CN112020137A - Intelligent factory positioning method and device for indoor terminal - Google Patents
Intelligent factory positioning method and device for indoor terminal Download PDFInfo
- Publication number
- CN112020137A CN112020137A CN202010800216.1A CN202010800216A CN112020137A CN 112020137 A CN112020137 A CN 112020137A CN 202010800216 A CN202010800216 A CN 202010800216A CN 112020137 A CN112020137 A CN 112020137A
- Authority
- CN
- China
- Prior art keywords
- indoor
- characteristic information
- original position
- node
- information
- 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
Images
Classifications
-
- 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
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/309—Measuring or estimating channel quality parameters
- H04B17/318—Received signal strength
-
- 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
Abstract
The invention discloses an intelligent factory positioning method and device for an indoor terminal. The method comprises the following steps: s1, an offline stage, specifically including: the method comprises the steps that the strength of received signals from all indoor wireless nodes collected on each reference node is used as original position characteristic information, and an original position characteristic space is constructed; carrying out nonlinear mapping conversion on the original position feature space to obtain a target feature space; s2, an online stage, specifically comprising: the received signal strength from each indoor wireless node collected on the test node is used as the characteristic information of the test position; carrying out nonlinear mapping conversion on the test position characteristic information to obtain target characteristic information; and determining the position of the terminal in the intelligent factory according to the target characteristic information and the target characteristic space by adopting a weighted neighbor method. The invention can effectively improve the positioning precision of the terminal in the intelligent factory.
Description
Technical Field
The invention relates to the technical field of wireless positioning, in particular to an indoor terminal-oriented intelligent factory positioning method and device.
Background
An intelligent factory is a new stage of the current factory based on equipment intellectualization, management modernization and information computerization, and an internal real-time positioning system continuously tracks factory equipment, an AGV (automatic guided vehicle), personnel, materials and the like through a real-time positioning indoor terminal and sends positioning data to an upper-layer software system, so that the refined production management of the intelligent factory is realized.
Currently, the mainstream intelligent factory positioning method is a method based on clustering and decision tree, that is, the received signal strength (RSS value) from the best indoor wireless node in the factory area is selected as the position characteristic information. In fact, a plurality of indoor wireless nodes in a factory area have a signal interference phenomenon, and although the existing intelligent factory positioning method is simple and easy to use and is rapid in positioning, the following outstanding defects still exist:
1. unselected indoor wireless nodes and corresponding RSS values in the factory area are abandoned, and the influence of RSS values from other indoor wireless nodes on the positioning result is difficult to fully consider;
2. when the number of selectable indoor wireless nodes in the factory area is small, the RSS value from the best indoor wireless node is selected as the position characteristic information, and the accuracy of the positioning result is limited to a certain extent.
Therefore, the positioning accuracy of the existing intelligent factory positioning method is low.
Disclosure of Invention
In order to overcome the defects of the prior art, the invention provides an intelligent factory positioning method and device facing an indoor terminal, which can effectively improve the positioning accuracy of the indoor terminal in an intelligent factory.
In order to solve the foregoing technical problem, in a first aspect, an embodiment of the present invention provides an intelligent factory positioning method for an indoor terminal, including:
s1, an offline stage, specifically including:
the method comprises the steps that the strength of received signals from all indoor wireless nodes collected on each reference node is used as original position characteristic information, and an original position characteristic space is constructed;
carrying out nonlinear mapping conversion on the original position feature space to obtain a target feature space;
s2, an online stage, specifically comprising:
the received signal strength from each indoor wireless node collected on the test node is used as the characteristic information of the test position;
carrying out nonlinear mapping conversion on the test position characteristic information to obtain target characteristic information;
and determining the position of the terminal in the intelligent factory according to the target characteristic information and the target characteristic space by adopting a weighted neighbor method.
Further, before constructing an original location feature space by using the received signal strength from each indoor wireless node collected at each reference node as original location feature information, the method further includes:
and laying a plurality of reference nodes in a wireless coverage area, and constructing an original position physical space according to the physical positions of all the reference nodes.
Further, the constructing of the original location feature space by using the received signal strength from each indoor wireless node collected at each reference node as original location feature information specifically includes:
acquiring identification information of each indoor wireless node on each reference node, acquiring the strength of received signals from the same indoor wireless node for multiple times, and taking the average value of the strength of the received signals from the same indoor wireless node as the original position characteristic information;
and when the acquisition is finished at all the reference nodes, constructing the original position feature space according to all the original position feature information.
Further, the taking the strength of the received signal from each of the indoor wireless nodes collected at the test node as the test location characteristic information specifically includes:
and acquiring identification information of each indoor wireless node on the test node, acquiring the strength of received signals from the same indoor wireless node for multiple times, and taking the average value of the strength of the received signals from the same indoor wireless node as the characteristic information of the test position.
Further, the determining, by using a weighted neighbor method, the position of the internal chamber terminal of the intelligent factory according to the target feature information and the target feature space specifically includes:
calculating Euclidean distance between the target feature information and each original position feature information in the target feature space;
and screening a plurality of original position characteristic information according to the sequence of the Euclidean distance from small to large, and determining the position of the indoor terminal according to the screened original position characteristic information.
In a second aspect, an embodiment of the present invention provides an intelligent factory positioning apparatus for an indoor terminal, including:
the offline module specifically comprises:
the system comprises an original position characteristic space construction unit, a position information acquisition unit and a position information acquisition unit, wherein the original position characteristic space construction unit is used for constructing an original position characteristic space by taking the received signal strength from each indoor wireless node acquired on each reference node as original position characteristic information;
the target characteristic space obtaining unit is used for carrying out nonlinear mapping conversion on the original position characteristic space to obtain a target characteristic space;
the online module specifically comprises:
the test position characteristic information acquisition unit is used for taking the received signal strength from each indoor wireless node acquired on the test node as test position characteristic information;
the target characteristic information acquisition unit is used for carrying out nonlinear mapping conversion on the test position characteristic information to obtain target characteristic information;
and the indoor terminal positioning unit is used for determining the position of the indoor terminal in the intelligent factory according to the target characteristic information and the target characteristic space by adopting a weighted neighbor method.
Further, the original location feature space construction unit is further configured to, before constructing an original location feature space by using the received signal strength from each indoor wireless node acquired at each reference node as original location feature information, lay a plurality of reference nodes in a wireless coverage area, and construct an original location physical space according to physical locations of all the reference nodes.
Further, the constructing of the original location feature space by using the received signal strength from each indoor wireless node collected at each reference node as original location feature information specifically includes:
acquiring identification information of each indoor wireless node on each reference node, acquiring the strength of received signals from the same indoor wireless node for multiple times, and taking the average value of the strength of the received signals from the same indoor wireless node as the original position characteristic information;
and when the acquisition is finished at all the reference nodes, constructing the original position feature space according to all the original position feature information.
Further, the taking the strength of the received signal from each of the indoor wireless nodes collected at the test node as the test location characteristic information specifically includes:
and acquiring identification information of each indoor wireless node on the test node, acquiring the strength of received signals from the same indoor wireless node for multiple times, and taking the average value of the strength of the received signals from the same indoor wireless node as the characteristic information of the test position.
Further, the determining, by using a weighted neighbor method, the position of the internal chamber terminal of the intelligent factory according to the target feature information and the target feature space specifically includes:
calculating Euclidean distance between the target feature information and each original position feature information in the target feature space;
and screening a plurality of original position characteristic information according to the sequence of the Euclidean distance from small to large, and determining the position of the indoor terminal according to the screened original position characteristic information.
The embodiment of the invention has the following beneficial effects:
in the off-line stage, the received signal strength from each indoor wireless node acquired on each reference node is used as original position characteristic information to construct an original position characteristic space, the original position characteristic space is subjected to nonlinear mapping conversion to obtain a target characteristic space, in the on-line stage, the received signal strength from each indoor wireless node acquired on a test node is used as test position characteristic information, the test position characteristic information is subjected to nonlinear mapping conversion to obtain target characteristic information, and the position of a terminal in an intelligent factory is determined according to the target characteristic information and the target characteristic space by adopting a weighted neighbor method, so that accurate positioning of the terminal in the intelligent factory is realized. Compared with the prior art, the embodiment of the invention adopts the nonlinear mapping conversion feature extraction algorithm to train the original position feature information in the off-line stage, the nonlinear mapping conversion feature extraction algorithm can expand the original position feature space to a high-dimensional space for analysis, and the original position feature information is utilized to a greater extent, so that the positioning precision of the indoor terminal of the intelligent factory can be effectively improved.
Drawings
Fig. 1 is a schematic flowchart of an intelligent factory location method for an indoor terminal according to a first embodiment of the present invention;
fig. 2 is another schematic flow chart of an intelligent factory positioning method for an indoor terminal according to a first embodiment of the present invention;
FIG. 3 is a flow chart illustrating an offline stage according to a first embodiment of the present invention;
FIG. 4 is a schematic flow chart of the online phase in the first embodiment of the present invention;
fig. 5 is a schematic structural diagram of an intelligent factory positioning apparatus facing an indoor terminal according to a second embodiment of the present invention.
Detailed Description
The technical solutions in the present invention will be described clearly and completely with reference to the accompanying drawings, and it is obvious that the described embodiments are only some embodiments of the present invention, not all embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It should be noted that, the step numbers in the text are only for convenience of explanation of the specific embodiments, and do not serve to limit the execution sequence of the steps.
The first embodiment:
as shown in fig. 1 to 4, a first embodiment provides an intelligent factory location method for an indoor terminal, including steps S1 to S2:
s1, an off-line stage, which comprises steps S11-S12:
and S11, constructing an original position characteristic space by taking the received signal strength from each indoor wireless node collected on each reference node as original position characteristic information.
And S12, carrying out nonlinear mapping conversion on the original position feature space to obtain a target feature space.
S2, an online stage, which specifically comprises the following steps S21-S23:
and S21, taking the strength of the received signal from each indoor wireless node collected on the test node as the characteristic information of the test position.
And S22, carrying out nonlinear mapping conversion on the test position characteristic information to obtain target characteristic information.
And S23, determining the position of the terminal in the intelligent factory according to the target characteristic information and the target characteristic space by adopting a weighted neighbor method.
In a preferred embodiment, before constructing the raw location feature space by using the received signal strength from each indoor wireless node collected at each reference node as raw location feature information, the method further includes: and laying a plurality of reference nodes in a wireless coverage area, and constructing an original position physical space according to the physical positions of all the reference nodes.
Illustratively, in the process of constructing the original position feature space in the off-line stage, N reference nodes are pre-arranged, and the physical position of each reference node is li(xi,yi) The physical locations of the N reference nodes form an original location physical space L ═ L1,l2,…,lN)TCollecting received signal strength RSS from n indoor wireless nodes on each reference node to obtain n-dimensional vector Fi＝(RSS1,RSS2,…,RSSn)TI ∈ (1, N), i.e., the original location feature information, stores the original location feature information of all reference nodes in a relevant database, and constructs an N × N-dimensional original location feature space F ═ F (F)1,F2,…,FN)TEach row vector in the matrix F represents the original location characteristic information of one reference node.
In the process of obtaining the target feature space in the off-line stage, the specific steps of performing nonlinear mapping conversion on the original position feature space are as follows:
1. taking the original position feature space F as an input space, and calculating a nonlinear mapping transformation matrix K according to the formula (1):
in the formula (1), the width is Gaussian nucleus width; i, j is less than or equal to N; k is an NxN dimensional matrix;
2. correcting the nonlinear mapping transformation matrix K according to the formula (2) to obtain a corrected kernel matrix
In the formula (2), C is an NxN-order matrix, and each element is
3. Calculating the first K maximum eigenvalues lambda of the nonlinear mapping transformation matrix K1≥λ2...λk-1≥λkAnd corresponding feature vector v1,v2,…,vk(ii) a Such as:
nonlinear mapping transformation matrix
The characteristic polynomial of (a) is | λ E-K |, i.e.
Simply to get | λ E-K | ═ λ +2)2(lambda-4), then the eigenvalues of K are 4, -2, -2, respectively, the first 3 largest eigenvalues are lambda3＝4≥λ1＝λ2＝-2；
When lambda is1＝λ2When-2, the characteristic equation set (λ E-a) X is substituted with 0, i.e.
Obtaining basic solution system X1＝[1,1,0]TI.e. k1X1Is the entire eigenvector (k) belonging to eigenvalue-21≠0)；
The same can obtain lambda3Basic solution system X when being 42＝[0,1,1]T，k2X2Is the overall feature direction belonging to the feature value 4
Quantity (k)2≠0)；
4. V is orthogonalized by1,v2,…,vkOrthogonalizing the units to obtain a1,a2,…,ak；
5. The target feature space F' is calculated according to equation (3):
in the formula (3), a ═ a1,a2,…,ak)TIs an Nxn dimensional matrix, and K-is a corrected kernel matrix.
The process of collecting the test location feature information at the online stage is similar to the process of collecting the original location feature information at the offline stage, and an n-dimensional vector S ═ is obtained by collecting the received signal strength RSS from n indoor wireless nodes at the test node (RSS)1,RSS2,…,RSSn)TI.e. test location signature information.
In the process of acquiring the target characteristic information at the online stage, performing nonlinear mapping transformation processing on the test position characteristic information S according to the formula (4) to obtain target characteristic information S':
in the formula (4), atIs a feature vector; k (F)t,S)＝Φ(Ft) Phi (S), wherein phi is a mapping function from a data space to a feature space; k (F)i,S)＝Φ(Fi)Φ(S)；KtjConverting a matrix K (F) for a non-linear mappingtS) elements of the jth row and jth column; kijConverting a matrix K (F) for a non-linear mappingiS) elements of the ith row and the jth column.
In the process of positioning the indoor terminal at the online stage, a weighted neighbor method is adopted, and the specific steps of determining the position of the indoor terminal in the intelligent factory according to the target characteristic information S 'and the target characteristic space F' are as follows:
1. calculating target characteristic information S 'and each original position characteristic information F in the target characteristic space F' according to the formula (5)iThe euclidean distance of':
in the formula (5), Di(S′,Fi') can characterize S' and Fi' degree of similarity between them, the smaller the value, the more similar the two are;
2. according to Di(S′,Fi') size arrangement, find the top m < N minimum Euclidean distances DiAnd m original position characteristic information F corresponding to the m original position characteristic informationi' and reference node physical location li(xi,yi) So that it satisfies formula (6):
in the formula (6), 0 < d0Less than 1, preventing the denominator to be 0; σ is a positive number less than 1;
3. calculating the position of each indoor terminal in the intelligent factory according to the formula (7):
in the embodiment, in the off-line stage, the received signal strength from each indoor wireless node acquired on each reference node is used as original position feature information to construct an original position feature space, and the original position feature space is subjected to nonlinear mapping conversion to obtain a target feature space, in the on-line stage, the received signal strength from each indoor wireless node acquired on a test node is used as test position feature information, and the test position feature information is subjected to nonlinear mapping conversion to obtain target feature information, so that the position of the indoor terminal in the intelligent factory is determined according to the target feature information and the target feature space by adopting a weighted neighbor method, and accurate positioning of the indoor terminal in the intelligent factory is realized. In the embodiment, the original position feature information is trained by adopting the nonlinear mapping conversion feature extraction algorithm in the offline stage, the nonlinear mapping conversion feature extraction algorithm can expand the original position feature space to a high-dimensional space for analysis, and the original position feature information is utilized to a greater extent, so that the positioning accuracy of the indoor terminal in the intelligent factory can be effectively improved.
In a preferred embodiment, the constructing an original location feature space by using the received signal strength from each indoor wireless node collected at each reference node as original location feature information specifically includes: acquiring identification information of each indoor wireless node on each reference node, acquiring the intensity of received signals from the same indoor wireless node for multiple times, and taking the average value of the intensity of the received signals from the same indoor wireless node as original position characteristic information; and when the acquisition is finished at all the reference nodes, constructing an original position feature space according to all the original position feature information.
Illustratively, identification information (such as MAC addresses) of indoor wireless nodes are collected on each reference node, and meanwhile, for each indoor wireless node such as the nth indoor wireless node, the received signal strength from the nth indoor wireless node is collected p times to obtain { RSSn,1，RSSn,2，…，RSSn,pAveraging the received signal strengths from the n-th indoor wireless node
Received information strength RSS as from nth indoor wireless nodenAnd obtaining original position characteristic information.
According to the embodiment, the average value of the received signal strength from the same indoor wireless node acquired on the reference node for multiple times is used as the original position characteristic information, so that the positioning accuracy of the indoor terminal in the intelligent factory is improved.
In a preferred embodiment, the taking the strength of the received signal from each indoor wireless node collected at the test node as the test location characteristic information specifically includes: the method comprises the steps of collecting identification information of each indoor wireless node on a test node, collecting the strength of received signals from the same indoor wireless node for multiple times, and taking the average value of the strength of the received signals from the same indoor wireless node as test position characteristic information.
Illustratively, identification information (such as MAC addresses) of indoor wireless nodes are collected on a test node, and meanwhile, for each indoor wireless node such as the nth indoor wireless node, the received signal strength from the nth indoor wireless node is collected p times to obtain { RSSn,1，RSSn,2，…，RSSn,pAveraging the received signal strengths from the n-th indoor wireless node
Received information strength RSS as from nth indoor wireless nodenAnd obtaining the characteristic information of the test position.
The average value of the received signal strength from the same indoor wireless node acquired on the test node for multiple times is used as the characteristic information of the test position, and the positioning accuracy of the indoor terminal in the intelligent factory is improved.
In a preferred embodiment, the determining, by using a weighted neighbor method, the position of the internal chamber terminal of the intelligent factory according to the target feature information and the target feature space specifically includes: calculating Euclidean distance between the target characteristic information and each original position characteristic information in the target characteristic space; and screening a plurality of original position characteristic information according to the sequence of the Euclidean distance from small to large, and determining the position of the indoor terminal according to the screened original position characteristic information.
In the embodiment, a weighted neighbor method is adopted, and the position of the indoor terminal in the intelligent factory is determined according to the target characteristic information and the target characteristic space obtained by nonlinear mapping conversion, so that the positioning accuracy of the indoor terminal in the intelligent factory is improved.
Second embodiment:
as shown in fig. 5, a second embodiment provides an intelligent factory positioning apparatus facing an indoor terminal, including: the offline module 21 specifically includes: an original location feature space construction unit 211, configured to construct an original location feature space by using the received signal strength from each indoor wireless node collected at each reference node as original location feature information; a target feature space obtaining unit 212, configured to perform nonlinear mapping conversion on the original position feature space to obtain a target feature space; the online module 22 specifically includes: a test location characteristic information acquisition unit 221 configured to use the received signal strength from each indoor wireless node acquired at the test node as test location characteristic information; a target characteristic information obtaining unit 222, configured to perform nonlinear mapping conversion on the test position characteristic information to obtain target characteristic information; and the indoor terminal positioning unit 223 is used for determining the position of the indoor terminal in the intelligent factory according to the target characteristic information and the target characteristic space by adopting a weighted neighbor method.
In a preferred embodiment, the original location feature space constructing unit 211 is further configured to lay a plurality of reference nodes in the wireless coverage area before constructing the original location feature space by using the received signal strength from each indoor wireless node collected at each reference node as the original location feature information, and construct the original location physical space according to the physical locations of all the reference nodes.
Illustratively, N reference nodes are pre-arranged through the original position feature space construction unit 211, and the physical position of each reference node is li(xi,yi) The physical locations of the N reference nodes form an original location physical space L ═ L1,l2,…,lN)TCollecting received signal strength RSS from n indoor wireless nodes on each reference node to obtain n-dimensional vector Fi＝(RSS1,RSS2,…,RSSn)TI ∈ (1, N), i.e., the original location feature information, stores the original location feature information of all reference nodes in a relevant database, and constructs an N × N-dimensional original location feature space F ═ F (F)1,F2,…,FN)TEach row vector in the matrix F represents the original location characteristic information of one reference node.
The specific steps of performing nonlinear mapping conversion on the original position feature space by the target feature space obtaining unit 212 are as follows:
1. taking the original position feature space F as an input space, and calculating a nonlinear mapping transformation matrix K according to an equation (8):
in formula (8), the width is the Gaussian kernel width; i, j is less than or equal to N; k is an NxN dimensional matrix;
2. correcting the nonlinear mapping transformation matrix K according to the formula (9) to obtain a corrected kernel matrix
In the formula (9), C is an NxN order matrix, and each element is
3. Calculating the first K maximum eigenvalues lambda of the nonlinear mapping transformation matrix K1≥λ2...λk-1≥λkAnd corresponding feature vector v1,v2,…,vk(ii) a Such as:
nonlinear mapping transformation matrix
The characteristic polynomial of (a) is | λ E-K |, i.e.
Simply to get | λ E-K | ═ λ +2)2(lambda-4), the eigenvalues of K are 4, -2, -2, the first 3 largest, respectively
Characteristic value of lambda3＝4≥λ1＝λ2＝-2；
When lambda is1＝λ2When-2, the characteristic equation set (λ E-a) X is substituted with 0, i.e.
Obtaining basic solution system X1＝[1,1,0]TI.e. k1X1Is the entire eigenvector (k) belonging to eigenvalue-21≠0)；
The same can obtain lambda3Basic solution system X when being 42＝[0,1,1]T，k2X2Is the complete eigenvector (k) belonging to eigenvalue 42≠0)；
4. V is orthogonalized by1,v2,…,vkOrthogonalizing the units to obtain a1,a2,…,ak；
5. The target feature space F' is calculated according to equation (10):
in the formula (10), a ═ a1,a2,…,ak)TIs a matrix with the dimension of N multiplied by N,
is the modified kernel matrix.
The process of collecting the feature information of the test location at the online stage is similar to the process of collecting the feature information of the original location at the offline stage, and the RSS of the received signal strength from n indoor wireless nodes is collected on the test node by the test location feature information collecting unit 221 to obtain an n-dimensional vector S ═ RSS (RSS)1,RSS2,…,RSSn)TI.e. test location signature information.
By the target feature information obtaining unit 222, the test position feature information S is subjected to the nonlinear mapping transformation processing according to equation (11), so as to obtain target feature information S':
in the formula (11), atIs a feature vector; k (F)t,S)＝Φ(Ft) Phi (S), wherein phi is a mapping function from a data space to a feature space; k (F)i,S)＝Φ(Fi)Φ(S)；KtjConverting a matrix K (F) for a non-linear mappingtS) elements of the jth row and jth column; kijConverting a matrix K (F) for a non-linear mappingiS) elements of the ith row and the jth column.
The specific steps of determining the position of the indoor terminal in the intelligent factory according to the target feature information S 'and the target feature space F' by the indoor terminal positioning unit 223 and using the weighted neighbor method are as follows:
1. calculating target feature information S 'and each original position feature information F in target feature space F' according to equation (12)iThe euclidean distance of':
in formula (12), Di(S′,Fi') can characterize S' and Fi' degree of similarity between them, the smaller the value, the more similar the two are;
2. according to Di(S′,Fi') size arrangement, find the top m < N minimum Euclidean distances DiAnd m original position characteristic information F corresponding to the m original position characteristic informationi' and reference node physical location li(xi,yi) So that it satisfies formula (13):
in the formula (13), 0 < d0Less than 1, preventing the denominator to be 0; σ is a positive number less than 1;
3. calculating the position of each indoor terminal in the intelligent factory according to the formula (14):
in this embodiment, the original location feature space is constructed by using the received signal strength from each indoor wireless node collected at each reference node as original location feature information by the original location feature space construction unit 211, and the original location feature space is subjected to nonlinear mapping conversion by the target feature space acquisition unit 212 to obtain a target feature space, the received signal strength from each indoor wireless node collected at a test node is used as test location feature information by the test location feature information collection unit 221, and the test location feature information is subjected to nonlinear mapping conversion by the target feature information acquisition unit to obtain target feature information, so that the location of the terminal in the intelligent factory is determined according to the target feature information and the target feature space by the indoor terminal positioning unit by using the weighted neighbor method, therefore, accurate positioning of the indoor terminal in the intelligent factory is achieved. In the embodiment, the original position feature information is trained by adopting the nonlinear mapping conversion feature extraction algorithm in the offline stage, the nonlinear mapping conversion feature extraction algorithm can expand the original position feature space to a high-dimensional space for analysis, and the original position feature information is utilized to a greater extent, so that the positioning accuracy of the indoor terminal in the intelligent factory can be effectively improved.
In a preferred embodiment, the constructing an original location feature space by using the received signal strength from each indoor wireless node collected at each reference node as original location feature information specifically includes: acquiring identification information of each indoor wireless node on each reference node, acquiring the intensity of received signals from the same indoor wireless node for multiple times, and taking the average value of the intensity of the received signals from the same indoor wireless node as original position characteristic information; and when the acquisition is finished at all the reference nodes, constructing an original position feature space according to all the original position feature information.
As an indicationFor example, through the original location feature space constructing unit 211, the identification information (such as the MAC address) of each indoor wireless node is collected on each reference node, and for each indoor wireless node, such as the nth indoor wireless node, the received signal strength from the nth indoor wireless node is collected p times to obtain { RSS { (RSS) }n,1，RSSn,2，…，RSSn,pAveraging the received signal strengths from the n-th indoor wireless node
Received information strength RSS as from nth indoor wireless nodenAnd obtaining original position characteristic information.
In this embodiment, the original location feature space construction unit 211 is used to take the average value of the received signal strengths from the same indoor wireless node, which are acquired on the reference node for multiple times, as the original location feature information, which is beneficial to improving the positioning accuracy of the indoor terminal in the intelligent factory.
In a preferred embodiment, the taking the strength of the received signal from each indoor wireless node collected at the test node as the test location characteristic information specifically includes: the method comprises the steps of collecting identification information of each indoor wireless node on a test node, collecting the strength of received signals from the same indoor wireless node for multiple times, and taking the average value of the strength of the received signals from the same indoor wireless node as test position characteristic information.
Illustratively, through the test location characteristic information collecting unit 221, identification information (such as MAC address) of each indoor wireless node is collected on the test node, and meanwhile, for each indoor wireless node such as the nth indoor wireless node, the received signal strength from the nth indoor wireless node is collected p times to obtain { RSSn,1，RSSn,2，…，RSSn,pAveraging the received signal strengths from the n-th indoor wireless node
Received information strength RSS as from nth indoor wireless nodenGet measuredAnd testing the position characteristic information.
In this embodiment, the average value of the received signal strengths from the same indoor wireless node, which is acquired on the test node for multiple times, is used as the test position characteristic information through the test position characteristic information acquisition unit 221, which is beneficial to improving the positioning accuracy of the indoor terminal in the intelligent factory.
In a preferred embodiment, the determining, by using a weighted neighbor method, the position of the internal chamber terminal of the intelligent factory according to the target feature information and the target feature space specifically includes: calculating Euclidean distance between the target characteristic information and each original position characteristic information in the target characteristic space; and screening a plurality of original position characteristic information according to the sequence of the Euclidean distance from small to large, and determining the position of the indoor terminal according to the screened original position characteristic information.
In this embodiment, the indoor terminal positioning unit 223 determines the position of the indoor terminal in the intelligent factory according to the target feature information and the target feature space obtained by the nonlinear mapping conversion by using the weighted neighbor method, which is favorable for improving the positioning accuracy of the indoor terminal in the intelligent factory.
In summary, the embodiment of the present invention has the following advantages:
the method comprises the steps that identification information and received signal strength corresponding to each indoor wireless node collected on each reference node are used as original position feature information in an off-line stage, an original position feature space is constructed, nonlinear mapping conversion is conducted on the original position feature space, a target feature space is obtained, the identification information and the received signal strength corresponding to each indoor wireless node collected on a test node are used as test position feature information in an on-line stage, the nonlinear mapping conversion is conducted on the test position feature information, the target feature information is obtained, and the position of a terminal in an intelligent factory inner chamber is determined according to the target feature information and the target feature space by adopting a weighted neighbor method, so that accurate positioning of the terminal in the intelligent factory inner chamber is achieved. According to the embodiment of the invention, the original position feature information is trained by adopting the nonlinear mapping conversion feature extraction algorithm in an off-line stage, the nonlinear mapping conversion feature extraction algorithm can expand the original position feature space to a high-dimensional space for analysis, and the original position feature information is utilized to a greater extent, so that the positioning accuracy of the indoor terminal in the intelligent factory can be effectively improved.
While the foregoing is directed to the preferred embodiment of the present invention, it will be understood by those skilled in the art that various changes and modifications may be made without departing from the spirit and scope of the invention.
It will be understood by those skilled in the art that all or part of the processes of the above embodiments may be implemented by hardware related to instructions of a computer program, and the computer program may be stored in a computer readable storage medium, and when executed, may include the processes of the above embodiments. The storage medium may be a magnetic disk, an optical disk, a Read-Only Memory (ROM), a Random Access Memory (RAM), or the like.
Claims (10)
1. An intelligent factory positioning method for an indoor terminal is characterized by comprising the following steps:
s1, an offline stage, specifically including:
the method comprises the steps that the strength of received signals from all indoor wireless nodes collected on each reference node is used as original position characteristic information, and an original position characteristic space is constructed;
carrying out nonlinear mapping conversion on the original position feature space to obtain a target feature space;
s2, an online stage, specifically comprising:
the received signal strength from each indoor wireless node collected on the test node is used as the characteristic information of the test position;
carrying out nonlinear mapping conversion on the test position characteristic information to obtain target characteristic information;
and determining the position of the terminal in the intelligent factory according to the target characteristic information and the target characteristic space by adopting a weighted neighbor method.
2. The intelligent factory positioning method for indoor terminals as claimed in claim 1, wherein before said constructing a raw location feature space using the received signal strength from each indoor wireless node collected at each reference node as raw location feature information, further comprising:
and laying a plurality of reference nodes in a wireless coverage area, and constructing an original position physical space according to the physical positions of all the reference nodes.
3. The intelligent factory positioning method for indoor terminals as claimed in claim 1, wherein the original location feature space is constructed by using the received signal strength from each indoor wireless node collected at each reference node as original location feature information, specifically:
acquiring identification information of each indoor wireless node on each reference node, acquiring the strength of received signals from the same indoor wireless node for multiple times, and taking the average value of the strength of the received signals from the same indoor wireless node as the original position characteristic information;
and when the acquisition is finished at all the reference nodes, constructing the original position feature space according to all the original position feature information.
4. The intelligent factory positioning method for the indoor terminal according to claim 1, wherein the received signal strength from each indoor wireless node collected at the test node is used as the test location characteristic information, specifically:
and acquiring identification information of each indoor wireless node on the test node, acquiring the strength of received signals from the same indoor wireless node for multiple times, and taking the average value of the strength of the received signals from the same indoor wireless node as the characteristic information of the test position.
5. The intelligent factory positioning method facing the indoor terminal as claimed in claim 1, wherein the determining the position of the indoor terminal in the intelligent factory according to the target feature information and the target feature space by using the weighted neighbor method specifically comprises:
calculating Euclidean distance between the target feature information and each original position feature information in the target feature space;
and screening a plurality of original position characteristic information according to the sequence of the Euclidean distance from small to large, and determining the position of the indoor terminal according to the screened original position characteristic information.
6. The utility model provides a towards indoor terminal's intelligent mill positioner which characterized in that includes:
the offline module specifically comprises:
the system comprises an original position characteristic space construction unit, a position information acquisition unit and a position information acquisition unit, wherein the original position characteristic space construction unit is used for constructing an original position characteristic space by taking the received signal strength from each indoor wireless node acquired on each reference node as original position characteristic information;
the target characteristic space obtaining unit is used for carrying out nonlinear mapping conversion on the original position characteristic space to obtain a target characteristic space;
the online module specifically comprises:
the test position characteristic information acquisition unit is used for taking the received signal strength from each indoor wireless node acquired on the test node as test position characteristic information;
the target characteristic information acquisition unit is used for carrying out nonlinear mapping conversion on the test position characteristic information to obtain target characteristic information;
and the indoor terminal positioning unit is used for determining the position of the indoor terminal in the intelligent factory according to the target characteristic information and the target characteristic space by adopting a weighted neighbor method.
7. The intelligent factory positioning apparatus for indoor terminals as claimed in claim 6, wherein said home location feature space constructing unit is further configured to lay a plurality of said reference nodes in a wireless coverage area before constructing a home location feature space by using the received signal strength from each indoor wireless node collected at each reference node as home location feature information, and construct a home location physical space according to the physical locations of all said reference nodes.
8. The intelligent factory positioning apparatus for indoor terminals as claimed in claim 6, wherein the raw location feature space is constructed by using the received signal strength from each indoor wireless node collected at each reference node as raw location feature information, specifically:
acquiring identification information of each indoor wireless node on each reference node, acquiring the strength of received signals from the same indoor wireless node for multiple times, and taking the average value of the strength of the received signals from the same indoor wireless node as the original position characteristic information;
and when the acquisition is finished at all the reference nodes, constructing the original position feature space according to all the original position feature information.
9. The intelligent factory positioning apparatus for indoor terminals as claimed in claim 6, wherein the received signal strength from each indoor wireless node collected at a test node is used as the test location characteristic information, specifically:
and acquiring identification information of each indoor wireless node on the test node, acquiring the strength of received signals from the same indoor wireless node for multiple times, and taking the average value of the strength of the received signals from the same indoor wireless node as the characteristic information of the test position.
10. The intelligent factory positioning apparatus facing an indoor terminal as claimed in claim 6, wherein the determining the position of the indoor terminal in the intelligent factory according to the target feature information and the target feature space by using the weighted neighbor method specifically comprises:
calculating Euclidean distance between the target feature information and each original position feature information in the target feature space;
and screening a plurality of original position characteristic information according to the sequence of the Euclidean distance from small to large, and determining the position of the indoor terminal according to the screened original position characteristic information.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010800216.1A CN112020137A (en) | 2020-08-10 | 2020-08-10 | Intelligent factory positioning method and device for indoor terminal |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010800216.1A CN112020137A (en) | 2020-08-10 | 2020-08-10 | Intelligent factory positioning method and device for indoor terminal |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112020137A true CN112020137A (en) | 2020-12-01 |
Family
ID=73504240
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010800216.1A Pending CN112020137A (en) | 2020-08-10 | 2020-08-10 | Intelligent factory positioning method and device for indoor terminal |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112020137A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130065605A1 (en) * | 2011-09-13 | 2013-03-14 | Piotr Mirowski | KL-Divergence Kernel Regression For Non-Gaussian Fingerprint Based Localization |
| CN103686999A (en) * | 2013-12-12 | 2014-03-26 | 中国石油大学(华东) | Indoor wireless locating method based on WiFi signals |
| CN105813192A (en) * | 2014-12-29 | 2016-07-27 | 上海奕通信息科技有限公司 | Combined fingerprint feature-based WiFi fingerprint positioning method |
| CN106714302A (en) * | 2017-01-23 | 2017-05-24 | 吉林大学 | Indoor positioning device based on BP-Landmarc neural network and control method |
| CN108900970A (en) * | 2018-07-06 | 2018-11-27 | 中国民航大学 | A kind of terminal indoor orientation method returning Kernel discriminant analysis based on spectrum |
| CN110958583A (en) * | 2019-10-31 | 2020-04-03 | 西南科技大学 | Indoor positioning method based on position and strength dual clustering |
| US20200142045A1 (en) * | 2018-06-04 | 2020-05-07 | Central China Normal University | Fingerprint positioning method and system in smart classroom |
-
2020
- 2020-08-10 CN CN202010800216.1A patent/CN112020137A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20130065605A1 (en) * | 2011-09-13 | 2013-03-14 | Piotr Mirowski | KL-Divergence Kernel Regression For Non-Gaussian Fingerprint Based Localization |
| CN103686999A (en) * | 2013-12-12 | 2014-03-26 | 中国石油大学(华东) | Indoor wireless locating method based on WiFi signals |
| CN105813192A (en) * | 2014-12-29 | 2016-07-27 | 上海奕通信息科技有限公司 | Combined fingerprint feature-based WiFi fingerprint positioning method |
| CN106714302A (en) * | 2017-01-23 | 2017-05-24 | 吉林大学 | Indoor positioning device based on BP-Landmarc neural network and control method |
| US20200142045A1 (en) * | 2018-06-04 | 2020-05-07 | Central China Normal University | Fingerprint positioning method and system in smart classroom |
| CN108900970A (en) * | 2018-07-06 | 2018-11-27 | 中国民航大学 | A kind of terminal indoor orientation method returning Kernel discriminant analysis based on spectrum |
| CN110958583A (en) * | 2019-10-31 | 2020-04-03 | 西南科技大学 | Indoor positioning method based on position and strength dual clustering |
Non-Patent Citations (7)
| Title |
|---|
| 乔柯: "飞行器的声探测与定位技术研究", 《中国优秀硕士论文电子期刊网》 * |
| 张彬: "基于WLAN的室内指纹定位关键技术研究", 《中国优秀硕士论文电子期刊网》 * |
| 李华亮等: "基于核函数特征提取的室内定位算法研究", 《通信学报》 * |
| 李飞腾: "基于信号特征提取的WLAN室内定位算法研究", 《中国优秀硕士论文电子期刊网》 * |
| 薛卫星等: "一种RSSI不同特征值融合的提取算法", 《测绘工程》 * |
| 赵永生: "基于WLAN信号强度指纹的室内定位技术研究", 《中国优秀硕士论文电子期刊网》 * |
| 韩萍: "SAR 自动目标识别及相关技术研究", 《中国优秀博士论文电子期刊网》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108037520B (en) | Neural network-based direct positioning deviation correction method under array amplitude-phase error condition | |
| CN115018021B (en) | Machine room abnormity detection method and device based on graph structure and abnormity attention mechanism | |
| Du et al. | Shape recognition based on radial basis probabilistic neural network and application to plant species identification | |
| CN103889051B (en) | Indoor WLAN fingerprint positioning methods based on AP ID filterings and Kalman filtering | |
| CN110929847A (en) | Converter transformer fault diagnosis method based on deep convolutional neural network | |
| CN114449452B (en) | Wi-Fi indoor positioning method based on CNN-RNN | |
| CN104853435A (en) | Probability based indoor location method and device | |
| CN111586605B (en) | KNN indoor target positioning method based on adjacent weighted self-adaptive k value | |
| CN111050282A (en) | Multi-time fuzzy inference weighted KNN positioning method | |
| CN110674940B (en) | Multi-index anomaly detection method based on neural network | |
| CN109068349B (en) | Indoor intrusion detection method based on small sample iterative migration | |
| CN110673568A (en) | Method and system for determining fault sequence of industrial equipment in glass fiber manufacturing industry | |
| Terbuch et al. | Hybrid machine learning for anomaly detection in industrial time-series measurement data | |
| CN112333652B (en) | WLAN indoor positioning method and device and electronic equipment | |
| CN112020137A (en) | Intelligent factory positioning method and device for indoor terminal | |
| CN111757257A (en) | Dynamic fuzzy matching indoor positioning method for overcoming equipment difference | |
| CN116470504A (en) | Self-healing control system of distributed power distribution network | |
| CN114814776B (en) | PD radar target detection method based on graph attention network and transfer learning | |
| CN112946567B (en) | Moving target fingerprint indoor positioning method based on domain antagonism neural network | |
| CN114374931B (en) | Fingerprint positioning method based on metric learning of neighbor component analysis | |
| CN115563312A (en) | Medicine-disease-target triple target entity completion method and application | |
| CN112348175B (en) | Method for performing feature engineering based on reinforcement learning | |
| CN112437485A (en) | Positioning method and device of fingerprint space interpolation method based on neural network | |
| Mathew et al. | Disease Classification in Bell Pepper Plants Based on Deep Learning Network Architecture | |
| Liu et al. | Surface defect classification of steels based on ensemble of extreme learning machines |
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: 20201201 |
|
| RJ01 | Rejection of invention patent application after publication |