CN105722028A - Indoor pedestrian positioning system and indoor pedestrian positioning method based on WIFI and magnetic field two-level search - Google Patents

Abstract

The invention discloses an indoor pedestrian positioning system and an indoor pedestrian positioning method based on WIFI and magnetic field two-level search, wherein the system is mainly composed of a mobile terminal and a positioning server. When a pedestrian walks indoors with a mobile terminal in the hand, firstly, primary positioning is carried out through the WIFI, and after a primary coordinate is determined through WIFI positioning, magnetic field information within an area with a radius R is searched through the coordinate to construct a temporary fingerprint database. Matching calculation is carried out on the magnetic field information measured online and the temporary fingerprint database, and finally, the coordinate of the to-be-positioned position is obtained. The positioning method makes full use of validity of coarse positioning of the WIFI positioning and accuracy of small-range positioning of the magnetic field, simultaneously reduces dependence on WIFI signals, and improves accuracy and reliability of positioning.

Description

Based on WIFI, Two step Search indoor, magnetic field pedestrian's alignment system and method

Technical field

The present invention relates to the indoor orientation method of area of pattern recognition, particularly relate to based on WIFI, Two step Search indoor, magnetic field pedestrian's alignment system and method.

Background technology

Along with being continuously increased of position-based service (LBS) demand, indoor positioning received more and more attention in recent years.It was widely available research in recent years owing to extra hardware need not be increased, there is the advantage such as low cost, widespread deployment based on WIFI framework indoor locating system.But, owing to radio signal is highly susceptible to environmental disturbances, bringing great challenge to WIFI positioning framework location, its positioning precision obtains very big restriction, and WIFI positioning precision can only meet in the scope of 3～5 meters under normal circumstances.Along with popularizing of smart mobile phone built-in sensors, magnetic field sensor is embedded in smart mobile phone to detect magnetic field of the earth information；In indoor environment, magnetic field of the earth information is easily subject to the distortion of Indoor environment environment and produces exception, for instance reinforced concrete, pillar, common electric etc., these devices or equipment meeting disturbing magnetic field, make indoor earth's magnetic field have different distributions in zones of different, and Magnetic Field is highly stable.Then experiments show that, have identical Magnetic Field in zones of different, this brings certain problem to single utilization Magnetic oriented.Single utilize WIFI or Magnetic oriented generally can not get satisfied result.

Summary of the invention

In order to solve above-mentioned technical problem, the present invention provides based on WIFI, Two step Search indoor, magnetic field pedestrian's alignment system and method, make full use of the respective feature of WIFI and magnetic field, play the advantage in its respective position fixing process, adopt Two step Search strategy, to obtain positioning result more accurately, concrete technical scheme is as follows.

Based on WIFI, Two step Search indoor, magnetic field pedestrian's alignment system, including one by mobile terminal, location-server, database server and the web server for safeguarding.Mobile terminal is mainly carried by location user, and the built-in WIFI of mobile terminal gathers and magnetic field sensor module.Can obtaining the WIFI information that can receive of current location and the Magnetic Field of current location in real time, mobile terminal and location-server communication are taked data packing and are sent in the location-server of fixing IP with TCP or udp protocol；Database server is mainly used in save location finger print data and region, location cartographic information；Web server is mainly used in alignment system maintenance, monitoring, location fingerprint updates and map is uploaded.

Further, customer location and user trajectory can be shown by mobile terminal in real time, can show location title when clicking customer location.

Further, location-server and web server are deployed in high in the clouds, and location-server performs location algorithm, and web server also has data management, user terminal management function and user feedback management.

Based on WIFI, Two step Search indoor, magnetic field pedestrian's localization method, comprise the following steps:

Step one: when carrying out floor data collection, arranges WIFI sampled point in floor area to be positioned according to setting interval, gathers L data at each WIFI sampled point, build WIFI location fingerprint data storehouse；Magnetic field sampled point is set according to the interval of 1 meter simultaneously, in each magnetic field, sampled point also gathers L data and obtains meansigma methods, the geographical position coordinates of each sampled point is recorded when gathering WIFI and magnetic field data, build WIFI, Magnetic oriented fingerprint database, record the MAC Address of floor porch WIFI simultaneously, entrance WIFI access point MAC Address, WIFI, magnetic field fingerprint database and floor maps are bound packing together and uploads to location-server；Wherein, L takes 300～500 according to WIFI signal stable case；

Step 2: during inlet chamber environment, the MAC Address of the WIFI data that first can receive according to porch in porch obtains location packet to web server, the map datum bag inquired downloads to mobile terminal this locality and carries out resolving adaptation mobile terminal screen；

Step 3: during real-time positioning, gathers WIFI data and the magnetic field data of current location, first the WIFI data separate KNN algorithm of Real-time Collection carries out coarse positioning calculating and obtains coarse positioning coordinate (x_w, y_w)；

Step 4: getting WIFI coarse positioning coordinate (x_w, y_w) after, with (x_w, y_w) for the center of circle, add set S to the magnetic field finger print information that radius is R₁In；With (x_k-1, y_k-1) for the center of circle, add set S to the magnetic field finger print information that radius is R₂In；Take S₁And S₂Union composition finger print information enough build interim fingerprint database S_M, by the magnetic field data of Real-time Collection and set S_MIn Magnetic Field utilize KNN algorithm to position calculating, export the final elements of a fix；Wherein (x_k-1, y_k-1) it is the last elements of a fix, R value is determined according to practical situation and positioning precision, and wherein k represents kth time location, S when wherein k is equal to 1₁Can be taken as empty set.3, as above-mentioned based on WIFI, Two step Search indoor, magnetic field pedestrian's localization method: in step one, the resolution that the sampled point employing linear interpolation method at 1 meter of interval makes when building magnetic field fingerprint database magnetic field reference point is 0.1 meter；When building Magnetic Field Standard storehouse, the magnetic field data component B that will gather_x, B_y, B_zConverting accordingly, build BHV fingerprint vector, wherein transformation for mula is as follows:

$BHV=(\sqrt{{B_x}^2+{B_y}^2+{B_z}^2},\sqrt{{B_x}^2+{B_y}^2},B_z)---\left(1\right)$

Wherein B_x, B_y, B_zIt it is the mobile terminal magnetic-field component along tri-directions of terminal x, y, z.

The present invention, by adopting technique scheme, compared with prior art, the invention have the advantages that and technique effect:

1) decreasing fingerprint collecting workload, WIFI fingerprint it is frequently necessary to safeguard owing to being easily subject to environmental disturbances, and in the present invention, the WIFI fingerprint spacing gathered is very big, the fingerprint of collection is very little, and during later maintenance, workload is fewer；The earth magnetism fingerprint simultaneously gathered in the present invention need not be safeguarded substantially due to its stability very good later stage, and the present invention adopts linear interpolation method that magnetic field fingerprint is interpolated, and decreases workload during collection.

2) make full use of WIFI and Magnetic Field, improve positioning precision.The present invention make full use of WIFI in the advantage of coarse positioning and magnetic field the advantage at local area localization, a little formation one two-stage in conjunction with two kinds of localization methods searches element localization method, improves positioning precision.

Accompanying drawing explanation

Fig. 1 is based on WIFI, Two step Search indoor, magnetic field pedestrian positioning system structure figure in example.

Fig. 2 is acquisition for mobile terminal ground map flow chart.

Fig. 3 implements overall flow figure based on WIFI, Two step Search indoor, magnetic field pedestrian's localization method in example.

Fig. 4 builds interim fingerprint database schematic diagram in example.

Detailed description of the invention

Below in conjunction with accompanying drawing, specific embodiment of the invention is described further, but the enforcement of the present invention and protection are not limited to this.

Fig. 1 is the system architecture diagram of the present invention, and the present invention is mainly made up of mobile terminal, location-server, database server and the web server for safeguarding.

Mobile terminal communicates with server through the Internet mainly through WIFI or GPRS, and mobile terminal possesses WIFI and gathers and magnetic field collection sensor；Mobile terminal is based primarily upon Android and ios platform development and location APP application, and APP application can cross display customer location and user's run trace in real time, and mobile terminal is communicated by TCP/UDP agreement and location-server；Location-server is mainly used in receiving mobile terminal request and positions computing restoring to normal position result simultaneously to mobile terminal, location-server is taken into account server after floor identification number is sent to location-server by map download function and mobile terminal and is inquired about map datum according to identification number simultaneously, if map datum being sent to mobile terminal after inquiring map datum, the service that otherwise returns can not obtain and please build map datum, and flow chart is as shown in Figure 2；Web server is mainly used in carrying out back-stage management platform, and including data management, for terminal management, map structuring and user feedback management etc., alignment system manager can carry out background maintenance by browser.

Fig. 3 is that localization method of the present invention implements overall flow figure, and localization method of the present invention is mainly made up of two parts, location database establishment stage: this stage mainly sets up the location fingerprint database data for positioning；Online real-time positioning process: this process mainly obtains the location information of Real-time Collection and goes to make comparisons with data in data base to obtain positioning result.

Referring to Fig. 3, localization method detailed description of the invention of the present invention is as follows:

Step one: when carrying out floor data collection, in floor area to be positioned, certain WIFI sampled point is set according to 6m interval, gathers 500 data at each WIFI sampled point, build WIFI location fingerprint data storehouse；A certain amount of magnetic field sampled point is set according to the interval of 1 meter simultaneously, in each magnetic field, sampled point also gathers 500 secondary data and obtains meansigma methods, the geographical position coordinates of each sampled point is recorded when gathering WIFI and magnetic field data, build WIFI, Magnetic oriented fingerprint database, record the MAC Address of the WIFI access point of floor porch simultaneously, the MAC Address of entrance WIFI, WIFI, magnetic field fingerprint database and floor maps one piece are bound packing and uploads to location-server.

Step 2: during inlet chamber environment, first in porch according to porch can receive to the MAC Address of WIFI data carry out floor inquiry to server, the map inquired is downloaded to mobile terminal local and carry out parsing and be fitted to mobile terminal.

Step 3: during real-time positioning, gathers WIFI data and the magnetic field data of current location, first the WIFI data separate KNN algorithm of Real-time Collection carries out coarse positioning calculating and obtains coarse positioning coordinate (x_w, y_w), relative to plane coordinate system.

Step 3: getting WIFI coarse positioning coordinate (x_w, y_w) after, with (x_w, y_w) for the center of circle, add set S to the magnetic field finger print information that radius is R₁In；With (x_k-1, y_k-1) for the center of circle, add set S to the magnetic field finger print information that radius is R₂In；Take S₁And S₂Union composition finger print information enough build interim fingerprint database S_M, by the Magnetic Field of Real-time Collection and set S_MIn Magnetic Field utilize KNN algorithm to position calculating, export the final elements of a fix.(wherein (x_k-1, y_k-1) be the last position coordinates positioned)

1) as above-mentioned based in WIFI, Two step Search indoor, magnetic field pedestrian's localization method step one, WIFI and magnetic field being sampled at differing intervals, the resolution that the sampled point employing linear interpolation method at 1 meter of interval makes when building Magnetic Field Standard storehouse magnetic field reference point simultaneously is 0.1 meter.

2) as above-mentioned based in WIFI, Two step Search indoor, magnetic field pedestrian's localization method step one, when building Magnetic Field Standard storehouse, the magnetic field data component B that will gather_p=(B_x, B_y, B_z) convert accordingly, build BHV fingerprint vector, wherein transformation for mula is as follows:

$BHV=(\sqrt{{B_x}^2+{B_y}^2+{B_z}^2},\sqrt{{B_x}^2+{B_y}^2},B_z)---\left(2\right)$

Wherein B_x、B_y、B_zThe respectively magnetic field x, y, z component of intelligent terminal, these three component can be directly obtained from the Magnetic Sensor of mobile terminal.WhereinIt is a rotational invariants,It is one and horizontally rotates invariant, B_zWhen highly not changing, stability is fine.

Claims (5)

1. based on WIFI, Two step Search indoor, magnetic field pedestrian's alignment system, it is characterised in that include a mobile terminal, location-server, database server and the web server for safeguarding；Database server is mainly used in save location finger print data and region, location cartographic information；Web server is mainly used in alignment system maintenance, monitoring, location fingerprint updates and map is uploaded；Location-server is mainly used in running location algorithm program；Mobile terminal is carried by positioning user, the built-in WIFI of mobile terminal gathers and magnetic field sensor module, obtaining the WIFI information that can receive of current location and the Magnetic Field of current location in real time, mobile terminal is by the location-server communication of TCP or udp protocol and fixing IP；Database server is mainly used in save location finger print data and region, location cartographic information；Web server is mainly used in alignment system maintenance, monitoring, location fingerprint updates and map is uploaded.

2. as claimed in claim 1 based on WIFI, Two step Search indoor, magnetic field pedestrian's alignment system, it is characterised in that: customer location and user trajectory can be shown by mobile terminal in real time, can show location title when clicking customer location.

3. as claimed in claim 1 based on WIFI, Two step Search indoor, magnetic field pedestrian's alignment system, it is characterized in that: location-server, database server and web server are deployed in high in the clouds, location-server performs location algorithm, and web server also has data management, user terminal management and user feedback management function.

4. utilize system described in any one of claim 1～3 based on WIFI, Two step Search indoor, magnetic field pedestrian's localization method, it is characterised in that comprise the following steps:

Step one: when carrying out floor data collection, arranges WIFI sampled point in floor area to be positioned according to setting interval, gathers L data at each WIFI sampled point, build WIFI location fingerprint data storehouse；Magnetic field sampled point is set according to the interval set simultaneously, in each magnetic field, sampled point also gathers L data and obtains meansigma methods, the geographical position coordinates of each sampled point is recorded when gathering WIFI and magnetic field data, build WIFI, Magnetic oriented fingerprint database, record the MAC Address of floor porch WIFI simultaneously, WIFIMAC address, porch, WIFI, magnetic field fingerprint database and floor maps are bound packing together and uploads to location-server；Wherein, L takes 300～500 according to WIFI signal stable case；

Step 2: during inlet chamber environment, the MAC Address of the WIFI signal that first can receive according to porch in porch obtains location packet to web server, the map datum bag inquired downloads to mobile terminal this locality and carries out resolving adaptation mobile terminal screen；

Step 3: during real-time positioning, gathers WIFI data and the magnetic field data of current location, first the WIFI data separate KNN algorithm of Real-time Collection carries out coarse positioning calculating and obtains coarse positioning coordinate (x_w, y_w)；

Step 4: getting WIFI coarse positioning coordinate (x_w, y_w) after, with (x_w, y_w) for the center of circle, add set S to the magnetic field finger print information that radius is R₁In；With (x_k-1, y_k-1) for the center of circle, add set S to the magnetic field finger print information that radius is R₂In；Take S₁And S₂Union composition finger print information enough build interim fingerprint database S_M, by the magnetic field data of Real-time Collection and set S_MIn magnetic field data utilize KNN algorithm to position calculating, export the final elements of a fix；Wherein (x_k-1, y_k-1) it is the last elements of a fix, R value is determined according to practical situation and positioning precision, and wherein k represents kth time location, S when wherein k is equal to 1₁Can be taken as empty set.

5. as claimed in claim 4 based on WIFI, Two step Search indoor, magnetic field pedestrian's localization method, it is characterised in that: in step one, the resolution that the sampled point employing linear interpolation method at 1 meter of interval makes when building magnetic field fingerprint database magnetic field reference point is 0.1 meter；When building Magnetic Field Standard storehouse, the magnetic field data component B that will gather_x, B_y, B_zConverting accordingly, build BHV fingerprint vector, wherein transformation for mula is as follows:

$BHV=(\sqrt{{B_x}^2+{B_y}^2+{B_z}^2},\sqrt{{B_x}^2+{B_y}^2},B_z)---\left(1\right)$

Wherein B_x, B_y, B_zIt it is the mobile terminal magnetic-field component along tri-directions of terminal x, y, z.