CN106793078B - Bluetooth indoor positioning method based on RSSI correction value dual positioning - Google Patents

Abstract

The invention discloses a bluetooth indoor positioning method based on dual positioning of RSSI correction value. The scheme is as follows: 1. Collect the received signal strength values received by each iBeacon node from other iBeacon nodes and correct them to obtain the received signal strength values of the iBeacon nodes Correction matrix; 2. Calculate the distance matrix from each node to other nodes according to the correction matrix; 3. Estimate the coordinate value of each node from the distance matrix to obtain the coordinate error; 4. Collect the received signal strength value of the iBeacon node received by the node to be positioned and compare Its correction obtains the signal strength correction matrix of the node to be located; 5. Obtains the distance matrix from the node to be located to other iBeacon nodes according to the correction matrix, and then obtains the estimated coordinate value of the node to be located; 6. The result of step 3 and step 5 Get the final coordinates of the node to be located. The invention has high positioning accuracy and can be used in nursing homes and smart communities.

Description

Bluetooth indoor positioning method based on RSSI correction value dual positioning

technical field

The invention belongs to the technical field of communication, and in particular relates to a Bluetooth indoor positioning method, which can be used in shopping malls, nursing homes, smart communities and fire places.

Background technique

With the rapid development of the Internet of Things society, sensing smart devices are more and more widely used in the smart society due to their advantages of low power consumption, self-organization, and convenient layout, and indoor positioning has become an important supporting field of the Internet of Things society. The indoor positioning technology based on Bluetooth iBeacon will become one of the hotspots of indoor positioning due to its unique advantages.

At present, the indoor positioning algorithms based on Bluetooth mainly include the algorithm TOA based on the signal transmission time, the algorithm TDOA based on the signal transmission time difference, the algorithm AOA based on the signal angle of arrival, the algorithm based on the received signal strength value RSSI and the single centroid algorithm, among which:

TOA algorithm, because the indoor environment space is relatively narrow, the signal propagation will be delayed by the interference of obstacles, and the superposition of time delay will occur. At the same time, this algorithm requires precise synchronization of the clock between the positioning node and the reference node , the hardware requirements are very high, and the complexity and cost of the system are relatively large, resulting in low practicability.

The TDOA algorithm is an improvement of the TOA algorithm. This algorithm does not directly use the absolute time when the signal arrives at the reference node, but uses the time difference between the signal arrival at two reference nodes to determine the position of the positioning node, so there is no need for a link between the reference node and the positioning node. The precise synchronization of clocks only needs to be synchronized between reference nodes. Although this algorithm reduces the requirements for clock synchronization, its accuracy is low.

AOA algorithm, this algorithm needs to install an antenna matrix on the node to obtain angle information, but because most of the antennas of the nodes are omnidirectional, it is impossible to distinguish which direction the signal comes from, and installing the antenna on the Bluetooth node requires special hardware equipment , such as antenna arrays, cause Bluetooth nodes to exceed ordinary sensor nodes in terms of energy consumption, size, and price, which is contrary to the characteristics of low cost and low power consumption of wireless smart devices, so the practicability is poor.

RSSI algorithm, this algorithm is to obtain the relationship between the received signal strength value RSSI and the distance, and obtain the model between the distance and the RSSI, so as to perform indoor positioning. However, due to the free attenuation effect, signal absorption effect, non-line-of-sight propagation effect, multipath effect, and shadow effect of the radio signal in the indoor environment, the RSSI value of the Bluetooth signal is changing drastically with the change of the distance, resulting in the traditional receiver-based The positioning algorithm of the signal strength value RSSI has a large error range. However, due to its low requirements for clock processing, low cost, and low complexity, general indoor positioning algorithms are based on RSSI.

The single centroid algorithm is to receive the information of all reference nodes within its communication range through the positioning node, and use the geometric centroid of the reference node as its own estimated position to locate, but the number of centroids is single and far away from the node to be positioned The influence degree of the reference node is small, and the contribution degree of the reference node farther away from the node to be positioned cannot be highlighted, which brings about a large error and does not further improve the positioning accuracy.

Contents of the invention

The object of the present invention is to address the above deficiencies in the prior art, and propose a Bluetooth indoor positioning method based on RSSI correction value dual positioning, so as to improve the indoor positioning accuracy of Bluetooth iBeacon.

For achieving above object, technical scheme of the present invention comprises as follows:

(1) Collect the received signal strength values of other iBeacon nodes received by each iBeacon node deployed indoors to form an iBeacon node signal strength matrix R, and use the Dixon detection method and Gaussian filter algorithm to correct it to obtain each iBeacon node The signal strength correction value of the iBeacon node constitutes the signal strength correction matrix R':

in:

i, j are iBeacon node numbers, n is the number of iBeacon nodes, i∈[1,n], j∈[1,n], n>3;

R _ij ＝[R _{ij1 R ij2 R ij3} ... R _ij30 ] is a 1×30-dimensional matrix, indicating that the i-th iBeacon node receives 30 sets of signal values sent by the j-th iBeacon node;

R _ij ' represents the correction value of the signal strength received by the i-th iBeacon node from the j-th iBeacon node.

(2) According to the iBeacon node signal strength correction matrix R', using the logarithmic distance path loss model, the distance value from the iBeacon node to other iBeacon nodes is obtained to form the iBeacon node distance matrix D:

Where: D _ij represents the distance value from the i-th iBeacon node to the j-th iBeacon node, when i=j, D _ij =0.

(3) According to the iBeacon node distance matrix D, the multi-centroid algorithm is used to obtain the estimated coordinate value Q _{(xi ,} y _{i )} of each iBeacon node, and the actual coordinates e(v _i , z _i ) comparison, get the error point P(α,β);

(4) Collect the received signal strength value sent by the node to be positioned to receive the iBeacon node to form the signal strength matrix r of the positioned node, and use the Dixon detection method and Gaussian filter algorithm to correct it to obtain the strength correction value of the node to be positioned. Constitute the signal strength correction matrix r' of the node to be located:

r=[r ₁ r ₂ r ₃ ... r _i ... r _n ], r'=[r ₁ ' r ₂ ' r ₃ ' ... r _i ' ... r _n ']

in:

r _i is a 1×30-dimensional matrix, indicating that the node to be positioned receives 30 sets of signal values sent by the i-th iBeacon node;

r _i ' represents the correction value of the signal strength received by the node to be positioned from the i-th iBeacon node.

(5) According to the signal strength correction matrix r' of the node to be located, the logarithmic distance path loss model is used to obtain the distance value from the node to be located to other iBeacon nodes, and the distance matrix d of the node to be located is formed:

d=[d ₁ d ₂ d ₃ ... d _i ... d _n ]

Among them: d _i is the distance value from the node to be positioned to the i-th iBeacon node.

(6) According to the distance matrix d of the node to be located, the multi-centroid algorithm is used to obtain the estimated coordinate value W(x, y) of the node to be located;

(7) According to the estimated coordinate value W(x, y) and the error point P(α, β), the final coordinate of the node to be positioned is: W(x+α, y+β).

The main advantage of the present invention relative to prior art is:

First, compared with the existing preprocessing in the previous stage, the present invention processes the Dixon detection algorithm on the received signal strength values with large fluctuations in the early stage, eliminates the received signal strength values with severe jitter, and then performs the processing on the retained data Gaussian filtering is used to preserve the received signal strength value with less jitter to the greatest extent.

Second, compared with the existing path loss model, the present invention provides the best environmental coefficient reference range based on a large amount of experimental data, which is not only simple to implement, but also the selected coefficients are more suitable for indoor environments.

Thirdly, compared with the existing positioning algorithm, the present invention calculates the coordinate error of the system and the estimated coordinate of the node to be positioned due to the dual positioning algorithm, so it can calculate the coordinate error of the system and the estimated coordinate of the node to be positioned according to the coordinate error The final coordinates of the bit node.

Fourth, compared with the existing positioning algorithm, the present invention groups iBeacon nodes far away from the node to be positioned, after finding its centroid, applies the centroid to the group of iBeacon nodes that are closer to the node to be positioned, and then calculates The center of mass is taken to further improve the positioning accuracy, and at the same time, it can prevent the fitness from changing with the state of the iBeacon node in the environment.

Description of drawings

Fig. 1 is the realization flowchart of the present invention;

Fig. 2 is a comparison diagram of positioning effects between the present invention and the existing single centroid algorithm.

Detailed ways

With reference to Fig. 1, the bluetooth indoor positioning method based on RSSI correction value dual positioning of the present invention, its specific implementation steps are as follows:

Step 1, construct the iBeacon node signal strength matrix R.

Divide the indoor environment into several areas with regular hexagons, place an iBeacon node at each vertex of the area, place a total of n iBeacon nodes, each iBeacon node periodically broadcasts its own number, its own coordinates and received signal strength values, and collects The received signal strength values of other iBeacon nodes received by each iBeacon node form the iBeacon node signal strength matrix R:

Where: i, j are iBeacon node numbers, n is the number of iBeacon nodes, i∈[1,n], j∈[1,n], n>3;

R _ij =[R _{ij1 R ij2 R ij3} ... R _ij30 ] is a 1×30-dimensional matrix, indicating that the i-th iBeacon node receives 30 sets of signal values sent by the j-th iBeacon node.

In step 2, the iBeacon node signal strength matrix R is corrected to obtain the iBeacon node Dixon matrix R ^* .

2a) Use the Dixon detection method to eliminate the received signal strength value of the 30 sets of signal values received by each iBeacon node from other iBeacon nodes, and receive the i-th iBeacon node The 30 groups of received signal strength values are arranged in ascending order, which are R _{ij1 , R ij2 ,} ...,R _ij30 ;

2b) Determine the detection level of outlier detection as a=0.02, and determine the Dixon test critical value M(a,n);

2c) According to the statistical formula of the Dixon test method, calculate the abnormal value G of the highest end of the iBeacon node and the abnormal value G' of the lowest end of the iBeacon node:

2d) Compare the highest-end outlier G of the iBeacon node and the lowest-end outlier G' of the iBeacon node with the critical value M(a,n):

If G>M(a,n) or G'>M(a,n), then remove the received signal strength value R _ijN corresponding to the abnormal value;

If G≤M(a,n) or G'≤M(a,n), then retain the received signal strength value R _ijN corresponding to the abnormal value, and execute 2e), where N is the number of 30 sets of signal values;

2e) Reorder the retained RSSI values, repeat steps 2a)-2d), until all RSSI values with drastic changes are eliminated, and use the final retained values as the output of the Dixon detection algorithm to form an iBeacon node Dixon matrix R ^* :

in:

R ^* _ij ＝[R ^* _{ij1 R * ij2} ^R _* ^ij3 _... R ^* _ijM ] is a 1×M dimensional matrix, M is the number of signal values retained after the Dixon detection algorithm among the 30 groups of signal values, R ^* _ij represents the signal value matrix retained after the Dixon detection algorithm among the 30 sets of signal values sent by the i-th iBeacon node received by the j-th iBeacon node;

In step 3, the iBeacon node Dixon matrix R ^* is corrected to obtain the iBeacon node signal strength correction matrix R'.

3a) Perform Gaussian filtering on the iBeacon node Dixon matrix R ^* to obtain the iBeacon node Gaussian matrix R";

3b) Take the arithmetic mean value of each element matrix of the Gaussian matrix R" of the iBeacon node, and obtain the received signal strength correction value of each iBeacon node, and form the iBeacon node signal strength correction matrix R':

Where: R _ij ' represents the correction value of the signal strength received by the i-th iBeacon node from the j-th iBeacon node.

Step 4, construct the iBeacon node distance matrix D.

4a) Establish the path loss relationship between the received signal strength correction value and the distance between iBeacon nodes:

in:

R _ij 'indicates that the i-th iBeacon node receives the signal strength correction value of the j-th iBeacon node;

q ₀ represents the reference signal strength correction value when the distance L ₀ =1 meter between two iBeacon nodes;

D _ij represents the distance value from the i-th iBeacon node to the j-th iBeacon node, when i=j, D _ij =0;

x _ε and u are environmental parameters, indicating the degree of influence of the space environment, x _ε is the absolute value of the average power received at a distance of 1 meter from the iBeacon node, u is the path loss factor, and the best reference range of x _ε is 41-47, u The best reference range is 2.15-4.3;

4b) Convert the path loss relational expression of step 4a) into the following form:

4c) According to step 4b), the distance matrix D from each iBeacon node to all other iBeacon nodes is obtained:

Step 5, calculate the estimated coordinate value Q( _xi , y _i ) of each iBeacon node.

5a) Sort the distance values D _ij from the i-th iBeacon node to the j-th iBeacon node in ascending order to obtain the distance set: D _i1 , D _i2 , D _i3 ,...D _ij ,..., D _in , where D _i1 <D _i2 <D _i3 <...<D _ij <...<D _in ;

5b) Subtract the minimum value D _i1 from each element in the distance set to obtain the difference set:

0,ΔD _i1 ,ΔD _i2 ,ΔD _i3 ,...,ΔD _ij ,...,ΔD _in ;

5c) Calculate the average value of each element in the difference set

Among them: ΔD _ij =D _ij -D _i1 , indicating the difference between the distance value D _ij and the minimum value D _i1 between the i-th iBeacon node and the j-th iBeacon node;

5d) Divide iBeacon nodes into The set A of and The set B of the set B, assuming that there are m nodes in the set B, and divide the three iBeacon nodes with the closest difference in the set B into a group, and divide them into m-2 groups in total, calculate the centroid of each group, and its coordinates (x _k , y _k ):

Where: k∈[1,m-2], (x _k1 ,y _k1 ), (x _k2 ,y _k2 ), (x _k3 ,y _k3 ) are the coordinates of 3 iBeacon nodes in each group;

5e) Reconsider the centroid obtained in step 5d) as a newly added iBeacon node, and then form a polygon with m-2 centroid points and nm iBeacon nodes in set A, and use the centroid algorithm to calculate the centroid of this polygon to obtain the i-th iBeacon nodes estimate the coordinates Q( _xi ,y _i ).

Step 6, calculate the error point P(α,β).

Calculate the abscissa and ordinate of the error point P according to the actual coordinates e(v _i , _zi ) and estimated coordinates Q( _xi , y _i ) of the i-th iBeacon node:

get error points

Step 7: Construct the signal strength matrix r of the nodes to be located.

Divide the indoor environment into several areas with regular hexagons, place an iBeacon node at each vertex of the area, place a total of n iBeacon nodes, each iBeacon node periodically broadcasts its own number, its own coordinates and received signal strength values, and collects The node to be positioned receives the received signal strength value sent by the iBeacon node to form the signal strength matrix r of the node to be positioned:

r=[r ₁ r ₂ r ₃ ... r _i ... r _n ]

Among them: r _i is a 1×30 dimensional matrix, indicating that the node to be positioned receives 30 sets of signal values sent by the i-th iBeacon node;

Step 8: Correct the signal strength matrix r of the node to be located to obtain the Dixon matrix r ^* of the node to be located.

8a) Use the Dixon detection method to eliminate the received signal strength values that change drastically among the 30 sets of signal values received by the node to be located from other iBeacon nodes, and receive the 30 sets of signal values from the i-th iBeacon node to the node to be located. The received signal strength values are arranged in ascending order, r _i1 , r _i2 ,...,r _i30 ;

8b) According to the statistical formula of the Dixon test method, calculate the highest outlier g of the node to be located and the lowest outlier g' of the node to be located:

8c) Compare the highest outlier g of the node to be located and the lowest outlier g' of the node to be located with the critical value M(a,n):

If g>M(a,n) or g'>M(a,n), then remove the received signal strength value r _iN corresponding to the abnormal value;

If g≤M(a,n) or g'≤M(a,n), then retain the received signal strength value r _iN corresponding to the abnormal value, and execute 8d), where N is the number of 30 sets of signal values;

8d) Reorder the reserved received signal strength values, and repeat steps 8a)-8c), until all received signal strength values with drastic changes are eliminated, and the final reserved values are used as the output of the Dixon detection algorithm to constitute the location-to-be Node Dixon matrix r ^* :

r ^* = [r ₁ ^* r ₂ ^* r ₃ ^* ... r _i ^* ... r _n ^* ]

Among them: r _i ^* =[r ^* _i1 r ^* _i2 r ^* _i3 ... r ^* _iM ] is a 1×M dimensional matrix, and M is the number of signal values retained after the Dixon detection algorithm among the 30 groups of signal values , r _i ^* represents the signal value matrix retained after the Dixon detection algorithm among the 30 sets of signal values sent by the i-th iBeacon node received by the node to be located.

Step 9: Modify the Dixon matrix r ^* of the node to be located to obtain the signal strength correction matrix r' of the node to be located.

9a) using Gaussian filtering to process the Dixon matrix r ^* of the node to be positioned to obtain the Gaussian matrix r″ of the node to be positioned;

9b) Take the arithmetic mean value of each element matrix of the Gaussian matrix r" of the node to be positioned to obtain the correction value of the received signal strength of the node to be positioned, and form the signal strength correction matrix r' of the node to be positioned:

r'=[r ₁ ' r ₂ ' r ₃ ' ... r _i ' ... r _n ']

Among them: r _i ' represents the correction value of the signal strength received by the i-th iBeacon node received by the node to be positioned.

Step 10, construct the distance matrix d of the nodes to be located.

10a) Establish the path loss relationship between the received signal strength correction value and the distance between the node to be positioned and the i-th iBeacon node:

in:

r _i 'represents the signal strength correction value received by the node to be positioned from the i-th iBeacon node;

q ₁ represents the reference signal strength correction value when the distance between the node to be positioned and the iBeacon node is L ₀ =1 meter;

d _i is the distance value from the node to be positioned to the i-th iBeacon node.

10b) convert the path loss relational expression of step 10a) into the following form:

10c) According to step 10b), obtain the distance matrix d from the node to be positioned to all other iBeacon nodes:

d=[d ₁ d ₂ d ₃ ... d _i ... d _n ]

Step 11, calculate the estimated coordinate value W(x, y) of the node to be located.

11a) Sort the distance value d _i from the node to be positioned to the i-th iBeacon node in ascending order to obtain a set of distances to be positioned: d _ε ={d ₁ ,d ₂ ,d ₃ ,...,d _i ,...,d _n }, where d ₁ <d ₂ <d ₃ <...<d _i <...<d _n ;

11b) Subtract the minimum value d ₁ from each element in the distance set d _ε to be positioned to obtain the difference set d _ε ' to be positioned:

d _ε '={0,Δd ₁ ,Δd ₂ ,Δd ₃ ,...,Δd _i ,...,Δd _n },

Among them: Δd _i = d _i -d ₁ , which represents the difference between the distance value d _i from the node to be positioned to the i-th iBeacon node and the minimum value d ₁ ;

11c) Calculate the average value of each element in the difference set d _ε ' to be positioned

11d) Divide the iBeacon nodes into two sets respectively, namely The set C ₁ of and Set C ₂ , assuming that there are z nodes in the set C ₂ , divide the three iBeacon nodes with the closest difference in the set C ₂ into a group, and divide them into z-2 groups. Calculate the centroid of each group, and its coordinates (x _L ,y _L ) is:

Where: L∈[1,z-2], (x _L1 ,y _L1 ), (x _L2 ,y _L2 ), (x _L3 ,y _L3 ) are the coordinates of the three iBeacon nodes in each group.

11e) Reconsider the centroid obtained in step 11d) as a newly added iBeacon node, and then form a polygon with z-2 centroid points and nz iBeacon nodes in the set C ₁ , and use the centroid algorithm to calculate the centroid of the polygon to obtain Estimated coordinates W(x,y) of bit nodes.

Step 12, calculate the final coordinate W(x+α, y+β) of the node to be positioned.

According to the estimated coordinates W(x, y) of the node to be positioned obtained in step 11 and the abscissa α and ordinate β of the error point P obtained in step 6, the final coordinates of the node to be positioned are calculated as:

In the following, the positioning effect of the present invention will be further analyzed in combination with simulation experiments.

1. Experimental conditions

In this experiment, 16 iBeacon nodes and 10 nodes to be positioned are arranged in an indoor environment of 50*50 meters.

2. Experimental content

Using the method of the present invention and the existing single-centroid algorithm to test the above-mentioned 10 nodes to be positioned respectively, the coordinate values of the nodes to be positioned are obtained, and the result is shown in Fig. 2 .

As can be seen from Fig. 2, the coordinate value of the node to be positioned measured by the present invention is compared with the actual coordinate value of the node to be positioned, the maximum value of the positioning accuracy error is 2.413 meters, the minimum value is 0.75 meters, and the average value is 1.59 meters. Comparing the coordinates of the nodes to be positioned by the existing centroid algorithm with the actual coordinates of the nodes to be positioned, the maximum error of the positioning accuracy is 4.089 meters, the minimum is 1.56 meters, and the average is 2.75 meters.

Experiments show that the present invention improves the positioning accuracy by 25%-35%, and has great advantages in indoor positioning.

Claims (5)

1. A Bluetooth indoor positioning method based on RSSI correction value dual positioning, comprising: (1) Collect the received signal strength values of other iBeacon nodes received by each iBeacon node deployed indoors to form an iBeacon node signal strength matrix R, and use the Dixon detection method and Gaussian filter algorithm to correct it to obtain each iBeacon node The signal strength correction value of the iBeacon node constitutes the signal strength correction matrix R': in: i, j are iBeacon node numbers, n is the number of iBeacon nodes, i∈[1,n], j∈[1,n], n>3; R _ij ＝[R _{ij1 R ij2 R ij3} ... R _ij30 ] is a 1×30-dimensional matrix, indicating that the i-th iBeacon node receives 30 sets of signal values sent by the j-th iBeacon node; R _ij 'indicates that the i-th iBeacon node receives the signal strength correction value of the j-th iBeacon node; (2) According to the iBeacon node signal strength correction matrix R', using the logarithmic distance path loss model, the distance value from the iBeacon node to other iBeacon nodes is obtained to form the iBeacon node distance matrix D: Wherein: D _ij represents the distance value between the i-th iBeacon node and the j-th iBeacon node, when i=j, D _ij =0; (3) According to the iBeacon node distance matrix D, the multi-centroid algorithm is used to obtain the estimated coordinate value Q _{(xi ,} y _{i )} of each iBeacon node, and the actual coordinates e( v _i , z _i ) to get the error point P(α,β), follow the steps below: 3.1) Estimate the estimated coordinate value Q( _xi ,y _i ) of each iBeacon node: 3.1a) Sort the distance value D _ij from the i-th iBeacon node to the j-th iBeacon node in ascending order to obtain the distance set: D _i1 , D _i2 , D _i3 ,...D _ij ,... ,D _in , where D _i1 <D _i2 <D _i3 <...<D _ij <...<D _in ; 3.1b) Subtract the minimum value D _i1 from each element in the distance set to obtain the difference set: 0,ΔD _i1 ,ΔD _i2 ,ΔD _i3 ,...,ΔD _ij ,...,ΔD _in 3.1c) Calculate the average value of each element in the difference set Among them: ΔD _ij =D _ij -D _i1 , indicating the difference between the distance value D _ij and the minimum value D _i1 between the i-th iBeacon node and the j-th iBeacon node; 3.1d) Divide iBeacon nodes into and There are two sets A and B of the set B, assuming that there are m nodes in the set B, divide the three iBeacon nodes with the closest difference in the set B into a group, and divide them into m-2 groups, calculate the centroid of each group, and its coordinates ( x _k ,y _k ): Where: k∈[1,m-2], (x _k1 ,y _k1 ), (x _k2 ,y _k2 ), (x _k3 ,y _k3 ) are the coordinates of 3 iBeacon nodes in each group; 3.1e) Reconsider the centroid obtained in step 3.1d) as a newly added iBeacon node, and then form a polygon with m-2 centroid points and nm iBeacon nodes in set A, and use the centroid algorithm to calculate the centroid of this polygon, and get The i-th iBeacon node estimates coordinates Q( _xi , y _i ); 3.2) Calculate the abscissa and ordinate of the error point P according to the actual coordinates e(v _i , z _i ) and estimated coordinates Q( _xi , y _i ) of the i-th iBeacon node: (4) Collect the received signal strength value sent by the node to be positioned to receive the iBeacon node to form the signal strength matrix r of the positioned node, and use the Dixon detection method and Gaussian filter algorithm to correct it to obtain the strength correction value of the node to be positioned. Constitute the signal strength correction matrix r' of the node to be located: r=[r ₁ r ₂ r ₃ ... r _i ... r _n ], r'=[r ₁ ' r ₂ ' r ₃ ' ... r _i ' ... r _n '] in: r _i is a 1×30-dimensional matrix, indicating that the node to be positioned receives 30 sets of signal values sent by the i-th iBeacon node; r _i 'represents the signal strength correction value received by the node to be positioned from the i-th iBeacon node; (5) According to the signal strength correction matrix r' of the node to be located, the logarithmic distance path loss model is used to obtain the distance value from the node to be located to other iBeacon nodes, and the distance matrix d of the node to be located is formed: d=[d ₁ d ₂ d ₃ ... d _i ... d _n ] Among them: d _i is the distance value from the node to be positioned to the i-th iBeacon node; (6) According to the distance matrix d of the node to be located, the multi-centroid algorithm is used to obtain the estimated coordinate value W(x, y) of the node to be located, and the steps are as follows: 6.1) Sort the distance value d _i from the node to be positioned to the i-th iBeacon node in ascending order to obtain a set of distances to be positioned: d _ε ={d ₁ ,d ₂ ,d ₃ ,...,d _i ,...,d _n }, where d ₁ <d ₂ <d ₃ <...<d _i <...<d _n ; 6.2) Subtract the minimum value d ₁ from each element in the distance set d _ε to be positioned to obtain the difference set d _ε ' to be positioned: d _ε '={0,Δd ₁ ,Δd ₂ ,Δd ₃ ,...,Δd _i ,...,Δd _n }, Among them: Δd _i = d _i -d ₁ , which represents the difference between the distance value d _i and the minimum value d ₁ between the node to be positioned and the i-th iBeacon node; 6.3) Calculate the average value of each element in the difference set d _ε ' to be positioned 6.4) Divide iBeacon nodes into The set C ₁ of and Set C ₂ , assuming that there are z nodes in the set C ₂ , divide the three iBeacon nodes with the closest difference in the set C ₂ into a group, and divide them into z-2 groups. Calculate the centroid of each group, and its coordinates (x _L ,y _L ) is: Where: L∈[1,z-2], (x _L1 ,y _L1 ), (x _L2 ,y _L2 ), (x _L3 ,y _L3 ) are the coordinates of the three iBeacon nodes in each group; 6.5) Reconsider the centroid obtained in step 6.4) as a newly added iBeacon node, and then form a polygon with z-2 centroid points and nz iBeacon nodes in the set C ₁ , use the centroid algorithm to obtain the centroid of the polygon, and calculate The estimated coordinate W(x,y) of the node to be located; (7) According to the estimated coordinate value W(x, y) and the error point P(α, β), the final coordinate of the node to be positioned is: W(x+α, y+β). 2. The method according to claim 1, wherein in the step (1), utilize Dixon detection method and Gaussian filter algorithm to correct the iBeacon node signal strength matrix R, according to the following steps: 1.1) Divide the indoor environment into several areas with regular hexagons, place an iBeacon node at each vertex of the area, and place n iBeacon nodes in total, and each iBeacon node periodically broadcasts its own number, its own coordinates and received signal strength values ; 1.2) Use the Dixon detection method to eliminate the received signal strength values that each iBeacon node receives from among the 30 sets of signal values sent by other iBeacon nodes that change drastically: 1.2a) Arrange the 30 sets of received signal strength values received by the i-th iBeacon node from the j-th iBeacon node in ascending order, followed by R _{ij1 , R ij2 ,} ...,R _ij30 ; 1.2b) Determine the detection level of outlier detection as a=0.02, and determine the Dixon test critical value M(a,n); 1.2c) According to the statistical formula of the Dixon test method, calculate the abnormal value G of the highest end of the iBeacon node and the abnormal value G' of the lowest end of the iBeacon node: 1.2d) Compare the highest-end outlier G of the iBeacon node and the lowest-end outlier G' of the iBeacon node with the critical value M(a,n): If G>M(a,n) or G'>M(a,n), remove the received signal strength value R _ijN corresponding to the abnormal value; If G≤M(a,n) or G'≤M(a,n), then retain the received signal strength value R _ijN corresponding to the abnormal value, and execute 1.2e), N is the number of 30 sets of signal values; 1.2e) Reorder the retained received signal strength values, repeat steps 1.2a)-1.2d), until all received signal strength values with drastic changes are eliminated, and use the final retained value as the output of the Dixon detection algorithm, Constitute the iBeacon node Dixon matrix R ^* : Among them: R ^* _ij = [R ^* _{ij1 R * ij2} ^R _* ^ij3 _... R ^* _ijM ] is a 1×M dimensional matrix, and M is the number of signal values retained after the Dixon detection algorithm among the 30 groups of signal values , R ^* _ij represents the signal value matrix retained after the Dixon detection algorithm among the 30 sets of signal values sent by the i-th iBeacon node received by the j-th iBeacon node; 1.3) Perform Gaussian filtering on the Dixon matrix R ^* of the iBeacon node to obtain the Gaussian matrix R" of the iBeacon node, and then take the arithmetic mean value of each element of the Gaussian matrix R" of the iBeacon node to obtain the received signal strength correction value of each iBeacon node , forming the iBeacon node signal strength correction matrix: 3. method according to claim 1, wherein in the step (2), calculate the distance value D _ij between the i-th iBeacon node to the j-th iBeacon node, carry out according to the following steps: 2.1) Establish the path loss relationship between the received signal strength value and the distance: in: R _ij 'indicates that the i-th iBeacon node receives the signal strength correction value of the j-th iBeacon node; q ₀ represents the reference signal strength correction value when the distance L ₀ =1 meter between two iBeacon nodes; x _ε and u are environmental parameters, indicating the degree of influence of the space environment, x _ε is the absolute value of the average power received at a distance of 1 meter from the iBeacon node, u is the path loss factor, and the best reference range of x _ε is 41-47, u The best reference range is 2.15-4.3; 2.2) the path loss relation of step 2.1) is converted into the following form: 2.3) According to step 2.2), the distance matrix D from each iBeacon node to all other iBeacon nodes is obtained: 4. method according to claim 1, utilize Dixon detection method and Gaussian filter algorithm to revise in the step (4) node signal strength matrix r to be positioned, carry out according to the following steps: 4.1) Divide the indoor environment into several areas with regular hexagons, place an iBeacon node at each vertex of the area, place a total of n iBeacon nodes, and each iBeacon node periodically broadcasts its own number, its own coordinates and received signal strength values ; 4.2) Use the Dixon detection method to eliminate the received signal strength values that change drastically among the 30 sets of signal values received by the node to be located from other iBeacon nodes: 4.2a) Arrange the 30 sets of received signal strength values received by the node to be positioned from the i-th iBeacon node in ascending order, which are r _i1 , r _i2 ,...,r _i30 ; 4.2b) According to the statistical formula of the Dixon test method, calculate the highest outlier g of the node to be located and the lowest outlier g' of the node to be located: 4.2c) Compare the highest outlier g of the node to be located and the lowest outlier g' of the node to be located with the critical value M(a,n): If g>M(a,n) or g'>M(a,n), remove the received signal strength value r _iN corresponding to the abnormal value; If g≤M(a,n) or g'≤M(a,n), then retain the received signal strength value r _iN corresponding to the abnormal value, and execute 4.2d), where N is the number of 30 sets of signal values; 4.2d) Reorder the retained RSSI values, repeat steps 4.2a)-4.2c), until all RSSI values with drastic changes are eliminated, and use the final retained values as the output of the Dixon detection algorithm, Constitute the Dixon matrix r ^* of the node to be located: r ^* = [r ₁ ^* r ₂ ^* r ₃ ^* ... r _i ^* ... r _n ^* ] Among them: r _i ^* =[r ^* _i1 r ^* _i2 r ^* _i3 ... r ^* _iM ] is a 1×M dimensional matrix, and M is the number of signal values retained after the Dixon detection algorithm among the 30 groups of signal values , r _i ^* represents the signal value matrix retained by the Dixon detection algorithm among the 30 sets of signal values sent by the i-th iBeacon node received by the node to be located; 4.3) Perform Gaussian filtering on the Dixon matrix r _i ^* of the node to be positioned to obtain the Gaussian matrix r" of the node to be positioned, and then take the arithmetic mean value of each element matrix of the Gaussian matrix r" of the node to be positioned to obtain the received signal of the node to be positioned The strength correction value constitutes the signal strength correction matrix of the node to be located: r'=[r ₁ ' r ₂ ' r ₃ ' ... r _i ' ... r _n ']. 5. The method according to claim 1, wherein in the step (5), the distance value d _i between the node to be positioned and the i-th iBeacon node is calculated according to the following steps: 5.1) Establish the path loss relationship between the received signal strength value and the distance: in: r _i 'indicates that the node to be positioned receives the signal strength correction value of the i-th iBeacon node; q ₁ represents the reference signal strength correction value when the distance between the node to be positioned and the iBeacon node is L ₀ =1 meter; x _ε and u are environmental parameters, indicating the degree of influence of the space environment, x _ε is the absolute value of the average power received at a distance of 1 meter from the iBeacon node, u is the path loss factor, and the best reference range of x _ε is 41-47, u The best reference range is 2.15-4.3; 5.2) the path loss relation of step 5.1) is converted into the following form: 5.3) According to step 5.2), obtain the distance matrix d from the node to be positioned to all other iBeacon nodes: d=[d ₁ d ₂ d ₃ ... d _i ... d _n ].