CN111901749A - High-precision three-dimensional indoor positioning method based on multi-source fusion - Google Patents
High-precision three-dimensional indoor positioning method based on multi-source fusion Download PDFInfo
- Publication number
- CN111901749A CN111901749A CN202010890275.2A CN202010890275A CN111901749A CN 111901749 A CN111901749 A CN 111901749A CN 202010890275 A CN202010890275 A CN 202010890275A CN 111901749 A CN111901749 A CN 111901749A
- Authority
- CN
- China
- Prior art keywords
- positioning
- bluetooth
- dimensional
- value
- coordinate
- 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
- 230000004927 fusion Effects 0.000 title claims abstract description 48
- 238000000034 method Methods 0.000 title claims abstract description 32
- 230000001133 acceleration Effects 0.000 claims abstract description 19
- 238000012549 training Methods 0.000 claims description 16
- 238000004364 calculation method Methods 0.000 claims description 10
- 238000005259 measurement Methods 0.000 claims description 10
- 238000001914 filtration Methods 0.000 claims description 6
- 230000002146 bilateral effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 claims description 3
- 238000009499 grossing Methods 0.000 claims description 3
- 238000012545 processing Methods 0.000 claims description 3
- 230000001902 propagating effect Effects 0.000 claims description 3
- 230000007704 transition Effects 0.000 claims description 3
- 238000005516 engineering process Methods 0.000 description 5
- 230000000875 corresponding effect Effects 0.000 description 3
- 238000011161 development Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000002474 experimental method Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000012423 maintenance Methods 0.000 description 1
- 230000000452 restraining effect Effects 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/30—Services specially adapted for particular environments, situations or purposes
- H04W4/33—Services specially adapted for particular environments, situations or purposes for indoor environments, e.g. buildings
-
- 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
- H04W4/025—Services making use of location information using location based information parameters
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
- H04W4/80—Services using short range communication, e.g. near-field communication [NFC], radio-frequency identification [RFID] or low energy communication
-
- 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
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04W—WIRELESS COMMUNICATION NETWORKS
- H04W84/00—Network topologies
- H04W84/02—Hierarchically pre-organised networks, e.g. paging networks, cellular networks, WLAN [Wireless Local Area Network] or WLL [Wireless Local Loop]
- H04W84/10—Small scale networks; Flat hierarchical networks
- H04W84/12—WLAN [Wireless Local Area Networks]
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
Abstract
The invention discloses a high-precision three-dimensional indoor positioning method based on multi-source fusion, which comprises the steps of deploying a Wi-Fi environment in an area to be positioned, collecting a Wi-Fi signal strength value RSSI (received signal strength indicator), and establishing a Wi-Fi offline fingerprint database; collecting RSSI (received signal strength indicator) values of Wi-Fi signals at an online stage, and realizing Wi-Fi positioning by improving a weighted centroid positioning algorithm; deploying Bluetooth equipment in an area to be positioned, collecting a Bluetooth RSSI value, and realizing Bluetooth positioning by improving a weighted centroid positioning algorithm; realizing Wi-Fi and Bluetooth fusion positioning through average weighting; matching the result of the acceleration sensor by using a step size model, and performing PDR positioning; fusing Wi-Fi, Bluetooth and PDR through a UKF algorithm to obtain a two-dimensional positioning coordinate; utilizing a barometer to measure the height to judge the floor to obtain a height value in the vertical direction; and fusing the two-dimensional positioning coordinate and the height value in the vertical direction to obtain a three-dimensional position coordinate. The method not only can accurately position in the horizontal direction, but also has higher positioning accuracy in the vertical direction, and can carry out accurate three-dimensional position service on users.
Description
Technical Field
The invention relates to the technical field of indoor positioning, in particular to a high-precision three-dimensional indoor positioning method based on multi-source fusion.
Background
With the rapid development of the position service industry, the demand of people on position service is increased, the current mature GPS and Beidou positioning only support outdoor positioning, and due to the complexity of indoor environment, the indoor positioning needs to be improved in the aspects of implementation cost, positioning complexity, positioning precision and the like. The mainstream indoor positioning technologies include Wi-Fi positioning, bluetooth positioning, PDR positioning, ultra-wideband positioning, geomagnetic positioning and the like, and various positioning technologies have different advantages and disadvantages due to different positioning means, and table 1 shows the advantages and disadvantages of several mainstream indoor positioning technologies.
TABLE 1 comparison of different indoor positioning techniques
The Wi-Fi positioning is mainly realized by constructing a Wi-Fi fingerprint database, but the Wi-Fi positioning is not stable because Wi-Fi signals are unstable. Bluetooth Beacon broadcasts the ID of itself to around through the bluetooth, can take corresponding action after the cell-phone terminal obtains near Beacon's ID, if obtain the positional information that this ID corresponds etc. from the high in the clouds server, the terminal measures the received signal intensity of its position to this estimation and the distance between Beacon. PDR positioning is to position and track a target by means of pedestrian track calculation by utilizing a measuring unit formed by a gyroscope, an accelerometer, a direction sensor and the like arranged in a mobile phone, but the accumulated error in PDR positioning is large. The fusion positioning refers to the integration of multiple positioning technologies and information of multiple sensors for comprehensive positioning, so as to achieve the purpose of mutual complementation, improve the positioning accuracy and robustness and reduce the positioning cost. Regardless of Wi-Fi, Bluetooth Beacon or PDR, because indoor and outdoor environments are different, indoor environments are relatively complex, and characteristics of different places are completely different, at present, a single technology cannot meet requirements of precision, deployment and cost at the same time, and fusion positioning is bound to become a development direction of indoor positioning.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provides a high-precision three-dimensional indoor positioning method based on multi-source fusion.
The technical scheme for realizing the purpose of the invention is as follows:
a high-precision three-dimensional indoor positioning method based on multi-source fusion comprises the following steps:
1) deploying a Wi-Fi environment in an area to be positioned, and establishing a Wi-Fi offline fingerprint database in an offline stage by acquiring a Wi-Fi signal strength value RSSI of each small area in the area to be positioned;
2) collecting RSSI (received signal strength indicator) values of Wi-Fi signals at an online stage, and realizing Wi-Fi positioning by improving a weighted centroid positioning algorithm;
3) deploying Bluetooth equipment in an area to be positioned, collecting a Bluetooth RSSI value, and realizing Bluetooth positioning by improving a weighted centroid positioning algorithm;
4) realizing Wi-Fi and Bluetooth fusion positioning through average weighting;
5) matching the result of the acceleration sensor by using a step size model, and performing PDR positioning;
6) fusing Wi-Fi, Bluetooth and PDR through a UKF algorithm to obtain a two-dimensional positioning coordinate;
7) utilizing a barometer to measure the height to judge the floor to obtain a height value in the vertical direction;
8) and fusing the two-dimensional positioning coordinate and the vertical direction height value to obtain a final three-dimensional position coordinate, thereby realizing three-dimensional positioning of personnel in the building.
In step 1), the off-line stage is to divide a plurality of positions in the whole indoor scene and collect enough information at each positionCarrying out specific training on the Wi-Fi RSSI samples to obtain training results and coordinate information of the current position, storing the training results and the coordinate information of the current position in a database as training data, and constructing a Wi-Fi fingerprint spectrum after all training points are finished; the Wi-Fi signal acquisition for a plurality of times in a certain small area divided in an area to be positioned in a period of time t is as follows:
where p is the number of signal strengths acquired during the period of time t, rtqThe Wi-Fi signal strength of the p-th acquisition in t time is represented, and the constructed Wi-Fi offline fingerprint database is as follows:
I＝{(RSSV1,ο1),(RSSV2,ο2),...,(RSSVi,οi),...,(RSSVN,οN)}
wherein RSSVi＝(RSSi1,RSSi2,...,RSSiM) Is RSSI, from M Wi-Fi hotspotsi＝(x,y)∈R2Is RSSViThe corresponding position, x, y is the position information of the position.
In step 2), the online stage is that after the Wi-Fi offline fingerprint database is established, euclidean distances between all samples in the Wifi database are calculated, and an expression of the euclidean distances is as follows:
wherein DijRepresents the actual distance of sample i from sample j;
according to DijClustering, wherein each sample point is a class initially, two classes with the minimum distance are merged into a new class each time, a threshold value T is set, when the distance D between the two classes with the minimum distance is larger than T, clustering is finished, and the signal characteristic vector of the centroid of each class is calculated as
Let RSSI sample obtained online be denoted as T ═ RSSj1,RSSj2,...,RSSjM),(xj,yj) Wherein r) isj＝(RSSj1,RSSj2,...,RSSjM) RSSI representing M APs received online, a location vector o ═ x, y representing location information of RSSI obtained online, where the RSSI obtained online is known and the location vector (x, y) is unknown;
calculating the RSSI sample T obtained on line and the mass center M of all classesKFinding out the class P of the centroid point closest to the fingerprint distance of T, and calculating the K points closest to T in the Euclidean distance by the nearest neighbor algorithm, wherein the position information of the K points is (x)i,yi) If the number of the fingerprint points in P is less than K, extracting the physical positions of all the points in the class P, and determining the distribution of coordinate weights according to Euclidean distances between the fingerprints of the T and the K points, wherein the calculation formula of the weights is as follows:
w 'of'sjIs a function of the transition of the weight, wsjThe coordinate weight value distributed to the sample j is K, and K represents K points which are closest to S in Euclidean distance in P;
the Wi-Fi positioning result is obtained by the following formula,
in step 3), the bluetooth positioning is realized by improving the weighted centroid positioning algorithm, namely, bluetooth equipment is deployed in an area to be positioned, bluetooth RSSI values are collected, and m collection points (x) are subjected to the Bluetooth positioning1,y1)，(x2,y2)，...,(xm,ym)，Beacon
The signal intensity measured for m positions is S1,S2,...SmIs provided with
Then the weight is
The final positioning result is obtained as
In the step 4), the Wi-Fi and Bluetooth fusion positioning is realized by average weighting, namely 12 Wi-Fi routers are deployed in an area to be positioned to deploy a Wi-Fi positioning environment, the area to be positioned is divided into a plurality of small grids of 2 x 2m, Wi-Fi signal intensity values are collected every 2s for 10 times in total, the position information of a collection point is recorded, and a Wi-Fi offline fingerprint database is constructed by a mean value filtering method;
meanwhile, 12 Bluetooth nodes for transmitting Bluetooth signals are deployed in an area to be positioned, as the Bluetooth output frequency is higher than the Wi-Fi output frequency, when a Wi-Fi positioning result is usually obtained once, the Bluetooth positioning result is obtained for 3-5 times, and the Wi-Fi and the Bluetooth are fused through average weighting, namely the Wi-Fi positioning result in certain positioning is as follows:
LWifi＝(x,y)
during this time period, the bluetooth positioning result is:
Lbeacon(x,y)＝{(x1,y1),(x2,y2),...,(xn,yn)}
the average weighted coordinates for bluetooth positioning are:
at the moment, a distance threshold value sigma s judgment is added, and the obtained Wi-Fi positioning coordinate L is obtainedWifiAnd bluetooth weighted coordinates
Comparing the distance d with a distance threshold value sigma s, and determining the self-adaption of the positioning weight values of the distance d and the distance threshold value sigma s; when d is less than or equal to σ s, namely the positioning results of the two are close, the positioning results of the Bluetooth and the Wi-Fi are both in a normal range, when d is greater than σ s, namely the positioning results of the two are greatly different, the possible positioning error is larger, and the self-adaptive weight rule is as follows:
Lwifibeaconnamely the final positioning coordinate after the Wi-Fi and the Bluetooth are fused, and the distance threshold value sigma s is determined according to the environment and the positioning error of the Wi-Fi and the Bluetooth.
In step 5), the step length model is used for matching the result of the acceleration sensor to perform PDR positioning, and a system model for PDR positioning is as follows:
where K is the number of steps, positional information x after walkingk,ykFor positional information after walking, θkDenotes the azimuth after K steps, Wk-1In order to be a noise, the noise is,
the step length model is used for matching the result of the acceleration sensor for the step length average value,
is the orientation angle variation; the simulated measurement equation is as follows:
wherein x isk,ykRepresenting the pedestrian position obtained by Wi-Fi and Bluetooth fusion positioning; skRepresents the k-th step walking step length of the pedestrian, obtained by the result of the acceleration sensor, delta thetakThe orientation angle variation of the kth step after the pedestrian walks can be obtained through a built-in gyroscope of the intelligent terminal, and thetakRepresenting the orientation angle of the k step after the pedestrian walks; vkRepresenting noise.
In the step 6), the Wi-Fi, the Bluetooth and the PDR are fused through a UKF algorithm to obtain a final positioning result, specifically, a pedestrian walking system model is constructed, initial position coordinates of PDR positioning are obtained through Wi-Fi positioning, and an error threshold value is obtained through fusion of Wi-Fi positioning and PDR positioning; the position of the pedestrian after one step can be obtained through a built-in sensor, a gyroscope and a direction sensor of the intelligent terminal, and state information and position information are updated by using a measurement equation;
the UKF algorithm sets the system equations and the bilateral equations to have discrete forms, i.e.
Wherein X is an n-dimensional random vector and
z is a random observation vector of m dimensions, f and h are nonlinear vector functions, WkAnd VVkZero mean white noise sequences which are mutually uncorrelated; u. ofk-1A deterministic control item; z is obtained by propagating X through a nonlinear function f (·), and the statistical characteristic of Z is
According to
Designing a series of points xii(i ═ 1, 2., L), called Sigma point, and calculated by f (-) propagation to give γi(i ═ 1, 2.., L), then based on γiComputing
Usually the number of Sigma dots is reduced by 2n +1
Modeling a system model for positioning the pedestrian walking by the PDR as follows:
wherein the number of steps of walking is represented by K, and the position information after walking is represented by xk,ykIs represented by thetakDenotes the azimuth angle after K steps, Wk-1The representation of the noise is represented by,
the step length model is adopted to match the result of the acceleration sensor for the step length average value,
is the orientation angle variation; the simulated measurement equation is as follows:
wherein x isk,ykRepresenting the pedestrian position obtained by Wi-Fi and Bluetooth fusion positioning; skRepresenting the walking step length of the kth step of the pedestrian, and obtaining the result through an acceleration sensor; delta thetakRepresenting the orientation angle variation of the k step after the pedestrian walks, and acquiring the orientation angle variation through a built-in gyroscope of the intelligent terminal; thetakRepresenting the orientation angle of the k step after the pedestrian walks; vkRepresenting noise, and obtaining a final positioning result through UKF algorithm fusion, wherein the UKF filter initial value (x)0,y0) Given by Wi-Fi Bluetooth fusion positioning result, initial azimuth orientation angle theta0And (4) given by a PDR processing module, carrying out fusion positioning by using a UKF algorithm to obtain a two-dimensional positioning coordinate.
In step 7), the floor is determined by using the barometer to obtain a height value in the vertical direction, specifically: the method comprises the steps of obtaining a measured atmospheric pressure value by establishing an atmospheric pressure elevation model, and calculating the atmospheric pressure value through the atmospheric pressure elevation model to obtain an elevation value, wherein the calculation formula is as follows:
in the above formula, h is the height of the target to be positioned, h0Is the height of reference point, p0Is the reference point air pressure, p is the air pressure at the location of the target to be positioned, tmIs two equal pressure surfaces p0And average temperature in degrees Celsius between p, t0Is the reference point temperature, t is the temperature of the position where the target to be positioned is located;
reference point pressure p0The air pressure p of the position of the target to be positioned and the reference point temperature t0And measuring the temperature t of the position of the target to be positioned by adopting a base station placed indoors, smoothing the measured data by adopting a filtering method, and then substituting the smoothed data into a formula to carry out height calculation and floor height judgment.
In the step 8), the two-dimensional positioning coordinate and the vertical height value are fused to obtain a final three-dimensional position coordinate, so that three-dimensional positioning of the personnel in the building is realized, and the three-dimensional position coordinate is obtained by fusing the two-dimensional position coordinate obtained in the step 6) and the vertical height h obtained in the step 7), so that an indoor three-dimensional positioning effect is achieved.
The invention provides a high-precision three-dimensional indoor positioning method based on multi-source fusion, which comprises the steps of firstly realizing Wi-Fi positioning and Bluetooth positioning by improving a weighted centroid method, carrying out Wi-Fi and Bluetooth fusion positioning by average weighting, and restraining a fusion positioning result in a weight self-adaptive manner, so that the problem of unstable Wi-Fi signals is solved, realizing fusion positioning by fusion positioning result and PDR positioning fusion through UKF, solving the problem of large accumulated error in PDR positioning, and finally measuring height by using a barometer to judge height, so that indoor three-dimensional positioning is realized. Experiments prove that the Wi-Fi, Bluetooth and PDR fusion positioning result is higher than the positioning precision of single positioning, and the problems of unstable Wi-Fi positioning signals and large accumulated PDR errors are solved. The method can be used for positioning without additional equipment deployment, is less influenced by the environment, is relatively stable, and has lower operation and maintenance cost and better popularization prospect.
Drawings
FIG. 1 is an overall flow diagram of the present invention;
FIG. 2 is a Wi-Fi distribution map of a region to be located;
FIG. 3 is a flow chart of Wi-Fi fingerprint location;
FIG. 4 is a diagram of Bluetooth tag distribution in an area to be located;
fig. 5 shows a method for determining a high floor by barometer measurement.
Detailed Description
The invention is further illustrated but not limited by the following figures and examples.
Example (b):
a high-precision three-dimensional indoor positioning method based on multi-source fusion is disclosed, as shown in FIG. 1, and comprises the following steps:
1) deploying a Wi-Fi environment in an area to be positioned, and establishing a Wi-Fi offline fingerprint database in an offline stage by acquiring a Wi-Fi signal strength value RSSI of each small area in the area to be positioned;
2) collecting RSSI (received signal strength indicator) values of Wi-Fi signals at an online stage, and realizing Wi-Fi positioning by improving a weighted centroid positioning algorithm;
3) deploying Bluetooth equipment in an area to be positioned, collecting a Bluetooth RSSI value, and realizing Bluetooth positioning by improving a weighted centroid positioning algorithm;
4) realizing Wi-Fi and Bluetooth fusion positioning through average weighting;
5) matching the result of the acceleration sensor by using a step size model, and performing PDR positioning;
6) fusing Wi-Fi, Bluetooth and PDR through a UKF algorithm to obtain a two-dimensional positioning coordinate;
7) utilizing a barometer to measure the height to judge the floor to obtain a height value in the vertical direction;
8) and fusing the two-dimensional positioning coordinate and the vertical direction height value to obtain a final three-dimensional position coordinate, thereby realizing three-dimensional positioning of personnel in the building.
In step 1), the off-line stage is to divide a plurality of positions in the whole indoor scene, as shown in fig. 2, collect enough Wi-Fi RSSI samples at each position, perform specific training on the signal samples to obtain a training result and coordinate information of the current position, and store the training result and the coordinate information of the current position in the training result and the coordinate information of the current positionThe database is used as training data, and when all training points are finished, a Wi-Fi fingerprint is constructed; the Wi-Fi signal acquisition for a plurality of times in a certain small area divided in an area to be positioned in a period of time t is as follows:
where p is the number of signal strengths acquired during the period of time t, rtqThe Wi-Fi signal strength of the p-th acquisition in t time is represented, and the constructed Wi-Fi offline fingerprint database is as follows:
I＝{(RSSV1,ο1),(RSSV2,ο2),...,(RSSVi,οi),...,(RSSVN,οN)}
wherein RSSVi＝(RSSi1,RSSi2,...,RSSiM) Is RSSI, from M Wi-Fi hotspotsi＝(x,y)∈R2Is RSSViThe corresponding position, x, y is the position information of the position.
In step 2), the online stage, as shown in fig. 3, is to calculate the euclidean distance between every two samples in the Wifi database after the Wi-Fi offline fingerprint database is established, where the euclidean distance expression is:
wherein DijRepresents the actual distance of sample i from sample j;
according to DijClustering, wherein each sample point is a class initially, two classes with the minimum distance are merged into a new class each time, a threshold value T is set, when the distance D between the two classes with the minimum distance is larger than T, clustering is finished, and the signal characteristic vector of the centroid of each class is calculated as
Let RSSI sample obtained online be denoted as T ═ RSSj1,RSSj2,...,RSSjM),(xj,yj))Wherein r isj＝(RSSj1,RSSj2,...,RSSjM) RSSI representing M APs received online, a location vector o ═ x, y representing location information of RSSI obtained online, where the RSSI obtained online is known and the location vector (x, y) is unknown;
calculating the RSSI sample T obtained on line and the mass center M of all classesKFinding out the class P of the centroid point closest to the fingerprint distance of T, and calculating the K points closest to T in the Euclidean distance by the nearest neighbor algorithm, wherein the position information of the K points is (x)i,yi) If the number of the fingerprint points in P is less than K, extracting the physical positions of all the points in the class P, and determining the distribution of coordinate weights according to Euclidean distances between the fingerprints of the T and the K points, wherein the calculation formula of the weights is as follows:
w 'of'sjIs a function of the transition of the weight, wsjThe coordinate weight value distributed to the sample j is K, and K represents K points which are closest to S in Euclidean distance in P;
the Wi-Fi positioning result is obtained by the following formula,
in step 3), the bluetooth positioning is realized by improving the weighted centroid positioning algorithm, that is, bluetooth devices are deployed in an area to be positioned, as shown in fig. 4, bluetooth RSSI values are collected, and for m collection points (x)1,y1)，(x2,y2)，...,(xm,ym)，Beacon
The signal intensity measured for m positions is S1,S2,...SmIs provided with
Then the weight is
The final positioning result is obtained as
In the step 4), the Wi-Fi and Bluetooth fusion positioning is realized by average weighting, namely 12 Wi-Fi routers are deployed in an area to be positioned to deploy a Wi-Fi positioning environment, the area to be positioned is divided into a plurality of small grids of 2 x 2m, Wi-Fi signal intensity values are collected every 2s for 10 times in total, the position information of a collection point is recorded, and a Wi-Fi offline fingerprint database is constructed by a mean value filtering method;
meanwhile, 12 Bluetooth nodes for transmitting Bluetooth signals are deployed in an area to be positioned, and as the Bluetooth output frequency is higher than the Wi-Fi output frequency, namely when a Wi-Fi positioning result is usually obtained once, the Bluetooth positioning result is obtained for 3-5 times, in actual positioning, the Wi-Fi and the Bluetooth are fused through average weighting, namely the Wi-Fi positioning result in certain positioning is as follows:
LWifi＝(x,y)
during this time period, the bluetooth positioning result is:
Lbeacon(x,y)＝{(x1,y1),(x2,y2),...,(xn,yn)}
the average weighted coordinates for bluetooth positioning are:
at the moment, a distance threshold value sigma s judgment is added, and the obtained Wi-Fi positioning coordinate L is obtainedWifiWith bluetooth weighted coordinates LbeaconComparing the distance d of (x, y) with a distance threshold value sigma s, and determining the positioning weight value self-adaption of the distance d and the distance threshold value sigma s; when d is less than or equal to sigma s, namely the positioning results of the two are close, the Bluetooth is shownAnd Wi-Fi positioning results are in a normal range, when d is larger than sigma s, the positioning results of the two are larger in difference, and possibly the positioning error is larger, and the self-adaptive weight rule is as follows:
Lwifibeaconnamely the final positioning coordinate after the Wi-Fi and the Bluetooth are fused, and the distance threshold value sigma s is determined according to the environment and the positioning error of the Wi-Fi and the Bluetooth.
In step 5), the step length model is used for matching the result of the acceleration sensor to perform PDR positioning, and a system model for PDR positioning is as follows:
where K is the number of steps, positional information x after walkingk,ykFor positional information after walking, θkDenotes the azimuth after K steps, Wk-1In order to be a noise, the noise is,
the step length model is used for matching the result of the acceleration sensor for the step length average value,
is the orientation angle variation; the simulated measurement equation is as follows:
wherein x isk,ykRepresenting the pedestrian position obtained by Wi-Fi and Bluetooth fusion positioning; skRepresents the k-th step walking step length of the pedestrian, obtained by the result of the acceleration sensor, delta thetakThe orientation angle variation of the kth step after the pedestrian walks can be obtained through a built-in gyroscope of the intelligent terminal, and thetakRepresenting the orientation angle of the k step after the pedestrian walks; vkRepresenting noise.
In the step 6), the Wi-Fi, the Bluetooth and the PDR are fused through a UKF algorithm to obtain a final positioning result, specifically, a pedestrian walking system model is constructed, initial position coordinates of PDR positioning are obtained through Wi-Fi positioning, and an error threshold value is obtained through fusion of Wi-Fi positioning and PDR positioning; the position of the pedestrian after one step can be obtained through a built-in sensor, a gyroscope and a direction sensor of the intelligent terminal, and state information and position information are updated by using a measurement equation;
the UKF algorithm sets the system equations and the bilateral equations to have discrete forms, i.e.
Wherein X is an n-dimensional random vector and
z is a random observation vector of m dimensions, f and h are nonlinear vector functions, WkAnd VVkZero mean white noise sequences which are mutually uncorrelated; u. ofk-1A deterministic control item; z is obtained by propagating X through a nonlinear function f (·), and the statistical characteristic of Z is
According to
Designing a series of points xii(i ═ 1, 2., L), called Sigma point, and calculated by f (-) propagation to give γi(i ═ 1, 2.., L), then based on γiComputing
Usually the number of Sigma dots is reduced by 2n +1
Modeling a system model for positioning the pedestrian walking by the PDR as follows:
wherein the number of steps taken isK denotes the position information after walking as xk,ykIs represented by thetakDenotes the azimuth angle after K steps, Wk-1The representation of the noise is represented by,
the step length model is adopted to match the result of the acceleration sensor for the step length average value,
is the orientation angle variation; the simulated measurement equation is as follows:
wherein x isk,ykRepresenting the pedestrian position obtained by Wi-Fi and Bluetooth fusion positioning; skRepresenting the walking step length of the kth step of the pedestrian, and obtaining the result through an acceleration sensor; delta thetakRepresenting the orientation angle variation of the k step after the pedestrian walks, and acquiring the orientation angle variation through a built-in gyroscope of the intelligent terminal; thetakRepresenting the orientation angle of the k step after the pedestrian walks; vkRepresenting noise, and obtaining a final positioning result through UKF algorithm fusion, wherein the UKF filter initial value (x)0,y0) Given by Wi-Fi Bluetooth fusion positioning result, initial azimuth orientation angle theta0And (4) given by a PDR processing module, carrying out fusion positioning by using a UKF algorithm to obtain a two-dimensional positioning coordinate.
In step 7), the floor is determined by using the barometer to obtain a height value in the vertical direction, specifically: the method comprises the steps of obtaining a measured atmospheric pressure value by establishing an atmospheric pressure elevation model, and calculating the atmospheric pressure value through the atmospheric pressure elevation model to obtain an elevation value, wherein the calculation formula is as follows:
in the above formula, h is the height of the target to be positioned, h0Is the height of reference point, p0Is the reference point air pressure, p is the air pressure at the location of the target to be positioned, tmIs two equal pressure surfaces p0And average temperature in degrees Celsius between p, t0Is the reference point temperature, t is the temperature of the position where the target to be positioned is located;
the ground pressure is generally 970-1040 (hpa), and p is0With an assignment of 1000hpa, h0Is 10m, t0At 20 ℃, the expression for calculating the height becomes:
according to the formula, the curve of the height value and the change of the temperature and the air pressure is drawn by using MATLAB.
In an actual environment, a building is generally closed, and if the temperature and the air pressure measured by an outdoor base station are used as references, the indoor calculation result is inaccurate, and at this time, a floor judgment method needs to be designed. Reference point pressure p0The air pressure p of the position of the target to be positioned and the reference point temperature t0And measuring the temperature t of the position of the target to be positioned by adopting a base station placed indoors, smoothing the measured data by adopting a filtering method, and then substituting the smoothed data into a formula to carry out height calculation and floor height judgment. The specific flow is shown in fig. 5.
And 8), fusing the two-dimensional positioning coordinates and the vertical direction height value to obtain final three-dimensional position coordinates, and realizing three-dimensional positioning of personnel in the building. Fusing the two-dimensional position coordinate obtained in the step 6 and the vertical height h obtained in the step 7 to obtain a three-dimensional position coordinate, thereby achieving an indoor three-dimensional positioning effect.
Claims (9)
1. A high-precision three-dimensional indoor positioning method based on multi-source fusion is characterized by comprising the following steps:
1) deploying a Wi-Fi environment in an area to be positioned, and establishing a Wi-Fi offline fingerprint database in an offline stage by acquiring a Wi-Fi signal strength value RSSI of each small area in the area to be positioned;
2) collecting RSSI (received signal strength indicator) values of Wi-Fi signals at an online stage, and realizing Wi-Fi positioning by improving a weighted centroid positioning algorithm;
3) deploying Bluetooth equipment in an area to be positioned, collecting a Bluetooth RSSI value, and realizing Bluetooth positioning by improving a weighted centroid positioning algorithm;
4) realizing Wi-Fi and Bluetooth fusion positioning through average weighting;
5) matching the result of the acceleration sensor by using a step size model, and performing PDR positioning;
6) fusing Wi-Fi, Bluetooth and PDR through a UKF algorithm to obtain a two-dimensional positioning coordinate;
7) utilizing a barometer to measure the height to judge the floor to obtain a height value in the vertical direction;
8) and fusing the two-dimensional positioning coordinate and the vertical direction height value to obtain a final three-dimensional position coordinate, thereby realizing three-dimensional positioning of personnel in the building.
2. The multi-source fusion-based high-precision three-dimensional indoor positioning method is characterized in that in the step 1), in the off-line stage, a plurality of positions are divided in the whole indoor scene, enough Wi-Fi RSSI samples are collected at each position, specific training is carried out on the signal samples to obtain a training result and coordinate information of the current position, the training result and the coordinate information of the current position are stored in a database to serve as training data, and a Wi-Fi fingerprint map is constructed after all training points are finished; the Wi-Fi signal acquisition for a plurality of times in a certain small area divided in an area to be positioned in a period of time t is as follows:
where p is the number of signal strengths acquired during the period of time t, rtqThe Wi-Fi signal strength of the p-th acquisition in t time is represented, and the constructed Wi-Fi offline fingerprint database is as follows:
I＝{(RSSV1,ο1),(RSSV2,ο2),...,(RSSVi,οi),...,(RSSVN,οN)}
wherein RSSVi＝(RSSi1,RSSi2,...,RSSiM) Is RSSI, from M Wi-Fi hotspotsi＝(x,y)∈R2Is RSSViThe corresponding position, x, y is the position information of the position.
3. The multi-source fusion-based high-precision three-dimensional indoor positioning method as claimed in claim 1, wherein in the step 2), the online stage is that after a Wi-Fi offline fingerprint database is established, euclidean distances between every two samples in the Wifi database are calculated, and the euclidean distance expression is as follows:
wherein DijRepresents the actual distance of sample i from sample j;
according to DijClustering, wherein each sample point is a class initially, two classes with the minimum distance are merged into a new class each time, a threshold value T is set, when the distance D between the two classes with the minimum distance is larger than T, clustering is finished, and the signal characteristic vector of the centroid of each class is calculated as
Let RSSI sample obtained online be denoted as T ═ RSSj1,RSSj2,...,RSSjM),(xj,yj) Wherein r) isj＝(RSSj1,RSSj2,...,RSSjM) RSSI representing M APs received online, a location vector o ═ x, y representing location information of RSSI obtained online, where the RSSI obtained online is known and the location vector (x, y) is unknown;
calculating the RSSI sample T obtained on line and the mass center M of all classesKFinding out the class P of the centroid point closest to the fingerprint distance of T, and calculating the K points closest to T in the Euclidean distance by the nearest neighbor algorithm, wherein the position information of the K points is (x)i,yi) If the number of the fingerprint points in P is less than K, extracting the physical positions of all the points in the class P, and pointing according to T and K pointsDetermining the distribution of coordinate weights by the aid of the Wen Euclidean distance, wherein the weights are calculated according to the following formula:
w 'of'sjIs a function of the transition of the weight, wsjThe coordinate weight value distributed to the sample j is K, and K represents K points which are closest to S in Euclidean distance in P;
the Wi-Fi positioning result is obtained by the following formula,
4. the method according to claim 1, wherein in step 3), the bluetooth positioning is realized by improving the weighted centroid positioning algorithm, and the bluetooth device is deployed in the area to be positioned, the bluetooth RSSI value is collected, and for m collection points (x) in the area to be positioned, the bluetooth RSSI value is collected1,y1)，(x2,y2)，...,(xm,ym)，
The signal intensity measured for m positions is S1,S2,...SmIs provided with
Then the weight is
The final positioning result is obtained as
5. The multi-source fusion-based high-precision three-dimensional indoor positioning method is characterized in that in the step 4), the Wi-Fi and Bluetooth fusion positioning is realized through average weighting, 12 Wi-Fi routers are deployed in an area to be positioned to deploy a Wi-Fi positioning environment, the area to be positioned is divided into a plurality of small grids of 2 x 2m, Wi-Fi signal intensity values are collected every 2s for 10 times, the position information of a collection point is recorded, and a Wi-Fi offline fingerprint database is constructed through a mean value filtering method;
meanwhile, 12 Bluetooth nodes for transmitting Bluetooth signals are deployed in an area to be positioned, as the Bluetooth output frequency is higher than the Wi-Fi output frequency, when a Wi-Fi positioning result is usually obtained once, the Bluetooth positioning result is obtained for 3-5 times, and the Wi-Fi and the Bluetooth are fused through average weighting, namely the Wi-Fi positioning result in certain positioning is as follows:
LWifi＝(x,y)
during this time period, the bluetooth positioning result is:
Lbeacon(x,y)＝{(x1,y1),(x2,y2),...,(xn,yn)}
the average weighted coordinates for bluetooth positioning are:
at the moment, a distance threshold value sigma s judgment is added, and the obtained Wi-Fi positioning coordinate L is obtainedWifiAnd bluetooth weighted coordinates
Comparing the distance d with a distance threshold value sigma s, and determining the self-adaption of the positioning weight values of the distance d and the distance threshold value sigma s; when d is less than or equal to σ s, namely the positioning results of the two are close, the positioning results of the Bluetooth and the Wi-Fi are both in a normal range, when d is greater than σ s, namely the positioning results of the two are greatly different, the possible positioning error is larger, and the self-adaptive weight rule is as follows:
Lwifibeaconnamely the final positioning coordinate after the Wi-Fi and the Bluetooth are fused, and the distance threshold value sigma s is determined according to the environment and the positioning error of the Wi-Fi and the Bluetooth.
6. The multi-source fusion-based high-precision three-dimensional indoor positioning method according to claim 1, wherein in step 5), the step model is used to match the acceleration sensor result to perform PDR positioning, and the model of the PDR positioning system is as follows:
where K is the number of steps, positional information x after walkingk,ykFor positional information after walking, θkDenotes the azimuth after K steps, Wk-1In order to be a noise, the noise is,
the step length model is used for matching the result of the acceleration sensor for the step length average value,
is the orientation angle variation; the simulated measurement equation is as follows:
wherein x isk,ykRepresenting the pedestrian position obtained by Wi-Fi and Bluetooth fusion positioning; skRepresents the k-th step walking step length of the pedestrian, obtained by the result of the acceleration sensor, delta thetakThe orientation angle variation of the kth step after the pedestrian walks can be obtained through a built-in gyroscope of the intelligent terminal, and thetakRepresenting the orientation angle of the k step after the pedestrian walks; vkRepresenting noise.
7. The multi-source fusion-based high-precision three-dimensional indoor positioning method is characterized in that in the step 6), Wi-Fi, Bluetooth and PDR are fused through a UKF algorithm to obtain a final positioning result, specifically, a pedestrian walking system model is constructed, initial position coordinates of PDR positioning are obtained through Wi-Fi positioning, and an error threshold is obtained by fusing Wi-Fi positioning and PDR positioning; the position of the pedestrian after one step can be obtained through a built-in sensor, a gyroscope and a direction sensor of the intelligent terminal, and state information and position information are updated by using a measurement equation;
the UKF algorithm sets the system equations and the bilateral equations to have discrete forms, i.e.
Wherein X is an n-dimensional random vector and
z is a random observation vector of m dimensions, f and h are nonlinear vector functions, WkAnd VVkZero mean white noise sequences which are mutually uncorrelated; u. ofk-1A deterministic control item; z is obtained by propagating X through a nonlinear function f (·), and the statistical characteristic of Z is
According to
Designing a series of points xii(i ═ 1, 2., L), called Sigma point, and calculated by f (-) propagation to give γi(i ═ 1, 2.., L), then based on γiComputing
Usually the number of Sigma dots is reduced by 2n +1
Modeling a system model for positioning the pedestrian walking by the PDR as follows:
wherein the number of steps of walking is represented by K, and the position information after walking is represented by xk,ykIs represented by thetakDenotes the azimuth angle after K steps, Wk-1The representation of the noise is represented by,
the step length model is adopted to match the result of the acceleration sensor for the step length average value,
is the orientation angle variation; the simulated measurement equation is as follows:
wherein x isk,ykRepresenting the pedestrian position obtained by Wi-Fi and Bluetooth fusion positioning; skRepresenting the walking step length of the kth step of the pedestrian, and obtaining the result through an acceleration sensor; delta thetakRepresenting the orientation angle variation of the k step after the pedestrian walks, and acquiring the orientation angle variation through a built-in gyroscope of the intelligent terminal; thetakRepresenting the orientation angle of the k step after the pedestrian walks; vkRepresenting noise, and obtaining a final positioning result through UKF algorithm fusion, wherein the UKF filter initial value (x)0,y0) Given by Wi-Fi Bluetooth fusion positioning result, initial azimuth orientation angle theta0And (4) given by a PDR processing module, carrying out fusion positioning by using a UKF algorithm to obtain a two-dimensional positioning coordinate.
8. The multi-source fusion-based high-precision three-dimensional indoor positioning method according to claim 1, wherein in step 7), the barometer is used to determine the height of the floor to obtain a vertical height value, specifically: the method comprises the steps of obtaining a measured atmospheric pressure value by establishing an atmospheric pressure elevation model, and calculating the atmospheric pressure value through the atmospheric pressure elevation model to obtain an elevation value, wherein the calculation formula is as follows:
in the above formula, h is the height of the target to be positioned, h0Is the height of reference point, p0Is the reference point air pressure, p is the air pressure at the location of the target to be positioned, tmIs two equal pressure surfaces p0And average temperature in degrees Celsius between p, t0Is the reference point temperature, t is the temperature of the position where the target to be positioned is located;
reference point pressure p0The air pressure p of the position of the target to be positioned and the reference point temperature t0And measuring the temperature t of the position of the target to be positioned by adopting a base station placed indoors, smoothing the measured data by adopting a filtering method, and then substituting the smoothed data into a formula to carry out height calculation and floor height judgment.
9. The multi-source fusion-based high-precision three-dimensional indoor positioning method according to claim 1, wherein in step 8), the final three-dimensional position coordinate is obtained by fusing the two-dimensional positioning coordinate and the vertical height value to realize three-dimensional positioning of the personnel in the building, and the three-dimensional position coordinate is obtained by fusing the two-dimensional position coordinate obtained in step 6) and the vertical height h obtained in step 7), so that an indoor three-dimensional positioning effect is achieved.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010890275.2A CN111901749A (en) | 2020-08-29 | 2020-08-29 | High-precision three-dimensional indoor positioning method based on multi-source fusion |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010890275.2A CN111901749A (en) | 2020-08-29 | 2020-08-29 | High-precision three-dimensional indoor positioning method based on multi-source fusion |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111901749A true CN111901749A (en) | 2020-11-06 |
Family
ID=73225564
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010890275.2A Pending CN111901749A (en) | 2020-08-29 | 2020-08-29 | High-precision three-dimensional indoor positioning method based on multi-source fusion |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111901749A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113048977A (en) * | 2021-03-09 | 2021-06-29 | 杭州十域科技有限公司 | Indoor geomagnetic positioning method integrating radio waves and inertial sensor |
| CN113207169A (en) * | 2021-03-26 | 2021-08-03 | 福建新大陆软件工程有限公司 | Fusion positioning method based on multi-source heterogeneous data |
| CN113645568A (en) * | 2021-08-12 | 2021-11-12 | 山东卡尔电气股份有限公司 | Floor positioning system and method based on intelligent positioner |
| CN113852911A (en) * | 2021-09-26 | 2021-12-28 | 桂林电子科技大学 | Fingerprint library and PDR calculation-based fusion positioning method and fingerprint library updating method |
| WO2023287917A1 (en) * | 2021-07-14 | 2023-01-19 | Proc12 Inc. | Systems and methods for spatial tracking |
| CN116819591A (en) * | 2023-06-30 | 2023-09-29 | 常州市场景信息科技有限公司 | Positioning method and system based on RTK and Bluetooth |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110118549A (en) * | 2018-02-06 | 2019-08-13 | 刘禹岐 | A kind of Multi-source Information Fusion localization method and device |
| CN110320495A (en) * | 2019-08-01 | 2019-10-11 | 桂林电子科技大学 | A kind of indoor orientation method based on Wi-Fi, bluetooth and PDR fusion positioning |
-
2020
- 2020-08-29 CN CN202010890275.2A patent/CN111901749A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110118549A (en) * | 2018-02-06 | 2019-08-13 | 刘禹岐 | A kind of Multi-source Information Fusion localization method and device |
| CN110320495A (en) * | 2019-08-01 | 2019-10-11 | 桂林电子科技大学 | A kind of indoor orientation method based on Wi-Fi, bluetooth and PDR fusion positioning |
Non-Patent Citations (3)
| Title |
|---|
| 佟华等: "GRAPES模式后处理技术改进应用研究", 《气象科技》 * |
| 佟华等: "GRAPES模式后处理技术改进应用研究", 《气象科技》, no. 04, 15 August 2020 (2020-08-15) * |
| 邓中亮: "《室内外无线定位与导航》", 31 December 2013, pages: 4 * |
Cited By (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113048977A (en) * | 2021-03-09 | 2021-06-29 | 杭州十域科技有限公司 | Indoor geomagnetic positioning method integrating radio waves and inertial sensor |
| CN113207169A (en) * | 2021-03-26 | 2021-08-03 | 福建新大陆软件工程有限公司 | Fusion positioning method based on multi-source heterogeneous data |
| CN113207169B (en) * | 2021-03-26 | 2022-08-05 | 福建新大陆软件工程有限公司 | Fusion positioning method based on multi-source heterogeneous data |
| WO2023287917A1 (en) * | 2021-07-14 | 2023-01-19 | Proc12 Inc. | Systems and methods for spatial tracking |
| CN113645568A (en) * | 2021-08-12 | 2021-11-12 | 山东卡尔电气股份有限公司 | Floor positioning system and method based on intelligent positioner |
| CN113645568B (en) * | 2021-08-12 | 2024-02-27 | 山东卡尔电气股份有限公司 | Floor positioning system and method based on intelligent positioner |
| CN113852911A (en) * | 2021-09-26 | 2021-12-28 | 桂林电子科技大学 | Fingerprint library and PDR calculation-based fusion positioning method and fingerprint library updating method |
| CN113852911B (en) * | 2021-09-26 | 2024-05-07 | 桂林电子科技大学 | Fusion positioning method based on fingerprint library and PDR calculation and fingerprint library updating method |
| CN116819591A (en) * | 2023-06-30 | 2023-09-29 | 常州市场景信息科技有限公司 | Positioning method and system based on RTK and Bluetooth |
| CN116819591B (en) * | 2023-06-30 | 2024-01-19 | 常州市场景信息科技有限公司 | Positioning method and system based on RTK and Bluetooth |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111901749A (en) | High-precision three-dimensional indoor positioning method based on multi-source fusion | |
| CN109413578B (en) | Indoor positioning method based on fusion of WIFI and PDR | |
| CN110118549B (en) | Multi-source information fusion positioning method and device | |
| CN110320495A (en) | A kind of indoor orientation method based on Wi-Fi, bluetooth and PDR fusion positioning | |
| CN105813194B (en) | Indoor orientation method based on fingerprint database secondary correction | |
| CN106412839B (en) | Based on secondary partition and the matched indoor positioning of fingerprint gradient and tracking | |
| KR102116824B1 (en) | Positioning system based on deep learnin and construction method thereof | |
| CN106793086B (en) | Indoor positioning method | |
| CN110958575B (en) | Positioning method and system based on WiFi fusion prediction | |
| CN110557716B (en) | Indoor positioning method based on lognormal model | |
| CN105263113A (en) | Wi-Fi location fingerprint map building method and system based on crowd-sourcing | |
| WO2017121168A1 (en) | Cluster-based magnetic positioning method, device and system | |
| CN111970633A (en) | Indoor positioning method based on WiFi, Bluetooth and pedestrian dead reckoning fusion | |
| CN109151750A (en) | A kind of LTE indoor positioning floor method of discrimination based on Recognition with Recurrent Neural Network model | |
| CN111698774B (en) | Indoor positioning method and device based on multi-source information fusion | |
| CN109195104A (en) | A kind of indoor orientation method combined based on support vector regression and Kalman filtering | |
| CN112135246B (en) | RSSI (received Signal Strength indicator) updating indoor positioning method based on SSD (solid State disk) target detection | |
| CN112729301A (en) | Indoor positioning method based on multi-source data fusion | |
| CN115808170A (en) | Indoor real-time positioning method integrating Bluetooth and video analysis | |
| CN106197418B (en) | A kind of indoor orientation method merged based on the fingerprint technique of sliding window with sensor | |
| CN109541537B (en) | Universal indoor positioning method based on ranging | |
| CN109640253B (en) | Mobile robot positioning method | |
| Kawauchi et al. | FineMesh: High-Density Sampling Platform Using an Autonomous Robot | |
| CN115052245A (en) | Unmanned aerial vehicle-assisted wireless sensor network node positioning method based on deep learning | |
| CN108696818A (en) | The distance metric method and system of Wi-Fi fingerprint locations |
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 |
Application publication date: 20201106 |
|
| RJ01 | Rejection of invention patent application after publication |