CN115175306B - Indoor positioning method of electric power Internet of things based on convolutional neural network - Google Patents
Indoor positioning method of electric power Internet of things based on convolutional neural network Download PDFInfo
- Publication number
- CN115175306B CN115175306B CN202210729213.2A CN202210729213A CN115175306B CN 115175306 B CN115175306 B CN 115175306B CN 202210729213 A CN202210729213 A CN 202210729213A CN 115175306 B CN115175306 B CN 115175306B
- Authority
- CN
- China
- Prior art keywords
- rssi
- neural network
- convolutional neural
- training
- sample
- 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.)
- Active
Links
- 238000013527 convolutional neural network Methods 0.000 title claims abstract description 39
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000010586 diagram Methods 0.000 claims abstract description 15
- 238000010276 construction Methods 0.000 claims abstract description 8
- 238000009499 grossing Methods 0.000 claims description 12
- 238000005070 sampling Methods 0.000 claims description 7
- 238000007476 Maximum Likelihood Methods 0.000 claims description 6
- 238000005516 engineering process Methods 0.000 abstract description 8
- 238000000605 extraction Methods 0.000 abstract description 5
- 238000004891 communication Methods 0.000 description 3
- QVFWZNCVPCJQOP-UHFFFAOYSA-N chloralodol Chemical compound CC(O)(C)CC(C)OC(O)C(Cl)(Cl)Cl QVFWZNCVPCJQOP-UHFFFAOYSA-N 0.000 description 2
- 241000764238 Isis Species 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 239000000284 extract Substances 0.000 description 1
- 238000004377 microelectronic Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
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
- H04W64/006—Locating users or terminals or network equipment for network management purposes, e.g. mobility management with additional information processing, e.g. for direction or speed determination
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y10/00—Economic sectors
- G16Y10/35—Utilities, e.g. electricity, gas or water
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
- G16Y40/00—IoT characterised by the purpose of the information processing
- G16Y40/60—Positioning; Navigation
-
- 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
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Computing Systems (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Mathematical Physics (AREA)
- Artificial Intelligence (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Electromagnetism (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Software Systems (AREA)
- Quality & Reliability (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Economics (AREA)
- General Business, Economics & Management (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Image Analysis (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
Abstract
The invention discloses an indoor positioning method of an electric power Internet of things based on a convolutional neural network, which comprises the following steps: s1, in an offline fingerprint library construction stage, a probability density function diagram is generated for each acquisition position by using a nuclear density estimation method and is used for constructing an RSSI statistical sample fingerprint library; s2, training the convolutional neural network by using a statistical sample fingerprint library in a training stage; s3, in an online positioning stage, according to the RSSI samples acquired by equipment to be positioned, predicting the position of the equipment based on a convolutional neural network algorithm. The spatial statistical distribution characteristic of the signal intensity and the spatial feature extraction advantage of the convolutional neural network are fully utilized, and an accurate and stable indoor positioning technology can be provided in the application scene of the electric power Internet of things.
Description
Technical Field
The invention relates to the technical field of wireless sensor networks, in particular to an indoor positioning method of an electric power Internet of things based on a convolutional neural network.
Background
With the continuous progress of wireless communication, microelectronics and other technologies, the application of the internet of things in power systems is becoming wider and wider. A variety of wireless communication technologies have been applied to power systems such as LoRa, NB-IoT, OFDM, etc. In an electrical power system, such as a power plant, a substation, etc., many devices are located indoors and location information for these devices is needed when the system is running. However, the conventionally used outdoor positioning technology, such as GPS, is affected by obstacles such as walls, and is difficult to be applied to indoor environments, so indoor positioning in the electric power internet of things is a problem that needs to be solved.
Indoor positioning is being studied more and more at home and abroad. RSSI is the most commonly used indicator in indoor positioning because it is readily available in many wireless communication devices (e.g., loRa, NB-IoT, zigBee, OFDM). Although various indoor positioning methods based on RSSI exist, the methods are convenient to improve in precision, stability and the like, and in a power system, a great deal of problems of interference, noise, shielding and the like exist, so that a high-precision and stable indoor positioning method facing the power Internet of things is needed.
Disclosure of Invention
In order to overcome the defects, the invention aims to provide the indoor positioning method of the electric power Internet of things based on the convolutional neural network, which fully utilizes the spatial statistical distribution characteristic of the signal intensity and the spatial feature extraction advantage of the convolutional neural network and can provide an accurate and stable indoor positioning technology in the application scene of the electric power Internet of things.
In order to achieve the above purpose, the invention adopts the following technical scheme:
an indoor positioning method of an electric power Internet of things based on a convolutional neural network comprises the following steps:
S1, in an offline fingerprint library construction stage, a probability density function diagram is generated for each acquisition position by using a nuclear density estimation method and is used for constructing an RSSI statistical sample fingerprint library;
s2, training the convolutional neural network by using a statistical sample fingerprint library in a training stage;
S3, in an online positioning stage, according to the RSSI samples acquired by equipment to be positioned, predicting the position of the equipment based on a convolutional neural network algorithm.
Optionally, the step of the offline fingerprint library construction stage includes:
S11, let L= (L 1,…,ln) be n RSSI acquisition positions, wherein L i=(xi,yi,zi is the three-dimensional coordinate of a position L i; let a= (a 1,…,ah) be the set of APs in the positioning region, where a i is the i-th AP; collecting m RSSI samples at any RSSI collecting position l i and recording the RSSI sample set as Wherein/>For the j-th RSSI sample,/>The RSSI value of the kth AP in the jth sample at the acquisition position is obtained; after the RSSI samples are collected for each position to be collected, an original RSSI sample library/>
S12, given an original RSSI sample library R, training a kernel density estimator for a sample R i (i is more than or equal to 1 and less than or equal to n) at any RSSI acquisition position:
s121, give Let/>One kernel density estimator needs to be trained for each AP at RSSI acquisition position l i, where the kernel density estimator f i,k (r) for the kth AP is given by:
Wherein K (·) is a kernel function and u is a smoothing parameter; estimating by using a Gaussian kernel, and determining smoothing parameters by using a maximum likelihood estimation method; thus, there are a total of h kernel density estimators, denoted F i=(fi,1(r),…,fi,h (r), at RSSI acquisition location l i;
S13, let r min and r max be possible received RSSI minimum and maximum values, and let lambda be step size parameter; then a horizontal axis interval sample point x= (r min,rmin+λ,rmin+2λ,…,rmax) of the RSSI probability density distribution can be created; for any RSSI acquisition position l i, the following operations are performed:
S131, for any AP a k (k is more than or equal to 1 and less than or equal to h), generating corresponding probabilities Yi,k=(fi,k(rmin),fi,k(rmin+λ),fi,k(rmin+2λ),…,fi,k(rmax)); of all sampling points in X by using a kernel density estimator f i,k (r), and then, taking X= (r min,rmin+λ,rmin+2λ,…,rmax) as a horizontal axis coordinate ,Yi,k=(fi,k(rmin),fi,k(rmin+λ),fi,k(rmin+2λ),…,fi,k(rmax)) as a vertical axis coordinate to generate a line graph G i,k; obviously, there are h line graphs at any RSSI acquisition position l i, and the line graphs are recorded as G i=(Gi,1,…,Gi,h in a set;
And S132, after the operation of S131 is performed on all RSSI acquisition positions, a final offline RSSI fingerprint library G= (G 1,…,Gn) can be obtained.
Optionally, the step of training phase includes:
S21, constructing a convolutional neural network training set by using an offline RSSI fingerprint library G, taking i as a label of any RSSI acquisition position l i, and transversely splicing h line graphs to form an image which is marked as P i and is used as characteristic data, so that training data at the acquisition position is marked as D i=(i,Pi). Finally obtaining training set as D= (D 1,…Dn);
S22, constructing a convolutional neural network n-type classifier by using an existing framework, and marking the classifier as f;
s23, training the classifier f on the training set D.
Optionally, the step of the online positioning stage includes:
S31, for a certain device d to be positioned, continuously collecting m RSSI samples, and recording the RSSI sample set as Wherein/>For the j-th RSSI sample,/>The RSSI value of the kth AP in the jth sample at the acquisition position is obtained;
s32, training a kernel density estimator on the device d to be positioned for each AP, wherein the kernel density estimator f d,k (r) of the kth AP is given by:
Wherein K (·) is a kernel function and u is a smoothing parameter; considering that the signal intensity presents Gaussian distribution, the invention adopts Gaussian kernel to estimate, and the smoothing parameter is determined by adopting a maximum likelihood estimation method; thus, there are a total of h kernel density estimators, denoted as F d=(fd,1(r),…,fd,h (r)), on the device d to be located;
S33, for any AP a k (k is more than or equal to 1 and less than or equal to h), generating corresponding probabilities Yd,k=(fd,k(rmin),fd,k(rmin+λ),fd,k(rmin+2λ),…,fd,k(rmax)); of all sampling points in X by using a kernel density estimator f d,k (r), and then, taking X= (r min,rmin+λ,rmin+2λ,…,rmax) as a horizontal axis coordinate ,Yd,k=(fd,k(rmin),fd,k(rmin+λ),fd,k(rmin+2λ),…,fd,k(rmax)) as a vertical axis coordinate to generate a line graph G d,k; h line diagrams are arranged on the equipment d to be positioned, and the line diagrams are marked as G d=(Gd,1,…,Gd,h;
S34, positioning by using the trained convolutional neural network f.
Optionally, the step of using the trained convolutional neural network f to perform positioning includes:
S341, transversely splicing the h-piece line graphs in the G d to form an image which is recorded as P d as characteristic data;
S342, transmitting P d into f for classification, taking K positions with maximum f output layer softmax function estimated value, and recording the K positions with maximum probability as respectively Wherein/>
Respectively isIs a three-dimensional coordinate of (2);
s343, the positioning result of the final device d to be positioned is
The invention has the positive beneficial effects that:
Aiming at the characteristics of complex radio frequency environment, serious interference and the like of the electric power Internet of things, the spatial distribution characteristics of the signal intensity are extracted by utilizing the convolutional neural network, and the method is used for accurately and robustly positioning the indoor electric power equipment. The method comprises the following steps: an offline fingerprint library construction stage, a training stage and an online positioning stage. The spatial statistical distribution characteristic of the signal intensity and the spatial feature extraction advantage of the convolutional neural network are fully utilized, and an accurate and stable indoor positioning technology can be provided in the application scene of the electric power Internet of things.
Drawings
Fig. 1 is a schematic block diagram of an indoor positioning method of an electric power internet of things based on a convolutional neural network provided in embodiment 1 of the present invention;
Fig. 2 is a schematic diagram of an original RSSI acquisition provided in step S11 of embodiment 1 of the present invention;
FIG. 3 is a diagram illustrating the creation of a line graph according to step S131 of embodiment 1 of the present invention;
FIG. 4 is a schematic diagram of training set generation provided in step S21 of embodiment 1 of the present invention;
fig. 5 is a schematic diagram of final positioning generation of an output layer provided in step S342 of embodiment 1 of the present invention.
Detailed Description
The invention will be further described with reference to the following embodiments.
Example 1
As shown in fig. 1 to 5, an indoor positioning method of an electric power internet of things based on a convolutional neural network includes the steps of:
S1, in an offline fingerprint library construction stage, a probability density function diagram is generated for each acquisition position by using a nuclear density estimation method and is used for constructing an RSSI statistical sample fingerprint library;
s2, training the convolutional neural network by using a statistical sample fingerprint library in a training stage;
S3, in an online positioning stage, according to the RSSI samples acquired by equipment to be positioned, predicting the position of the equipment based on a convolutional neural network algorithm.
Aiming at the characteristics of complex radio frequency environment, serious interference and the like of the electric power Internet of things, the invention utilizes the convolutional neural network to extract the spatial distribution characteristic of the signal intensity for accurate and robust indoor positioning of the electric power equipment, fully utilizes the spatial statistical distribution characteristic of the signal intensity and the spatial characteristic extraction advantage of the convolutional neural network, and can provide an accurate and stable indoor positioning technology in the application scene of the electric power Internet of things.
The steps of the off-line fingerprint database construction stage comprise:
S11, let L= (L 1,…,ln) be n RSSI acquisition positions, wherein L i=(xi,yi,zi is the three-dimensional coordinate of a position L i; let a= (a 1,…,ah) be the set of APs in the positioning region, where a i is the i-th AP; collecting m RSSI samples at any RSSI collecting position l i and recording the RSSI sample set as Wherein/>For the j-th RSSI sample,/>The RSSI value of the kth AP in the jth sample at the acquisition position is obtained; after the RSSI samples are collected for each position to be collected, an original RSSI sample library/>
S12, given an original RSSI sample library R, training a kernel density estimator for a sample R i (i is more than or equal to 1 and less than or equal to n) at any RSSI acquisition position:
s121, give Let/>One kernel density estimator needs to be trained for each AP at RSSI acquisition position l i, where the kernel density estimator f i,k (r) for the kth AP is given by:
Wherein K (·) is a kernel function and u is a smoothing parameter; taking the Gaussian distribution of the signal intensity into consideration, estimating by adopting a Gaussian kernel, and determining a smoothing parameter by adopting a maximum likelihood estimation method; thus, there are a total of h kernel density estimators, denoted F i=(fi,1(r),…,fi,h (r), at RSSI acquisition location l i;
S13, let r min and r max be possible received RSSI minimum and maximum values, and let lambda be step size parameter; then a horizontal axis interval sample point x= (r min,rmin+λ,rmin+2λ,…,rmax) of the RSSI probability density distribution can be created; for any RSSI acquisition position l i, the following operations are performed:
S131, for any AP a k (k is more than or equal to 1 and less than or equal to h), generating corresponding probabilities Yi,k=(fi,k(rmin),fi,k(rmin+λ),fi,k(rmin+2λ),…,fi,k(rmax)); of all sampling points in X by using a kernel density estimator f i,k (r), and then, taking X= (r min,rmin+λ,rmin+2λ,…,rmax) as a horizontal axis coordinate ,Yi,k=(fi,k(rmin),fi,k(rmin+λ),fi,k(rmin+2λ),…,fi,k(rmax)) as a vertical axis coordinate to generate a line graph G i,k; obviously, there are h line graphs at any RSSI acquisition position l i, and the line graphs are recorded as G i=(Gi,1,…,Gi,h in a set;
And S132, after the operation of S131 is performed on all RSSI acquisition positions, a final offline RSSI fingerprint library G= (G 1,…,Gn) can be obtained.
The existing indoor positioning algorithm is mostly positioned based on the mean value, and the statistics and the spatial distribution characteristics of RSSI samples are ignored, so that the invention extracts the statistics characteristics of the original samples aiming at each RSSI acquisition position of the spatial distribution by using a nuclear density estimation method.
The training phase comprises the following steps:
S21, constructing a convolutional neural network training set by using an offline RSSI fingerprint library G, and specifically: for any RSSI acquisition position l i, i is used as its label, and the h line graphs are transversely spliced to form an image, which is denoted as P i as characteristic data, so that training data at the acquisition position is denoted as D i=(i,Pi). Finally obtaining training set as D= (D 1,…Dn);
s22, constructing a convolutional neural network n-type by using an existing framework (such as AlexNet, VGGNet and the like), and marking the classifier as f;
s23, training the classifier f on the training set D.
The online positioning stage comprises the following steps:
S31, for a certain device d to be positioned, continuously collecting m RSSI samples, and recording the RSSI sample set as Wherein/>For the j-th RSSI sample,/>The RSSI value of the kth AP in the jth sample at the acquisition position is obtained;
s32, training a kernel density estimator on the device d to be positioned for each AP, wherein the kernel density estimator f d,k (r) of the kth AP is given by:
Wherein K (·) is a kernel function and u is a smoothing parameter; considering that the signal intensity presents Gaussian distribution, the invention adopts Gaussian kernel to estimate, and the smoothing parameter is determined by adopting a maximum likelihood estimation method; thus, there are a total of h kernel density estimators, denoted as F d=(fd,1(r),…,fd,h (r)), on the device d to be located;
S33, for any AP a k (k is more than or equal to 1 and less than or equal to h), generating corresponding probabilities Yd,k=(fd,k(rmin),fd,k(rmin+λ),fd,k(rmin+2λ),…,fd,k(rmax)); of all sampling points in X by using a kernel density estimator f d,k (r), and then, taking X= (r min,rmin+λ,rmin+2λ,…,rmax) as a horizontal axis coordinate ,Yd,k=(fd,k(rmin),fd,k(rmin+λ),fd,k(rmin+2λ),…,fd,k(rmax)) as a vertical axis coordinate to generate a line graph G d,k; h line diagrams are arranged on the equipment d to be positioned, and the line diagrams are marked as G d=(Gd,1,…,Gd,h;
S34, positioning by using the trained convolutional neural network f.
The step of positioning by using the trained convolutional neural network f comprises the following steps:
S341, transversely splicing the h-piece line graphs in the G d to form an image which is recorded as P d as characteristic data;
S342, transmitting P d into f for classification, taking K positions with maximum f output layer softmax function estimated value instead of taking the classification result as the estimated position, and recording the K positions with maximum probability as respectively Wherein the method comprises the steps of Respectively/>Is a three-dimensional coordinate of (2);
s343, the positioning result of the final device d to be positioned is
Aiming at the characteristics of complex radio frequency environment, serious interference and the like of the electric power Internet of things, the spatial distribution characteristics of the signal intensity are extracted by utilizing the convolutional neural network, and the method is used for accurately and robustly positioning the indoor electric power equipment. The spatial statistical distribution characteristic of the signal intensity and the spatial feature extraction advantage of the convolutional neural network are fully utilized, and an accurate and stable indoor positioning technology can be provided in the application scene of the electric power Internet of things.
The concrete steps are as follows:
(1) Unlike available method, which adopts average value to estimate, the present invention utilizes kernel density estimating method to extract the probability statistic features of signal strength in each sampling point in space and the trained probability statistic model retains more features for positioning in image mode.
(2) In addition, the invention does not directly use the result of the convolutional neural network to locate, but uses the mean value of K positions with the maximum estimated value of the softmax function of the output layer to locate, thereby improving the stability of the locating system.
Finally, it is noted that the above-mentioned embodiments are merely for illustrating the technical solution of the present invention, and that other modifications and equivalents thereof by those skilled in the art should be included in the scope of the claims of the present invention without departing from the spirit and scope of the technical solution of the present invention.
Claims (2)
1. The indoor positioning method of the electric power Internet of things based on the convolutional neural network is characterized by comprising the following steps:
S1, in an offline fingerprint library construction stage, a probability density function diagram is generated for each acquisition position by using a nuclear density estimation method and is used for constructing an RSSI statistical sample fingerprint library;
s2, training the convolutional neural network by using a statistical sample fingerprint library in a training stage;
S3, in an online positioning stage, predicting the position of equipment based on a convolutional neural network algorithm according to an RSSI sample acquired by the equipment to be positioned;
The steps of the off-line fingerprint database construction stage comprise:
s11, let L= (L 1,…,ln) be n RSSI acquisition positions, wherein L i=(xi,yi,zi is the three-dimensional coordinate of a position L i; let a= (a 1,…,ah) be the set of APs in the positioning region, where a i is the i-th AP; collecting m RSSI samples at any RSSI collecting position l i and recording the RSSI sample set as Wherein/>For the j-th RSSI sample,/>The RSSI value of the kth AP in the jth sample at the acquisition position is obtained; after the RSSI samples are collected for each position to be collected, an original RSSI sample library/>;
S12, giving an original RSSI sample library R, and aiming at samples at any RSSI acquisition position(1. Ltoreq.i.ltoreq.n) training the kernel density estimator:
s121, give Let/>; One kernel density estimator needs to be trained for each AP at RSSI acquisition position l i, where the kernel density estimator f i,k (r) for the kth AP is given by:
,
wherein K (·) is a kernel function and u is a smoothing parameter; estimating by using a Gaussian kernel, and determining smoothing parameters by using a maximum likelihood estimation method; thus, there are a total of h kernel density estimators, denoted F i=(fi,1(r),…, fi,h (r), at RSSI acquisition location l i;
S13, let r min and r max be possible received RSSI minimum and maximum values, and let lambda be step size parameter; then a horizontal axis interval sample point x= (r min, rmin+λ, rmin+2λ,…, rmax) of the RSSI probability density distribution can be created; for any RSSI acquisition position l i, the following operations are performed:
S131, for any AP a k (k is more than or equal to 1 and less than or equal to h), generating corresponding probabilities Yi,k=( fi,k(rmin), fi,k(rmin+λ), fi,k(rmin+2λ),…, fi,k(rmax)); of all sampling points in X by using a kernel density estimator f i,k (r), and then, taking X= (r min, rmin+λ, rmin+2λ,…, rmax) as a horizontal axis coordinate ,Yi,k=( fi,k(rmin), fi,k(rmin+λ), fi,k(rmin+2λ),…, fi,k(rmax)) as a vertical axis coordinate to generate a line graph G i,k; obviously, there are h line graphs at any RSSI acquisition position l i, and the line graphs are recorded as G i=(Gi,1,…, Gi,h in a set;
s132, after the operation of S131 is performed on all RSSI acquisition positions, a final offline RSSI fingerprint library G= (G 1,…,Gn) can be obtained;
the online positioning stage comprises the following steps:
S31, for a certain device d to be positioned, continuously collecting m RSSI samples, and recording the RSSI sample set as Wherein/>For the j-th RSSI sample,/>The RSSI value of the kth AP in the jth sample at the acquisition position is obtained;
s32, training a kernel density estimator on the device d to be positioned for each AP, wherein the kernel density estimator f d,k (r) of the kth AP is given by:
Wherein K (·) is a kernel function and u is a smoothing parameter; taking the Gaussian distribution of the signal intensity into consideration, estimating by adopting a Gaussian kernel, and determining a smoothing parameter by adopting a maximum likelihood estimation method; thus, there are a total of h kernel density estimators, denoted as F d=( fd,1(r),…, fd,h (r)), on the device d to be located;
S33, for any AP a k (k is more than or equal to 1 and less than or equal to h), generating corresponding probabilities Yd,k=( fd,k(rmin), fd,k(rmin+λ), fd,k(rmin+2λ),…, fd,k(rmax)); of all sampling points in X by using a kernel density estimator f d,k (r), and then, taking X= (r min, rmin+λ, rmin+2λ,…, rmax) as a horizontal axis coordinate ,Yd,k=( fd,k(rmin), fd,k(rmin+λ), fd,k(rmin+2λ),…, fd,k(rmax)) as a vertical axis coordinate to generate a line graph G d,k; h line diagrams are arranged on the equipment d to be positioned, and the line diagrams are marked as G d=(Gd,1,…, Gd,h;
s34, positioning by using a trained convolutional neural network f;
the step of positioning by using the trained convolutional neural network f comprises the following steps:
S341, transversely splicing the h-piece line graphs in the G d to form an image which is recorded as P d as characteristic data;
S342, transmitting P d into f for classification, taking K positions with maximum f output layer softmax function estimated value, and recording the K positions with maximum probability as respectively Wherein/>,/>Respectively/>Is a three-dimensional coordinate of (2);
s343, the positioning result of the final device d to be positioned is 。
2. The indoor positioning method of the electric internet of things based on the convolutional neural network as set forth in claim 1, wherein the training phase comprises the steps of:
S21, constructing a convolutional neural network training set by using an offline RSSI fingerprint library G, taking i as a label of any RSSI acquisition position l i, transversely splicing h line graphs to form an image which is marked as P i and is used as characteristic data, marking training data on the acquisition position as D i=(i, Pi, and finally obtaining the training set as D= (D 1,…Dn);
s22, constructing a convolutional neural network classifier by using an existing framework, and marking the classifier as f;
s23, training the classifier f on the training set D.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210729213.2A CN115175306B (en) | 2022-06-24 | 2022-06-24 | Indoor positioning method of electric power Internet of things based on convolutional neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210729213.2A CN115175306B (en) | 2022-06-24 | 2022-06-24 | Indoor positioning method of electric power Internet of things based on convolutional neural network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN115175306A CN115175306A (en) | 2022-10-11 |
CN115175306B true CN115175306B (en) | 2024-05-07 |
Family
ID=83486988
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210729213.2A Active CN115175306B (en) | 2022-06-24 | 2022-06-24 | Indoor positioning method of electric power Internet of things based on convolutional neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115175306B (en) |
Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108282743A (en) * | 2018-03-05 | 2018-07-13 | 桂林理工大学 | Indoor orientation method, apparatus and system |
CN108924737A (en) * | 2018-06-20 | 2018-11-30 | 北京三快在线科技有限公司 | Localization method, device, equipment and computer readable storage medium |
US10620290B1 (en) * | 2019-02-19 | 2020-04-14 | Ideal Industries Lighting Llc | Indoor positioning system |
US10716089B1 (en) * | 2019-06-03 | 2020-07-14 | Mapsted Corp. | Deployment of trained neural network based RSS fingerprint dataset |
CN111669820A (en) * | 2020-05-27 | 2020-09-15 | 南京邮电大学 | Density peak value abnormity detection method and intelligent passive indoor positioning method |
CN112153736A (en) * | 2020-09-14 | 2020-12-29 | 南京邮电大学 | Personnel action identification and position estimation method based on channel state information |
CN112329685A (en) * | 2020-11-16 | 2021-02-05 | 常州大学 | Method for detecting crowd abnormal behaviors through fusion type convolutional neural network |
CN112712557A (en) * | 2020-12-17 | 2021-04-27 | 上海交通大学 | Super-resolution CIR indoor fingerprint positioning method based on convolutional neural network |
WO2021103027A1 (en) * | 2019-11-30 | 2021-06-03 | Beijing Didi Infinity Technology And Development Co., Ltd. | Base station positioning based on convolutional neural networks |
CN113316080A (en) * | 2021-04-19 | 2021-08-27 | 北京工业大学 | Indoor positioning method based on Wi-Fi and image fusion fingerprint |
CN113453148A (en) * | 2021-06-25 | 2021-09-28 | 南通大学 | Indoor position fingerprint positioning method combining deep learning and weighted K-neighbor algorithm |
CN114449452A (en) * | 2021-12-29 | 2022-05-06 | 华南师范大学 | Indoor positioning algorithm for heterogeneous equipment |
Family Cites Families (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US11983641B2 (en) * | 2017-06-05 | 2024-05-14 | Umajin Inc. | Asset tracking system and methods |
US11751057B2 (en) * | 2019-09-19 | 2023-09-05 | Colorado State University Research Foundation | Security-enhanced deep learning fingerprint-based indoor localization |
KR102204783B1 (en) * | 2020-07-09 | 2021-01-18 | 전남대학교산학협력단 | Deep learning-based beamforming communication system and method |
-
2022
- 2022-06-24 CN CN202210729213.2A patent/CN115175306B/en active Active
Patent Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108282743A (en) * | 2018-03-05 | 2018-07-13 | 桂林理工大学 | Indoor orientation method, apparatus and system |
CN108924737A (en) * | 2018-06-20 | 2018-11-30 | 北京三快在线科技有限公司 | Localization method, device, equipment and computer readable storage medium |
US10620290B1 (en) * | 2019-02-19 | 2020-04-14 | Ideal Industries Lighting Llc | Indoor positioning system |
US10716089B1 (en) * | 2019-06-03 | 2020-07-14 | Mapsted Corp. | Deployment of trained neural network based RSS fingerprint dataset |
WO2021103027A1 (en) * | 2019-11-30 | 2021-06-03 | Beijing Didi Infinity Technology And Development Co., Ltd. | Base station positioning based on convolutional neural networks |
CN111669820A (en) * | 2020-05-27 | 2020-09-15 | 南京邮电大学 | Density peak value abnormity detection method and intelligent passive indoor positioning method |
CN112153736A (en) * | 2020-09-14 | 2020-12-29 | 南京邮电大学 | Personnel action identification and position estimation method based on channel state information |
CN112329685A (en) * | 2020-11-16 | 2021-02-05 | 常州大学 | Method for detecting crowd abnormal behaviors through fusion type convolutional neural network |
CN112712557A (en) * | 2020-12-17 | 2021-04-27 | 上海交通大学 | Super-resolution CIR indoor fingerprint positioning method based on convolutional neural network |
CN113316080A (en) * | 2021-04-19 | 2021-08-27 | 北京工业大学 | Indoor positioning method based on Wi-Fi and image fusion fingerprint |
CN113453148A (en) * | 2021-06-25 | 2021-09-28 | 南通大学 | Indoor position fingerprint positioning method combining deep learning and weighted K-neighbor algorithm |
CN114449452A (en) * | 2021-12-29 | 2022-05-06 | 华南师范大学 | Indoor positioning algorithm for heterogeneous equipment |
Non-Patent Citations (5)
Title |
---|
Shixun Wu ; Wenxian Huang ; Min Li ; Kai Xu."A Novel RSSI Fingerprint Positioning Method Based on Virtual AP and Convolutional Neural Network".《IEEE Sensors Journal》.2022,全文. * |
Yu Lin ; Yao Tong ; Qinkun Zhong ; Ruipeng Gao ; Buyi Yin ; Lei Liu ; Li Ma ; ."DCCP: Deep Convolutional Neural Networks for Cellular Network Positioning".《2021 IEEE Global Communications Conference (GLOBECOM)》.2022,全文. * |
刘帅 ; 王旭东 ; 吴楠."一种基于卷积神经网络的CSI指纹室内定位方法".《工程科学学报》.2021,全文. * |
张亮 ; 郭彤旭 ; 王佐强 ; 郭庆楠 ; 雷洛."一种单目视觉的水下目标定位与跟踪技术".《大众标准化》.2021,全文. * |
曹建荣 ; 张旭 ; 武欣莹 ; 吕俊杰 ; 杨红娟."结合CNN和WiFi指纹库的室内定位算法".《计算机系统应用》.2020,全文. * |
Also Published As
Publication number | Publication date |
---|---|
CN115175306A (en) | 2022-10-11 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112533149B (en) | Moving target positioning algorithm based on UWB mobile node | |
CN106793082B (en) | Mobile equipment positioning method in WLAN/Bluetooth heterogeneous network environment | |
CN109640269B (en) | Fingerprint positioning method based on CSI and time domain fusion algorithm | |
CN108174343B (en) | Wireless positioning method and system for power indoor communication operation and maintenance scene | |
CN113610243A (en) | Atmospheric pollutant tracing method based on coupled machine learning and correlation analysis | |
CN112881979B (en) | Initial state self-adaptive fusion positioning method based on EKF filtering | |
CN103916954A (en) | Probability locating method and locating device based on WLAN | |
CN110850363B (en) | Method for carrying out dynamic filtering optimization based on real-time positioning track data | |
CN114449452A (en) | Indoor positioning algorithm for heterogeneous equipment | |
CN105866732B (en) | The mixing indoor orientation method that a kind of improvement MK models and WKNN algorithms are combined | |
CN115908051A (en) | Method for determining energy storage capacity of power system | |
CN115100376A (en) | Electromagnetic spectrum map implementation method based on improved inverse distance interpolation method | |
CN115175306B (en) | Indoor positioning method of electric power Internet of things based on convolutional neural network | |
CN107656905B (en) | Air quality data real-time calibration method using error transfer | |
CN114236480A (en) | Airborne platform sensor system error registration algorithm | |
Wang et al. | Indoor position algorithm based on the fusion of wifi and image | |
CN108519579B (en) | WiFi fingerprint positioning method for analyzing optimal AP based on interval overlapping degree | |
CN113271542B (en) | Indoor mobile terminal positioning method based on Bluetooth and visible light | |
CN112040408B (en) | Multi-target accurate intelligent positioning and tracking method suitable for supervision places | |
Diao et al. | Enhancing trilateration localization by adaptive selecting distances | |
CN114286282A (en) | Fingerprint positioning method based on WiFi RSS data of mobile phone and dimension reduction algorithm | |
CN114818827A (en) | Non-invasive load decomposition method based on seq2point network | |
CN105682136A (en) | Indoor and outdoor separation method based on outdoor test characteristics | |
CN114449438A (en) | Indoor positioning method based on iBeacon fingerprint database | |
Yao et al. | Research on fingerprint location technology for indoor scene of power communication operation and maintenance |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |