position deviation general formula and update is iterated to estimated location, updated estimated location is equal to preceding primary estimation position
It sets plus the resulting position deviation of current iteration, the estimated location up to obtaining meeting iteration stopping condition estimates position as final
It sets；Wherein, each iteration also updates weight, and updated weight is equal to a preceding weight and multiplies in weight factor, the weight factor benefit
It is obtained with the resulting residual sum of squares (RSS) of current iteration.
The present invention also provides another three-dimensional Robust Estimation localization methods, and steps are as follows:
(1) multiple TDOA measurement groups about positioning target are obtained, select highest one group of measurement quality to survey as effective
Amount group；
(2) the corresponding initial weight of effective measurement group is calculated；
(3) the positioning equation group of effective measurement group about positioning target is constructed, solves that effective measurement group is corresponding initially to be estimated
Count position；
(4) it is directed to effective measurement group, corresponding is estimated according to its positioning equation group and its corresponding initial weight and initially
Position is counted, update is iterated to estimated location using Robust Estimation, until obtaining the estimated location for meeting iteration stopping condition
Final positioning result as positioning target；Wherein, each iteration also updates weight.
The beneficial effect of this method is: when solving the final positioning result of positioning target, not only carrying out to estimated location
Iteration updates, also update weight, can effectively overcome NLOS error bring to adversely affect, obtain higher positioning accuracy；Separately
Outside, by selecting the measurement highest TDOA measurement group of quality to be positioned, the influence of NLOS error, energy are also reduced from source
Enough further increase positioning accuracy.
In order to select the measurement highest TDOA measurement group of quality as effective measurement group, estimate as to above-mentioned three-dimensional robust
A kind of improvement of localization method is counted, the step (1) includes: the residual sum of squares (RSS) for calculating each TDOA measurement group, selects residual error
The smallest TDOA measurement group of quadratic sum is as effective measurement group.
In order to calculate the corresponding initial weight of effective measurement group, as to the another of above-mentioned three-dimensional Robust Estimation localization method
Kind is improved, and the step (2) includes: that the inverse for the residual sum of squares (RSS) for calculating effective measurement group is corresponding just as effective measurement group
Beginning weight.
In order to calculate the corresponding initial estimated location of effective measurement group, as to above-mentioned three-dimensional Robust Estimation localization method
Another is improved, and the step (3) includes: to solve the corresponding initial estimated location of effective measurement group using least-squares algorithm.
In order to realize the precise positioning to positioning target, change as another to above-mentioned three-dimensional Robust Estimation localization method
Into the step (4) includes: to carry out Taylor series expansion to positioning equation group, ignores the above component of second order, building is about fixed
The matrix equation of position target position deviation and residual error；The corresponding residual sum of squares (RSS) of the residual error is carried out using least-squares algorithm
Differential seeks extreme value, obtains the position deviation general formula of positioning target；By corresponding initial weight and corresponding initial estimated location generation
Enter the position deviation general formula and update is iterated to estimated location, updated estimated location is equal to a preceding estimated location and adds
The resulting position deviation of current iteration, until the estimated location for obtaining meeting iteration stopping condition is as the final fixed of positioning target
Position result；Wherein, each iteration also updates weight, and updated weight is equal to a preceding weight and multiplies in weight factor, the power because
Son is obtained using the resulting residual sum of squares (RSS) of current iteration.
Detailed description of the invention
Fig. 1 is the localization method flow chart of the embodiment of the present invention 1；
Fig. 2 is the simulation result diagram of the corresponding MS of the embodiment of the present invention 1；
Fig. 3 is the position error fluctuation situation map of the corresponding MS of the embodiment of the present invention 1；
The locating effect figure of MS when Fig. 4 is the variation of the corresponding range error of the embodiment of the present invention 1.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with attached drawing and specific implementation
The present invention will be described in further detail for example.
In three-dimensional cartesian coordinate system, it is assumed that the true location coordinate of MS is X (x, y, z)；I-th of base station BSiPosition
Coordinate is Xi(xi,yi,zi), i=1,2,3 ..., m.
MS to i-th base station BSiDistance are as follows:
Enable BS1For reference base station, di,1Indicate MS to base station BSiAnd BS1Actual range it is poor:
di,1=di-d1=| | Xi-X||-||X1-X|| (2)
The TDOA measured value obtained using ultra wide band (UWB) technology measurement, can be obtained MS to base station BSiAnd BS1Measurement away from
Deviation:
Ri,1=cti,1 (3)
Wherein, c is propagation velocity of electromagnetic wave, ti,1For TDOA measured value, i.e. MS to base station BSiAnd BS1Time difference.
Embodiment 1
In conjunction with Fig. 1, the localization method of the embodiment of the present invention 1 is mainly comprised the steps of:
Step (1): effective measurement group is determined
During being positioned using TDOA algorithm, the quality of TDOA measured value has very big shadow to positioning accuracy
It rings.In true environment, the direct path of MS and BS are always stopped by barrier, because of the time delay that NLOS is generated, will lead
Cause MS to base station BSiAnd BS1Time difference ti,1It is bigger than normal, so as to cause measurement range difference Ri,1With actual range difference di,1There are larger
Deviation seriously affects positioning accuracy.
If obtaining K TDOA measurement group using UWB technology, when base station number is m, every group includes m-1 TDOA measurement
Value, the residual sum of squares (RSS) for defining each TDOA measurement group is S (X), then
According to analysis above it is found that when leading to reaching time-difference t by NLOSi,1When bigger than normal, residual sum of squares (RSS) S (X) can become
Greatly.Therefore, can establish the relationship between residual sum of squares (RSS) S (X) and the measurement quality of TDOA measurement group: S (X) is bigger, and TDOA is surveyed
The measurement quality of amount group is poorer.
In order to obtain preferable positioning accuracy, it is desirable to which the positioning result of MS, which can be depended on more, contains NLOS measured value
Less measurement group.In order to realize the purpose:
Firstly, calculating the residual sum of squares (RSS) of each TDOA measurement group；
Then, enable measurement group of the residual sum of squares (RSS) less than given threshold S as effective measurement group.
This step measures the measurement quality of TDOA measurement group using residual sum of squares (RSS)；It, can also be with as other embodiments
The measurement quality of TDOA measurement group is measured with carat Metro lower bound (CRLB).
Step (2): initialization survey value weight
This step calculates the corresponding initial weight of each effectively measurement group, that is, calculates the corresponding survey of each effectively measurement group
The initial weight of magnitude.
Assuming that being screened shared k effectively measurement groups, S is usedk(X) residual sum of squares (RSS) of kth group is indicated.In order to eliminate residual error
The relationship of size between quadratic sum group calculates the average value of the residual sum of squares (RSS) of k effective measurement groups are as follows:
It defines in above-mentioned k effective measurement groups, the initial weight of each measured value are as follows:
This step calculates the average value of the residual sum of squares (RSS) of all effective measurement groups, and average value is inverted as each
The effectively corresponding initial weight of measurement group；As other embodiments, the residual error that can also calculate each effective measurement group is flat
Side obtains the corresponding initial weight of each effective measurement group multiplied by scheduled coefficient with after.
Step (3): MS initial position estimation
This step constructs each effectively positioning equation group of the measurement group about MS, and solves each effectively measurement group and correspond to
Initial estimated location, that is, solve the initial estimated location of the corresponding MS of each effectively measurement group.
In step (2), k effectively measurement groups are filtered out altogether, therefore can construct the k positioning equation groups about MS, into
And the initial estimated location of k MS can be acquired.
Below by taking individually effective measurement group as an example, it is discussed in detail and solves the initial of MS using total least square method (TLS)
The calculating process of the process of estimated location, remaining effective measurement group is similar:
By an effective measurement group, the positioning equation group about MS initial estimated location, i.e. TDOA equation group are constructed:
di,1=cti,1=di-d1 (7)
Since TDOA equation group is Nonlinear System of Equations, system of linear equations need to be translated into and then utilized totally most
Small square law solves, and detailed process is as follows:
It is easy to get by formula (7):
di=di,1+d1 (8)
Formula (8) both ends square can be obtained simultaneously:
di 2=di,1 2+d1 2+2di,1d1 (9)
From formula (1):
It substitutes into formula (9), can obtain:
(xi-x)2+(yi-y)2+(zi-z)2=di,1 2+d1 2+2di,1d1 (10)
As i=1:
d1 2=(x1-x)2+(y1-y)2+(z1-z)2
It substitutes into formula (10), can obtain:
2(xi-x1)x+2(yi-y1)y+2(zi-z1)z+2di,1d1=xi 2+yi 2+zi 2-x1 2-y1 2-z1 2-di,1 2
It enables
xi,1=xi-x1
yi,1=yi-y1
zi,1=zi-z1
ki=xi 2+yi 2+zi 2
It can obtain:
D againi,1=cti,1, then final formula (7) may be expressed as:
Wherein, ti,1It can be obtained by measurement, simultaneously because all base station BSsiCoordinate Xi(xi,yi,zi) it is known that then formula
It (12) is all known in addition to x, y, z in, therefore formula (12) is actually the system of linear equations that unknown number is x, y, z, it is real
The linearisation of formula (7) is showed.
Formula (12) vectorization can be obtained into matrix equation:
A β=B (13)
Wherein, A is coefficient matrix:
Vector β are as follows:
β=[x y z d1]T
Observation vector B are as follows:
Due to reaching time-difference ti,1Measured value there is error, cause coefficient matrices A and observation vector B can be by
To the interference of extraneous factor, therefore can be using the initial position of TLS estimation MS.
Augmented matrix H=[A B] is constructed first, is then solved using singular value decomposition method (SVD), solution needs three altogether
A step:
I. singular value decomposition is carried out to augmented matrix H=[A B], obtained:
Wherein, U=[U1 U2]∈Ra×aIt is [A B] [A B]TA feature vector composition orthogonal matrix；
It isB+1
The orthogonal matrix of a feature vector composition；
It is the matrix of the e+1 singular value composition of matrix H, and σ1≥σ2
≥…≥σt> σe+1。
II. theoretical according to Eckart-Young-Mirsky matrix approximation, the optimal solution of H=[A B] meets:
Wherein, Σ1=diag (σ1,σ2,…,σe,0)。
III. judge V22It is nonsingular, then initial estimation are as follows:
V can be readily available according to coefficient matrices A and observation vector B12And V22, to obtain formula (13) using TLS
Solution β0=[x0 y0 z0 d1] to get to MS initial estimated location be X0(x0,y0,z0)。
This step solves the initial estimated location of MS using total least square method；It, can be with as other embodiments
It is solved using Chan algorithm, Fang algorithm, weighted least-squares method etc..
Step (4): Robust Estimation
The essence of Robust Estimation is initialization weight and Gauss-Newton iterative algorithm based on Jacobian matrix.This step
Suddenly, for each effective measurement group, according to its positioning equation group and its corresponding initial weight and corresponding initial estimation position
It sets, solves the final estimated position of the corresponding MS of each effectively measurement group.
Below by taking individually effective measurement group as an example, the detailed process for solving the final estimated position of MS is discussed in detail, remaining
The calculating process of effective measurement group is similar:
1) Taylor series expansion
Taylor series expansion is carried out to formula (7), and ignores the above component of second order, is constructed about MS position deviation Δ
With the matrix equation of residual epsilon:
H=G Δ+ε (15)
In formula, vector ε indicates residual error；
Δ is the position deviation of MS:
Δ=[Δ x Δ y Δ z]T
G is coefficient matrix:
H is the difference between measured value and estimated distance difference:
Using weighted least square algorithm solution formula (15), the estimated value of MS position deviation Δ is obtained are as follows:
Δ=(GTPG-1)GTPh (16)
Wherein, P is measured value weight.
2) iteration updates
After each iteration, the estimated location of MS is updated, and judge whether the estimated location of updated MS meets repeatedly
For stop condition, if satisfied, then the estimated location of the updated MS is the final estimated position of MS；Otherwise, it is being no more than
In the case where maximum number of iterations, measured value weight is updated using residual sum of squares (RSS), and repeat iterative process, until meeting iteration
Stop condition.
Nth iteration process is as follows: n=1, and 2,3 ..., N, N are maximum number of iterations.
Firstly, the estimated location X of the MS obtained in preceding an iterationn-1(xn-1,yn-1,zn-1) at carry out Taylor expansion, obtain
To Gn-1、hn-1And the measured value weight P that preceding an iteration obtainsn-1Formula (16) are substituted into together obtains the position deviation Δ of MSn:
Δn=(Gn-1 TPn-1Gn-1 -1)Gn-1 TPn-1hn-1 (17)
Wherein,
To which the estimated location for updating MS is Xn:
Xn=Xn-1+Δn (20)
Then, X is judged using root-mean-square error (Root mean square error, RMSE)nWhether meet iteration to stop
Only condition, at this time XnFor (xn,yn,zn), root-mean-square error are as follows:
Set threshold value γ:
1. as RMSE < γ, then XnMeet iteration stopping condition, stop iteration at this time and returns to the estimated location X of MSn, Xn
The as final estimated position of MS；
2. then measured value weight is updated using residual sum of squares (RSS), into next time as RMSE > γ and the number of iterations n≤N
Iteration, updated measured value weight are Pn:
Pn=Pn-1ωn (22)
Wherein, ω is weight factor, also referred to as equivalent weight function, chooses IGGIII weight function herein and is updated to weight:
In formula, s indicates residual sum of squares (RSS), generally takes k0=1.0~2.5, k1=3.0~8.0.
Enable s s when nth iterationnIt indicates, ω ωnIt indicates, ε εnIt indicates, when updating measured value weight, first with
ΔnAcquire residual epsilonn:
εn=hn-1-Gn-1Δn (24)
Further acquire the residual sum of squares (RSS) s under the residual errorn, then by snIt substitutes into formula (23) and obtains ωn, finally substitute into
It is P that updated measured value weight, which can be obtained, in formula (22)n。
3. then output error and exiting operation when the number of iterations is greater than N.
Such as:
In initial weight P0With initial estimated location X0On the basis of carry out the 1st iteration, obtain the position deviation Δ of MS1
=(G0 TP0G0 -1)G0 TP0h0, in which:
Then, the estimated location for updating MS is X1=X0+Δ1, judge X1Whether meet iteration stopping condition, that is, judgesWhether threshold value γ is less than, if so, X1The final estimation of as MS
Position；Otherwise, in the case where being no more than maximum number of iterations, updating measured value weight using residual sum of squares (RSS) is P1=P0
ω1, ω1By by Δ1It substitutes into formula (24) and first acquires residual epsilon1, further acquire the residual sum of squares (RSS) s under the residual error1, then will
s1Formula (23) are substituted into obtain.
This step solves the general formula of MS position deviation using weighted least square algorithm；It, can also as other embodiments
To be solved using robust least squares algorithm.
This step updates weight factor used in measured value weight every time, is by IGGIII weight function by residual sum of squares (RSS)
It is converted to；As other embodiments, can also with Huber weight function, Andrews weight function, Tukey weight function or
IGGI weight function calculates weight factor.
Step (5): the final positioning result of MS is calculated
This step calculates the average value of the final estimated position of k MS as the final of MS on the basis of step (4)
Positioning result.
Embodiment 2
The present embodiment, the difference with embodiment 1 are only that: the process of step (2) initialization survey value weight are as follows: count respectively
The inverse for calculating the residual sum of squares (RSS) of each effective measurement group, the initial weight as the corresponding measured value of corresponding measurement group；Remaining step
Rapid all the same, details are not described herein again.
Embodiment 3
The present embodiment, the difference from embodiment 1 is that: effective measurement group in step (1) passes through selection residual sum of squares (RSS)
A smallest measurement group obtains；Correspondingly, the initial weight of measured value is the residual sum of squares (RSS) of the measurement group in step (2)
It is reciprocal.The initial estimated location of MS is solved for the measurement group later and solves the method and step of the final estimated position of MS
(3) and step (4) is identical, and details are not described herein again, it should be pointed out that since effective measurement group of the present embodiment only has 1, because
The final estimated position of this MS is the final positioning result of MS.
As other embodiments, the sequence of step (2) and step (3) in above-described embodiment 1, embodiment 2 and embodiment 3
It can exchange.
Specific simulating, verifying experiment is given below:
In simulation process, influenced caused by positioning for simulation NLOS non-market value, it is assumed that each TDOA measurement error
Between it is mutually indepedent, and obey zero-mean, variance δ2Gaussian Profile, herein by TDOA measurement error be converted into ranging miss
Difference, range error are the product of TDOA measurement error and velocity of electromagnetic waves.
The positioning to MS is realized using 5 locating base stations, and the true coordinate of MS is X (3,10,8), base station BS1、BS2、BS3、
BS4、BS5Coordinate be respectively as follows: X1(0,0,10)、X2(20,0,0)、X3(0,20,0)、X4(-20,0,0)、X5(0, -20,0),
Middle BS1For reference base station.
20 TDOA measurement groups about MS, the variance δ of range error are obtained using UWB technology2=4, utilize reality above
Localization method described in example 1 is applied, the positioning result of MS is emulated, simulation result is as shown in Figure 2.In figure, circle represents base
It stands, star represents the actual position of MS, and triangle represents the estimated location of MS.
It can be obtained by simulation result, the final positioning result of MS are as follows: (2.96,9.92,7.48), the actual position relative to MS
Coordinate (3,10,8) is 52.76cm by the position error that range error introduces.
For the randomness for avoiding position error, 100 emulation experiments are carried out to the positioning of MS, position error fluctuates feelings
Condition is as shown in Figure 3, it will thus be seen that in the variance δ of range error2When=4, position error RMSE is mainly between 0.3m to 0.5m
Fluctuation, i.e., position error is mainly between 30cm to 50cm, and relative to the localizing environment, error at this time is very little, can be with
It is considered to be accurately positioned.
Influence of the range error to locating effect is further investigated, as shown in Figure 4: engagement arithmetic A is described in embodiment 1
Localization method, algorithm B are the Taylor series expansion algorithms based on MS actual position.
As shown in Figure 4, in range error less (δ2< 7) when, the position error RMSE of algorithm A and algorithm B is almost the same,
Illustrate that the positioning accuracy of the two is almost the same；When range error is further enlarged, the position error RMSE of algorithm A and algorithm B
Difference remain within 1.5m, illustrate that the positioning accuracy of algorithm A is also closer to the positioning accuracy of algorithm B.
To sum up, three-dimensional Robust Estimation localization method proposed by the present invention is carried out using the high TDOA measurement group of measurement quality
Positioning, and positioning result is not only updated in an iterative process, measured value weight also is updated using residual sum of squares (RSS), it can be in complexity
LOS/NLOS hybird environment under obtain higher positioning accuracy, obtained MS positioning result can preferably reflect that MS's is true
Position.
Specific embodiment of the present invention is presented above, but the present invention is not limited to described embodiment.
Under the thinking that the present invention provides, to the skill in above-described embodiment by the way of being readily apparent that those skilled in the art
Art means are converted, are replaced, are modified, and play the role of with the present invention in relevant art means it is essentially identical, realize
Goal of the invention it is also essentially identical, the technical solution formed in this way is to be finely adjusted to be formed to above-described embodiment, this technology
Scheme is still fallen in protection scope of the present invention.
Claims (10)
1. a kind of three-dimensional Robust Estimation localization method, which is characterized in that steps are as follows:
(1) multiple TDOA measurement groups about positioning target are obtained, select 2 or more according to the measurement quality of TDOA measurement group
Effective measurement group；
(2) the corresponding initial weight of each effectively measurement group is calculated；
(3) it is corresponding just to solve each effectively measurement group for each positioning equation group of the effectively measurement group about positioning target of building
Beginning estimated location；
(4) for each effectively measurement group, corresponding estimate according to its positioning equation group and its corresponding initial weight and initially
Position is counted, update is iterated to estimated location using Robust Estimation, until obtaining the estimated location for meeting iteration stopping condition
As final estimated position；Wherein, each iteration also updates weight；
(5) final positioning result of the average value of all final estimated positions as positioning target is calculated.
2. three-dimensional Robust Estimation localization method according to claim 1, it is characterised in that: the step (1) includes: to calculate
The residual sum of squares (RSS) of each TDOA measurement group；Residual sum of squares (RSS) is selected to be less than the TDOA measurement group of given threshold as effective measurement
Group.
3. three-dimensional Robust Estimation localization method according to claim 2, it is characterised in that: the step (2) includes: to calculate
The average value of the residual sum of squares (RSS) of all effective measurement groups, average value is inverted corresponding initial as each effectively measurement group
Weight；Or the inverse of the residual sum of squares (RSS) of each effective measurement group is calculated separately, as the corresponding initial weight of corresponding measurement group.
4. three-dimensional Robust Estimation localization method according to claim 3, it is characterised in that: the step (3) includes: to utilize
Least-squares algorithm solves the corresponding initial estimated location of each effectively measurement group.
5. three-dimensional Robust Estimation localization method according to claim 4, it is characterised in that: the step (4) includes: to every
A positioning equation group carries out Taylor series expansion, ignores the above component of second order, and building is about positioning target position deviation and residual
The matrix equation of difference；Differential is carried out to the corresponding residual sum of squares (RSS) of the residual error using least-squares algorithm and seeks extreme value, is determined
The position deviation general formula of position target；Corresponding initial weight and corresponding initial estimated location are substituted into the position deviation general formula
Update is iterated to estimated location, updated estimated location is equal to a preceding estimated location and adds the resulting position of current iteration
Deviation, until the estimated location for obtaining meeting iteration stopping condition is as final estimated position；Wherein, each iteration also updates power
Weight, updated weight are equal to a preceding weight and multiply in weight factor, and the weight factor utilizes the resulting residuals squares of current iteration
With obtain.
6. a kind of three-dimensional Robust Estimation localization method, which is characterized in that steps are as follows:
(1) multiple TDOA measurement groups about positioning target are obtained, select highest one group of quality of measurement as effective measurement group；
(2) the corresponding initial weight of effective measurement group is calculated；
(3) positioning equation group of effective measurement group about positioning target is constructed, the corresponding initial estimation position of effective measurement group is solved
It sets；
(4) it is directed to effective measurement group, according to its positioning equation group and its corresponding initial weight and corresponding initial estimation position
It sets, update is iterated to estimated location using Robust Estimation, until obtaining the estimated location conduct for meeting iteration stopping condition
Position the final positioning result of target；Wherein, each iteration also updates weight.
7. three-dimensional Robust Estimation localization method according to claim 6, it is characterised in that: the step (1) includes: to calculate
The residual sum of squares (RSS) of each TDOA measurement group selects the smallest TDOA measurement group of residual sum of squares (RSS) as effective measurement group.
8. three-dimensional Robust Estimation localization method according to claim 7, it is characterised in that: the step (2) includes: to calculate
The inverse of the residual sum of squares (RSS) of effective measurement group is as the corresponding initial weight of effective measurement group.
9. three-dimensional Robust Estimation localization method according to claim 8, it is characterised in that: the step (3) includes: to utilize
Least-squares algorithm solves the corresponding initial estimated location of effective measurement group.
10. three-dimensional Robust Estimation localization method according to claim 9, it is characterised in that: the step (4) includes: pair
Positioning equation group carries out Taylor series expansion, ignores the above component of second order, building is about positioning target position deviation and residual error
Matrix equation；Differential is carried out to the corresponding residual sum of squares (RSS) of the residual error using least-squares algorithm and seeks extreme value, is positioned
The position deviation general formula of target；Corresponding initial weight and corresponding initial estimated location are substituted into the position deviation general formula pair
Estimated location is iterated update, and updated estimated location is equal to a preceding estimated location and adds the resulting position of current iteration inclined
Difference, until final positioning result of the estimated location for obtaining meeting iteration stopping condition as positioning target；Wherein, each iteration
Weight is also updated, updated weight is equal to a preceding weight and multiplies in weight factor, and the weight factor is resulting using current iteration
Residual sum of squares (RSS) obtains.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810998022.XA CN109151714A (en) | 2018-08-29 | 2018-08-29 | A kind of three-dimensional Robust Estimation localization method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810998022.XA CN109151714A (en) | 2018-08-29 | 2018-08-29 | A kind of three-dimensional Robust Estimation localization method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109151714A true CN109151714A (en) | 2019-01-04 |
Family
ID=64829276
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810998022.XA Pending CN109151714A (en) | 2018-08-29 | 2018-08-29 | A kind of three-dimensional Robust Estimation localization method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109151714A (en) |
Cited By (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110221244A (en) * | 2019-05-24 | 2019-09-10 | 宁波大学 | Based on the robust positioning method of reaching time-difference under the conditions of non line of sight |
| CN110221245A (en) * | 2019-05-28 | 2019-09-10 | 宁波大学 | The robust TDOA localization method of Combined estimator target position and non-market value |
| CN110730502A (en) * | 2019-10-23 | 2020-01-24 | 珠海优特电力科技股份有限公司 | Positioning method and device |
| CN110850366A (en) * | 2019-10-31 | 2020-02-28 | 宁波大学 | Positioning method based on received signal strength under mixed line-of-sight and non-line-of-sight environment |
| CN111308423A (en) * | 2019-12-11 | 2020-06-19 | 中国人民解放军陆军工程大学 | Robust sound source positioning system and method thereof |
| CN113341373A (en) * | 2021-05-31 | 2021-09-03 | 中国电子科技集团公司第三十六研究所 | Positioning method and device and electronic equipment |
| US20220132461A1 (en) * | 2020-10-28 | 2022-04-28 | Hitachi, Ltd. | Positioning Method |
| CN114760587A (en) * | 2022-06-14 | 2022-07-15 | 湖南工商大学 | Positioning method based on UWB sensor base station grouping strategy |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1623104A (en) * | 2000-10-18 | 2005-06-01 | 英特尔公司 | Remote terminal location algorithm |
| CN101730227A (en) * | 2009-11-10 | 2010-06-09 | 大连理工大学 | Multi-base station secondary positioning method based on toughness estimation and arrival time difference |
| CN102540202A (en) * | 2010-09-10 | 2012-07-04 | 塔莱斯公司 | Method for correcting position estimations by selecting pseudo-distance measurements |
| CN106707235A (en) * | 2017-03-08 | 2017-05-24 | 南京信息工程大学 | Indoor range finding positioning method based on improved traceless Kalman filtering |
-
2018
- 2018-08-29 CN CN201810998022.XA patent/CN109151714A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1623104A (en) * | 2000-10-18 | 2005-06-01 | 英特尔公司 | Remote terminal location algorithm |
| CN101730227A (en) * | 2009-11-10 | 2010-06-09 | 大连理工大学 | Multi-base station secondary positioning method based on toughness estimation and arrival time difference |
| CN102540202A (en) * | 2010-09-10 | 2012-07-04 | 塔莱斯公司 | Method for correcting position estimations by selecting pseudo-distance measurements |
| CN106707235A (en) * | 2017-03-08 | 2017-05-24 | 南京信息工程大学 | Indoor range finding positioning method based on improved traceless Kalman filtering |
Non-Patent Citations (1)
| Title |
|---|
| SHUAIXIN LI ET.AL: "A three-dimensional robust ridge estimation positioning method for UWB in a complex environment", 《ADVANCES IN SPACE RESEARCH》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110221244A (en) * | 2019-05-24 | 2019-09-10 | 宁波大学 | Based on the robust positioning method of reaching time-difference under the conditions of non line of sight |
| CN110221244B (en) * | 2019-05-24 | 2022-04-19 | 宁波大学 | Robust positioning method based on arrival time difference under non-line-of-sight condition |
| CN110221245A (en) * | 2019-05-28 | 2019-09-10 | 宁波大学 | The robust TDOA localization method of Combined estimator target position and non-market value |
| CN110730502A (en) * | 2019-10-23 | 2020-01-24 | 珠海优特电力科技股份有限公司 | Positioning method and device |
| CN110850366A (en) * | 2019-10-31 | 2020-02-28 | 宁波大学 | Positioning method based on received signal strength under mixed line-of-sight and non-line-of-sight environment |
| CN110850366B (en) * | 2019-10-31 | 2021-07-16 | 宁波大学 | Positioning method based on received signal strength under mixed line-of-sight and non-line-of-sight environment |
| CN111308423A (en) * | 2019-12-11 | 2020-06-19 | 中国人民解放军陆军工程大学 | Robust sound source positioning system and method thereof |
| CN111308423B (en) * | 2019-12-11 | 2022-05-24 | 中国人民解放军陆军工程大学 | Robust sound source positioning system and method thereof |
| US20220132461A1 (en) * | 2020-10-28 | 2022-04-28 | Hitachi, Ltd. | Positioning Method |
| CN113341373A (en) * | 2021-05-31 | 2021-09-03 | 中国电子科技集团公司第三十六研究所 | Positioning method and device and electronic equipment |
| CN113341373B (en) * | 2021-05-31 | 2024-05-14 | 中国电子科技集团公司第三十六研究所 | Positioning method and device and electronic equipment |
| CN114760587A (en) * | 2022-06-14 | 2022-07-15 | 湖南工商大学 | Positioning method based on UWB sensor base station grouping strategy |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109151714A (en) | A kind of three-dimensional Robust Estimation localization method | |
| Payami et al. | Channel measurements and analysis for very large array systems at 2.6 GHz | |
| CN103969622B (en) | A kind of time difference positioning method based on the receiving station that does more physical exercises | |
| El-Sayed et al. | Evaluation of localization methods in millimeter-wave wireless systems | |
| CN103476116B (en) | The anti-NLoS error location method merged based on positioning unit quality and many algorithm data | |
| CN106405533B (en) | Radar target combined synchronization and localization method based on constraint weighted least-squares | |
| CN110493742A (en) | A kind of indoor 3-D positioning method for ultra wide band | |
| CN110174643A (en) | A kind of localization method based on reaching time-difference without noise power information | |
| CN107820206B (en) | Non-line-of-sight positioning method based on signal intensity | |
| CN105898865A (en) | Cooperative location method based on EKF (Extended Kalman Filter) and PF (Particle Filter) under nonlinear and non-Gaussian condition | |
| CN110471049A (en) | Wireless communication interference source localization method based on measurement radiant power | |
| Priebe et al. | Calibrated broadband ray tracing for the simulation of wave propagation in mm and sub-mm wave indoor communication channels | |
| CN110784823B (en) | Outdoor target positioning method based on BP neural network and TDOA | |
| CN110954865A (en) | Short wave time difference positioning method based on ionosphere information | |
| CN106443655A (en) | Multiple-input-multiple-output radar near-field positioning algorithm | |
| CN102970749A (en) | Multi-base-station successive approximation positioning method | |
| CN108279411B (en) | MDS-based passive MIMO time difference positioning method | |
| Navarro et al. | TOA and DOA estimation for positioning and tracking in IR-UWB | |
| CN104683949A (en) | Antenna-array-based hybrid self-positioning method applied to wireless Mesh network | |
| CN111007456A (en) | Robust non-line-of-sight deviation elimination positioning method capable of realizing time domain combination | |
| CN110333481A (en) | A kind of joint positioning method based on distance than round algorithm and Taylor Series Expansion Method | |
| CN113708807B (en) | Channel modeling method based on MIMO-OTA base station static test | |
| Wang et al. | ToA and TDoA error models for NLoS propagation based on outdoor to indoor channel measurement | |
| CN110536410A (en) | The localization method measured under nlos environment based on RSS and TDOA | |
| Isa et al. | Enhanced location estimation with a single base station in WiMAX network |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190104 |