CN106412841A - Indoor location method based on DBSCAN algorithm - Google Patents

Abstract

The present invention provides an indoor location method based on a DBSCAN algorithm, belonging to the indoor location technology field. The method comprises: installing a signal emitter and a signal collector in a building requiring location, sending the number of the sample points, obtaining the signal intensity sequence of each sample point, performing recoding and numbering of the coordinates and the signal intensity sequence of each sample point, employing the DBSCAN algorithm to build an indoor location model, obtaining the signal intensity sequences of the location points, and obtaining the location coordinates of the points to be located according to the established indoor location model. The indoor location method based on the DBSCAN algorithm employs the unsupervised-learning algorithm to perform modeling and performs classification clustering according to the wireless signal intensity value (RSSI) without artificial grid division to allow the grid division to more correspond with the reality so as to improve the stability and the precision of the indoor location.

Description

Indoor positioning method based on DBSCAN algorithm

Technical Field

The invention belongs to the technical field of indoor positioning, and particularly relates to an indoor positioning method based on a DBSCAN algorithm.

Background

With the development of the internet of things, the demand for indoor positioning is increasing day by day, such as in indoor environments of shopping malls, schools, office buildings, hospitals, hotels, airports, warehouses, and the like, an indoor positioning technology is needed to efficiently manage resources and personnel, and a great amount of demand exists in the VR field to position indoor position information of players, so that how to better meet the increasing indoor positioning demand becomes a hotspot problem in the current positioning technology.

In the existing indoor positioning technology, a fingerprint positioning method based on wireless signal strength (RSSI) is widely applied to various indoor positioning systems, and in an indoor environment, the measured value of the wireless signal strength (RSSI) fluctuates greatly because the propagation of wireless signals is easily influenced by temperature, humidity and personnel movement; in order to improve the stability and the precision of measurement, indoor positioning is realized by a common method of matching a fingerprint database, and common methods of realizing indoor positioning by using the matching fingerprint database include a naive Bayes method, a K-nearest neighbor method (KNN), Bayesian estimation, a neural network method and the like, which are supervised learning methods, and usually require artificial grid division, and the networks are generally divided by bisecting areas.

Disclosure of Invention

In order to solve the defects of the prior art, the invention provides an indoor positioning method based on a DBSCAN algorithm, which comprises the following steps:

step 1, installing a plurality of signal transmitters and a plurality of signal collectors in a building to be positioned;

step 2, setting the number of the sample points, determining the coordinates of each sample point, and obtaining a signal intensity sequence of each sample point, namely the signal intensity from each sample point to each signal transmitter;

step 3, recording and numbering the coordinates and the signal intensity sequence of each sample point;

and 4, establishing an indoor positioning model by using a DBSCAN algorithm according to the numbered coordinates and signal intensity sequence of each sample point, wherein the method comprises the following steps:

step 4.1, setting density selection parameters, namely setting the length of a neighborhood and the number of sample points at least contained in each core object;

step 4.2, setting constraint conditions of the core objects according to the numbered signal intensity sequence of each sample point and the set density selection parameters, determining the core objects of the sample points according to the constraint conditions, and further determining the clustering clusters through the core objects; the constraint conditions of the core object are as follows: whether the number of other sample points contained in the neighborhood of each sample point is larger than or equal to the number of sample points at least contained in each set core object or not;

4.3, obtaining the average signal intensity of each cluster, and obtaining the average position coordinate of each cluster according to the numbered coordinates of each sample point, namely completing the establishment of an indoor positioning model;

step 5, obtaining a signal intensity sequence of the point to be located, and obtaining a locating coordinate of the point to be located according to the established indoor locating model, wherein the method specifically comprises the following steps: and acquiring the Euclidean distance between the to-be-positioned point and the average signal intensity of each cluster, and selecting the average position coordinate of the cluster with the minimum Euclidean distance as the positioning coordinate of the to-be-positioned point.

The constraint condition of the core object in step 4.2 is expressed by the following formula:

|N_θ(x_j)|≥Minpts (1)

wherein N is_θRepresenting the number of samples in the theta neighborhood of a certain sample point; θ represents the length of the neighborhood; x is the number of_jRepresents the jth sample point, j represents a natural number; minpts represents the number of sample points at least contained in each core object;

wherein,

N_θ(x_j)＝{x_i∈D|d(x_i,x_j)≤θ} (2)

wherein x is_iRepresents the ith sample point; i represents a natural number; d represents a data set of the numbered coordinates and signal intensity sequence of each sample point; d (x)_i,x_j) Denotes x_iAnd x_jThe euclidean distance between two sample points;

wherein,

wherein n represents the number of wireless signal strengths recorded at each sample point; k represents a natural number; r is_ikA k-th wireless signal strength value representing an i-th sample point; r is_jkRepresenting the kth wireless signal strength value for the jth sample point.

The invention has the advantages that:

the invention provides an indoor positioning method based on a DBSCAN algorithm, which utilizes an unsupervised learning algorithm to carry out modeling, carries out classification clustering according to a wireless signal strength value (RSSI) instead of artificially carrying out meshing, enables the meshing to be more practical and improves the stability and the accuracy of indoor positioning.

Drawings

Fig. 1 is a flow chart of an indoor positioning method based on DBSCAN algorithm according to an embodiment of the present invention;

fig. 2 is a schematic diagram of indoor positioning according to an embodiment of the present invention.

Detailed Description

An embodiment of the present invention will be further described with reference to the accompanying drawings.

In the embodiment of the present invention, a DBSCAN algorithm-based indoor positioning method, a flow chart of which is shown in fig. 1, includes the following steps:

step 1, installing a plurality of signal transmitters and a plurality of signal collectors in a building to be positioned;

step 2, setting the number of sample points, and determining the coordinate (Z) of each sample point₁,Z₂) And obtaining a signal strength sequence r for each sample point₁,r₂,···,r_nI.e. the signal strength of each sample point to the respective signal emitter;

and 3, recording and numbering the coordinate and signal intensity sequence of each sample point as sample data to obtain a data set D ═ x₁,x₂,···,x_m}; wherein m represents a natural number, i.e. a serial number of sample data;

and 4, establishing an indoor positioning model by using a DBSCAN algorithm according to the numbered coordinates and signal intensity sequence of each sample point, wherein the method comprises the following steps:

step 4.1, setting density selection parameters, namely setting the length theta of a neighborhood and the number Minpts of at least sample points contained in each core object;

step 4.2, setting constraint conditions of the core objects according to the numbered signal intensity sequence of each sample point and the set density selection parameters, determining all the core objects of the sample points according to the constraint conditions, and further determining the clustering cluster through the core objects; the constraint conditions of the core object are as follows: whether the number of other sample points contained in the neighborhood of each sample point is more than or equal to the set number of sample points at least contained in each core object or not is determined by the following specific steps:

step 4.2.1, setting constraint conditions of the core objects according to the numbered signal intensity sequence of each sample point and the set length of the neighborhood, and determining all the core objects in the sample points according to the constraint conditions, wherein the constraint condition formula is as follows:

|N_θ(x_j)|≥Minpts (1)

wherein N is_θRepresenting the number of samples in the theta neighborhood of a certain sample point; θ represents the length of the neighborhood; x is the number of_jRepresents the jth sample point, j represents a natural number, namely the serial number of the sample point; minpts represents the number of sample points at least contained in each core object;

in the embodiment of the invention, for x_j∈ D whose theta neighborhood includes the sum x in the sample set D_jIs less than or equal to θ, the formula is:

N_θ(x_j)＝{x_i∈D|d(x_i,x_j)≤θ} (2)

wherein x is_iRepresents the ith sample point; i represents a natural number, i.e. the serial number of the sample point; d represents a data set of the numbered coordinates and signal intensity sequence of each sample point;

in the examples of the present invention, d (x)_i,x_j) Denotes x_iAnd x_jThe Euclidean distance between two sample points, i.e. confidenceNumber intensity sequence r₁,r₂,···,r_nX in n-dimensional Euclidean space_iAnd x_jThe distance between two points is given by the following formula:

wherein n represents the number of wireless signal strengths recorded at each sample point; k represents a natural number, namely a serial number of the wireless signal strength value; r is_ikA k-th wireless signal strength value representing an i-th sample point; r is_jkA k-th wireless signal strength value representing a j-th sample point;

step 4.2.2, further determining all clustering clusters through the core object;

in the embodiment of the present invention, if x_jAt x_iIn the field of θ, and x_iIs a core object, then called x_jFrom x_iThe density is direct; for x_iAnd x_jIf there is a sample sequence { p }₁,p₂,···,p_nIn which p is₁＝x_i，p_n＝x_jAnd each sample in the sample sequence can be reached by the density of the previous sample, then x is called_jFrom x_iThe density can be reached; randomly selecting a core object in the sample set D as a 'seed', and starting to find out all sample points with reachable density, namely determining a corresponding cluster; accessing all core objects to obtain all cluster clusters;

4.3, obtaining the average signal intensity of each cluster, and calculating to obtain the average position coordinate of each cluster according to the numbered coordinates of each sample point, namely completing the establishment of an indoor positioning model;

in the embodiment of the present invention, the following formula is adopted to obtain the average signal intensity of each cluster:

wherein h represents the number of sample points in each cluster; r is_i1Representing a first wireless signal strength of an ith sample point in each cluster; r is_i2Representing a second wireless signal strength of an ith sample point in each cluster; r is_inRepresenting the nth wireless signal strength of the ith sample point in each cluster;

in the embodiment of the present invention, the average position coordinate of each cluster is obtained by using the following formula:

wherein Z is_i1An abscissa value representing the ith sample point in each cluster; z_i2Expressing the longitudinal coordinate value of the ith sample point in each cluster;

step 5, obtaining a signal intensity sequence of the point to be located through on-line measurement and the like, and obtaining a location coordinate of the point to be located according to the established indoor location model, wherein the method specifically comprises the following steps: acquiring the Euclidean distance between the to-be-positioned point and the average signal intensity of each cluster, and selecting the average position coordinate of the cluster with the minimum Euclidean distance as the positioning coordinate of the to-be-positioned point;

in the embodiment of the invention, as shown in fig. 2, the method is simulated on a computer, 14 sample points are set, a point to be positioned is set randomly, and the wireless signal strength sequence received by each sample point is { r }₁,r₂Dividing the sample points into two clustering clusters of C1 and C2 by using a DBSCAN algorithm, and respectively obtaining a central point of each clustering cluster, namely an average signal intensity point; and acquiring the Euclidean distance between the to-be-positioned point and the central point of each cluster, wherein the Euclidean distance between the to-be-positioned point and the central point of each cluster is the minimum, and the positioned coordinate is the average position coordinate of the cluster C1.

Claims (2)

1. An indoor positioning method based on DBSCAN algorithm is characterized in that: the method comprises the following steps:

step 1, installing a plurality of signal transmitters and a plurality of signal collectors in a building to be positioned;

step 2, setting the number of the sample points, determining the coordinates of each sample point, and obtaining a signal intensity sequence of each sample point, namely the signal intensity from each sample point to each signal transmitter;

step 3, recording and numbering the coordinates and the signal intensity sequence of each sample point;

and 4, establishing an indoor positioning model by using a DBSCAN algorithm according to the numbered coordinates and signal intensity sequence of each sample point, wherein the method comprises the following steps:

step 4.1, setting density selection parameters, namely setting the length of a neighborhood and the number of sample points at least contained in each core object;

step 4.2, setting constraint conditions of the core objects according to the numbered signal intensity sequence of each sample point and the set density selection parameters, determining the core objects of the sample points according to the constraint conditions, and further determining the clustering clusters through the core objects; the constraint conditions of the core object are as follows: whether the number of other sample points contained in the neighborhood of each sample point is larger than or equal to the number of sample points at least contained in each set core object or not;

4.3, obtaining the average signal intensity of each cluster, and obtaining the average position coordinate of each cluster according to the numbered coordinates of each sample point, namely completing the establishment of an indoor positioning model;

step 5, obtaining a signal intensity sequence of the point to be located, and obtaining a locating coordinate of the point to be located according to the established indoor locating model, wherein the method specifically comprises the following steps: and acquiring the Euclidean distance between the to-be-positioned point and the average signal intensity of each cluster, and selecting the average position coordinate of the cluster with the minimum Euclidean distance as the positioning coordinate of the to-be-positioned point.

2. The indoor positioning method based on DBSCAN algorithm as claimed in claim 1, wherein: the constraint condition of the core object in step 4.2 is expressed by the following formula:

|N_θ(x_j)|≥Minpts (1)

wherein N is_θRepresenting the number of samples in the theta neighborhood of a certain sample point; θ represents the length of the neighborhood; x is the number of_jRepresents the jth sample point, j represents a natural number; minpts represents the number of sample points at least contained in each core object;

wherein,

N_θ(x_j)＝{x_i∈D|d(x_i,x_j)≤θ} (2)

wherein x is_iRepresents the ith sample point; i represents a natural number; d represents a data set of the numbered coordinates and signal intensity sequence of each sample point; d (x)_i,x_j) Denotes x_iAnd x_jThe euclidean distance between two sample points;

wherein,

$d(x_i,x_j)=\sqrt{{\mathrm{\Σ}}_{k=1}^n{(r_{ik}-r_{jk})}^2}---\left(3\right)$

wherein n represents the number of wireless signal strengths recorded at each sample point; k represents a natural number; r is_ikA k-th wireless signal strength value representing an i-th sample point; r is_jkRepresenting the kth wireless signal strength value for the jth sample point.