CN118102444A - Self-adaptive self-updating indoor positioning method - Google Patents
Self-adaptive self-updating indoor positioning method Download PDFInfo
- Publication number
- CN118102444A CN118102444A CN202410487775.XA CN202410487775A CN118102444A CN 118102444 A CN118102444 A CN 118102444A CN 202410487775 A CN202410487775 A CN 202410487775A CN 118102444 A CN118102444 A CN 118102444A
- Authority
- CN
- China
- Prior art keywords
- indoor positioning
- indoor
- positioning
- target
- anchor
- 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
- 238000000034 method Methods 0.000 title claims abstract description 34
- 230000003044 adaptive effect Effects 0.000 claims description 13
- 239000011159 matrix material Substances 0.000 claims description 12
- 238000013528 artificial neural network Methods 0.000 claims description 7
- 238000007781 pre-processing Methods 0.000 claims description 7
- 238000004364 calculation method Methods 0.000 claims description 2
- 238000005457 optimization Methods 0.000 abstract description 2
- 238000005516 engineering process Methods 0.000 description 6
- 238000010586 diagram Methods 0.000 description 5
- 238000011161 development Methods 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000013507 mapping Methods 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 238000012549 training Methods 0.000 description 3
- 238000013527 convolutional neural network Methods 0.000 description 2
- 125000004122 cyclic group Chemical group 0.000 description 2
- 238000010295 mobile communication Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000000149 penetrating effect Effects 0.000 description 1
Landscapes
- Position Fixing By Use Of Radio Waves (AREA)
Abstract
The invention discloses a self-adaptive self-updating indoor positioning method, which comprises an indoor positioning system, wherein the indoor positioning system comprises a plurality of base stations deployed in an indoor environment, a plurality of anchor points and a server with a positioning network model, and comprises a self-adaptive self-updating process and an indoor positioning process; in the use process of the indoor positioning system, a base station acquires target signals and anchor point signals in real time to acquire and store IQ or covariance matrixes of the signals, the target signals and the anchor point signals are packaged and sent to a server after the system is used for a certain period of time, a real-time data set is manufactured by the server and provided for a positioning network model to perform incremental semi-supervised learning, parameter optimization of the positioning network model is performed, the optimized positioning network model is utilized to perform target positioning, positioning accuracy is high, and the indoor positioning system is suitable for indoor positioning scenes with frequent changes of indoor environments.
Description
Technical Field
The invention relates to the technical field of indoor positioning, in particular to a self-adaptive self-updating indoor positioning method.
Background
With the rapid development of information technology and mobile communication, there is a great increase in demand for location-based services. These location services have been widely used in a variety of fields including navigation, social media, and location identification in emergency situations. Currently, GNSS has made it possible to achieve sub-meter level positional accuracy outdoors. However, it is difficult to achieve indoor positioning using GNSS because satellite signals have difficulty penetrating a building. Because the scenes of people's work and life are mainly concentrated indoors in modern society, the demands for indoor positioning are very great, and how to improve the indoor positioning accuracy is very important.
At present, common indoor positioning technologies include WiFi, bluetooth, UWB and the like, most of the indoor positioning technologies adopt TDOA based on time difference or AOA based on angle, the positioning accuracy is easily affected by factors such as existence of direct paths, multipath phenomenon and the like, the influence of the distance between the base station and the area positioning accuracy is caused, and the area positioning accuracy is lower the farther the distance from the base station is; meanwhile, the indoor environment is generally complicated, and when the indoor environment changes such as object movement, personnel entering and exiting, and the like, the signal propagation path can be changed, so that the indoor positioning accuracy is affected. Accordingly, the present invention is directed to an adaptive and self-updating high-precision indoor positioning method to solve the above-mentioned problems.
Disclosure of Invention
The invention aims to provide a self-adaptive self-updating indoor positioning method, which solves the problems existing in the prior art, can adapt to the change of the environment, has the advantages of high positioning precision, strong instantaneity, wide generalization, good robustness and the like, and meets the indoor positioning requirements of various application scenes.
In order to achieve the above object, the solution of the present invention is:
An adaptive self-updating indoor positioning method comprises an indoor positioning system, wherein the indoor positioning system comprises a plurality of base stations deployed in an indoor environment, a plurality of anchor points and a server with a positioning network model, and comprises an adaptive self-updating process and an indoor positioning process;
(1) Adaptive update procedure
In the using process of the indoor positioning system, a pitch angle threshold value is set by taking the right lower side of a base station as a reference, a region with pitch angle lower than the threshold value represents a confidence coefficient high region, a region with pitch angle higher than the threshold value represents a confidence coefficient low region, a beacon positioned in the confidence coefficient high region and the confidence coefficient low region sends a target signal to the base station, the base station simultaneously and in real time collects the target signal in the confidence coefficient high region and the confidence coefficient low region of the indoor environment, and the base station preprocesses the target signal to obtain an IQ or covariance matrix of the target signal, so that region related data are formed and stored; transmitting an anchor signal to a base station by an anchor, preprocessing the anchor signal by the base station to obtain an IQ or covariance matrix of the anchor signal, forming anchor related data and storing the anchor related data; after the indoor positioning system reaches a preset value t in use, the base station packages and transmits the stored area related data and anchor point related data to a server, and the server makes a labeled real-time data set of an area with high confidence, a non-labeled real-time data set of an area with low confidence and a labeled anchor point data set of an anchor point and provides the data sets for a positioning network model to perform incremental semi-supervised learning so as to optimize parameters of the positioning network model;
(2) Indoor positioning process
Target signals at different positions in an indoor environment are collected by a plurality of base stations, target signals are preprocessed to obtain IQ or covariance matrixes of the target signals, the IQ or covariance matrixes of the target signals are input into an optimized positioning network model, AOA or TDOA of a target to be positioned is obtained, and position coordinates of the target are obtained through a position estimation algorithm.
In the indoor positioning process, the AOA of the target to be positioned is obtained through calculation of an angle estimation algorithm, and the TDOA of the target to be positioned is obtained through a channel parameter estimation algorithm.
Preferably, the angle estimation algorithm is MUSIC, CBF algorithm, or the like.
The positioning network model is a neural network such as a cyclic neural network, a convolutional neural network or a long-time and short-time memory network.
The position estimation algorithm is a triangular positioning method and the like.
After the technical scheme is adopted, the invention has the following technical effects:
According to the invention, through continuous iterative training of the positioning network model, parameters of the positioning network model are optimized in real time, so that the mapping relation between relevant characteristics of signals and relevant information such as AOA, TDOA and the like is accurately learned, the indoor positioning system can adapt to the change of the environment, is not limited by the fixed environment, has strong generalization and robustness, has high positioning precision, can effectively solve the problems of variability, multipath effect and the like of the indoor environment, is suitable for various indoor positioning systems, and provides a new solution for the development of the indoor positioning technology.
Drawings
FIG. 1 is a scene diagram of an adaptive self-updating indoor positioning system according to the present invention;
FIG. 2 is a deployment diagram of an indoor positioning system for multi-anchor monitoring according to the present invention;
FIG. 3 is a flow chart of the adaptive self-updating indoor positioning system of the present invention.
Detailed Description
In order to further explain the technical scheme of the invention, the invention is explained in detail by specific examples.
Referring to fig. 1, an adaptive self-updating indoor positioning system scene diagram is shown:
A plurality of base stations are deployed indoors, target signals emitted from targets are received in real time, the target signals are preprocessed to obtain IQ or covariance matrixes of the target signals, and the IQ or covariance matrixes are transmitted to a server in real time through a network cable; and a positioning network model is arranged in the server, and an IQ or covariance matrix of the target signal is input into the positioning network model, so that the position coordinates of the target can be obtained.
Referring to fig. 2, a deployment diagram of an indoor positioning system for multi-anchor monitoring is shown:
① As an anchor point, continuously transmitting an anchor point signal, and indirectly sensing the change of the environment;
② Receiving and preprocessing an anchor signal transmitted from an anchor for a base station;
③ As an indoor object, influencing the propagation path of the anchor point signal;
④ Is an object to be positioned and is in a room;
⑤ For a position resolving server (hereinafter referred to as "server"), a data set is received from a base station and optimization of a positioning network model and position resolving of a target are completed.
In this embodiment, an indoor positioning technology based on AOA is taken as an example, where a plurality of anchor points transmit anchor point signals to a base station at specific indoor positions, the base station acquires IQ or covariance matrixes of the anchor point signals through preprocessing, and forms a real-time data set with AOA information of the anchor points to be transmitted to a server, and a positioning network model is used on the server to perform incremental semi-supervised learning on the real-time data set, optimize parameters of the positioning network model, obtain a real-time positioning network model adapted to the current environment, and realize real-time high-precision positioning.
Referring to fig. 3, an adaptive self-updating indoor positioning system workflow diagram is shown:
(1) Indoor positioning system
The method comprises the steps that a plurality of base stations collect indoor data at different positions by using signal acquisition equipment, relevant characteristics such as IQ or covariance matrix of a target signal are obtained through preprocessing, then AOA is obtained by using a classical AOA estimation algorithm and a neural network method respectively, more accurate AOA is selected to serve as an estimation result of a final system through comparison, and finally position coordinates of the target are estimated through a triangular positioning position estimation method. The classical AOA estimation algorithm may be MUSIC, CBF, etc. and the neural network may be a cyclic neural network, a convolutional neural network, a long-short term memory network, etc. When the accuracy of the AOA obtained by the two methods is compared, whether a classical estimation algorithm is selected or not can be determined by using whether the ratio of the sum of the maximum eigenvalue of the covariance matrix of the signal and other eigenvalues exceeds a set threshold, if the ratio exceeds the set threshold, an AOA result obtained by the classical AOA estimation algorithm is selected, otherwise, an AOA result obtained by a neural network method is selected.
(2) Online updates
On the one hand, in order to improve the positioning accuracy, in the use process of the indoor positioning system, a pitch angle threshold value is set by taking the right lower side of a base station as a reference, a region with a pitch angle lower than the pitch angle threshold value represents a confidence level high region, a region with a pitch angle higher than the pitch angle threshold value represents a confidence level low region, beacons positioned in the confidence level high region and the confidence level low region send target signals to the base station, the base station simultaneously and in real time acquires the target signals in the confidence level high region and the confidence level low region, and the relevant data of the regions of an IQ or covariance matrix of the target signals are acquired through preprocessing and stored; after the indoor positioning system reaches a preset value t in use, the base station packages and transmits the stored area related data to the server, and the area with high confidence in the area related data is considered to have real position coordinates because the area with high confidence in the area related data is considered to have higher positioning accuracy, so that a real-time data set with labels of the area with high confidence and a real-time data set without labels of the area with low confidence can be manufactured by the server, and the real-time data set with labels and the real-time data set without labels of the area with low confidence are provided for a positioning network model to perform incremental semi-supervised learning, so that the relation between IQ or covariance matrix of signals and target AOA can be learned more accurately, and the positioning accuracy of the area with low confidence is improved.
On the other hand, in order to enable the indoor positioning system to sense the change of the environment, a method of anchor point monitoring can be used, a plurality of anchor points are deployed at specific indoor positions, anchor point signals are continuously transmitted to a base station, the anchor point signals are received by the base station after being transmitted through different paths in the indoor environment, when the indoor environment changes due to object movement, personnel access and the like, the paths of the anchor point signals reaching the base station change, the base station acquires the anchor point signals in real time, and the anchor point related data of an IQ or covariance matrix of the anchor point signals are acquired through preprocessing and stored; after the indoor positioning system reaches a preset value t in use, the base station packages and transmits the stored anchor related data to the server, and as the position of the anchor is known, a relatively real-time anchor data set with a label of the anchor can be manufactured by the server and provided for the positioning network model to perform incremental semi-supervised learning, so that the positioning network model can indirectly learn the change of the environment.
The number, the positions and the signal sending frequency and the signal sending intensity of the anchor points can be set according to actual conditions, so that the signals emitted by the anchor points can cover the whole indoor environment as much as possible in principle, and the change of the indoor environment can be perceived more accurately.
The labeled real-time data set in the high-confidence region, the unlabeled real-time data set in the low-confidence region and the labeled anchor point data set are synthesized, incremental semi-supervised learning is carried out on the data set synthesized by the positioning network model, and parameters of the positioning network model are optimized through continuous iterative training, so that the positioning network model can learn the mapping relation between the IQ or covariance matrix of the signal and the AOA more accurately, and further the optimized real-time positioning network model which is adaptive to the current environment can be obtained, and real-time high-precision positioning is realized.
(3) Indoor positioning
And (3) inputting an IQ or covariance matrix of the target signal, acquiring an AOA of the target by using the optimized real-time positioning network model, and then acquiring the position coordinates of the target by using a triangular positioning position estimation method to realize real-time high-precision positioning in a dynamic environment.
According to the invention, through continuous iterative training of the positioning network model, parameters of the positioning network model are optimized in real time, so that the mapping relation between relevant characteristics of signals and relevant information such as AOA, TDOA and the like is accurately learned, the indoor positioning system can adapt to the change of the environment, is not limited by the fixed environment, has strong generalization and robustness, has high positioning precision, can effectively solve the problems of variability, multipath effect and the like of the indoor environment, is suitable for various indoor positioning systems, and provides a new solution for the development of the indoor positioning technology.
The above examples and drawings are not intended to limit the form or form of the present invention, and any suitable variations or modifications thereof by those skilled in the art should be construed as not departing from the scope of the present invention.
Claims (3)
1. An adaptive self-updating indoor positioning method is characterized in that:
The system comprises an indoor positioning system, a control system and a control system, wherein the indoor positioning system comprises a plurality of base stations deployed in an indoor environment, a plurality of anchor points and a server with a positioning network model, and comprises a self-adaptive self-updating process and an indoor positioning process;
(1) Adaptive update procedure
In the using process of the indoor positioning system, a pitch angle threshold value is set by taking the right lower side of a base station as a reference, a region with pitch angle lower than the threshold value represents a confidence coefficient high region, a region with pitch angle higher than the threshold value represents a confidence coefficient low region, a beacon positioned in the confidence coefficient high region and the confidence coefficient low region sends a target signal to the base station, the base station simultaneously and in real time collects the target signal in the confidence coefficient high region and the confidence coefficient low region of the indoor environment, and the base station preprocesses the target signal to obtain an IQ or covariance matrix of the target signal, so that region related data are formed and stored; transmitting an anchor signal to a base station by an anchor, preprocessing the anchor signal by the base station to obtain an IQ or covariance matrix of the anchor signal, forming anchor related data and storing the anchor related data; after the indoor positioning system reaches a preset value t in use, the base station packages and transmits the stored area related data and anchor point related data to a server, and the server makes a labeled real-time data set of an area with high confidence, a non-labeled real-time data set of an area with low confidence and a labeled anchor point data set of an anchor point and provides the data sets for a positioning network model to perform incremental semi-supervised learning so as to optimize parameters of the positioning network model;
(2) Indoor positioning process
Target signals at different positions in an indoor environment are collected by a plurality of base stations, target signals are preprocessed to obtain IQ or covariance matrixes of the target signals, the IQ or covariance matrixes of the target signals are input into an optimized positioning network model, AOA or TDOA of a target to be positioned is obtained, and position coordinates of the target are obtained through a position estimation algorithm.
2. The adaptive self-updating indoor positioning method according to claim 1, wherein:
In the indoor positioning process, the AOA of the target to be positioned is obtained through calculation of an angle estimation algorithm, and the TDOA of the target to be positioned is obtained through a channel parameter estimation algorithm.
3. The adaptive self-updating indoor positioning method according to claim 1, wherein:
the positioning network model is a neural network.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410487775.XA CN118102444A (en) | 2024-04-23 | 2024-04-23 | Self-adaptive self-updating indoor positioning method |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202410487775.XA CN118102444A (en) | 2024-04-23 | 2024-04-23 | Self-adaptive self-updating indoor positioning method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN118102444A true CN118102444A (en) | 2024-05-28 |
Family
ID=91152028
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202410487775.XA Pending CN118102444A (en) | 2024-04-23 | 2024-04-23 | Self-adaptive self-updating indoor positioning method |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN118102444A (en) |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110933626A (en) * | 2019-11-05 | 2020-03-27 | 东南大学 | High-precision self-organizing network type indoor positioning method |
CN113993074A (en) * | 2021-11-19 | 2022-01-28 | 深圳市佳贤通信设备有限公司 | 5G base station signal transceiving device and target positioning method |
CN114924225A (en) * | 2022-04-28 | 2022-08-19 | 合肥工业大学 | High-precision indoor positioning method, device, equipment and medium |
CN116918403A (en) * | 2023-04-26 | 2023-10-20 | 上海移远通信技术股份有限公司 | Method and device for training positioning model, and method and device for positioning |
-
2024
- 2024-04-23 CN CN202410487775.XA patent/CN118102444A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110933626A (en) * | 2019-11-05 | 2020-03-27 | 东南大学 | High-precision self-organizing network type indoor positioning method |
CN113993074A (en) * | 2021-11-19 | 2022-01-28 | 深圳市佳贤通信设备有限公司 | 5G base station signal transceiving device and target positioning method |
CN114924225A (en) * | 2022-04-28 | 2022-08-19 | 合肥工业大学 | High-precision indoor positioning method, device, equipment and medium |
CN116918403A (en) * | 2023-04-26 | 2023-10-20 | 上海移远通信技术股份有限公司 | Method and device for training positioning model, and method and device for positioning |
Non-Patent Citations (1)
Title |
---|
刘公绪: "室内人员自主定位定向方法研究", 《中国博士学位论文全文数据库》, 15 May 2021 (2021-05-15), pages 1 - 194 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
KR102116824B1 (en) | Positioning system based on deep learnin and construction method thereof | |
CN103402258B (en) | Wi-Fi (Wireless Fidelity)-based indoor positioning system and method | |
CN106793087B (en) | Array antenna indoor positioning method based on AOA and PDOA | |
CN112533163B (en) | Indoor positioning method based on NB-IoT (NB-IoT) improved fusion ultra-wideband and Bluetooth | |
CN106842123B (en) | A kind of indoor orientation method and indoor locating system | |
Zhang et al. | A scalable indoor localization algorithm based on distance fitting and fingerprint mapping in Wi-Fi environments | |
CN109936837B (en) | Indoor positioning method and system based on Bluetooth | |
CN104181500A (en) | Real-time locating method based on inertia information and chance wireless signal characteristics | |
CN112533136B (en) | WLAN fingerprint positioning method based on deep learning | |
CN110187333B (en) | RFID label positioning method based on synthetic aperture radar technology | |
CN112925001A (en) | UWB positioning and satellite positioning dual-fusion greenhouse internal and external positioning system and method | |
CN111698774A (en) | Indoor positioning method and device based on multi-source information fusion | |
CN113534047B (en) | Equipment positioning method and device | |
CN109239661A (en) | A kind of RFID indoor locating system and algorithm based on depth Q network | |
Zhou-guo et al. | An improved indoor UHF RFID localization method based on deviation correction | |
CN109188360A (en) | A kind of indoor visible light 3-D positioning method based on bat algorithm | |
CN109839613A (en) | A kind of radio frequency positioning method and device calibrated using routing information | |
De Oliveira et al. | Low-cost indoor localization system combining multilateration and kalman filter | |
CN114758364A (en) | Industrial Internet of things scene fusion positioning method and system based on deep learning | |
CN118102444A (en) | Self-adaptive self-updating indoor positioning method | |
CN116772860A (en) | Novel indoor positioning system based on integration of wireless positioning technology and visual artificial intelligence | |
KR20180110392A (en) | Hybrid location positioning method for indoor location measurement based on smart device | |
Guo et al. | Indoor positioning optimization based on genetic algorithm and RBF neural network | |
CN114205741B (en) | TOA and position fingerprint combined indoor positioning method based on UWB | |
CN115079089A (en) | Differential correction-based non-cooperative unmanned aerial vehicle accurate positioning method and device |
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 |