CN104597469A - Intelligent positioning method based on GPS/AGPS/GPSOne and target behavior characteristics - Google Patents

Abstract

The invention discloses an intelligent positioning method based on GPS/AGPS/GPSOne and target behavior characteristics and solves the problems of absence of acuity of group behavior characteristics and identification advantages of specific group positioning in the prior art. By the intelligent positioning method based on existing electronic maps, various areas are divided according to interest points, and according to the target behavior characteristics, weight is set for each stage of division; within error range of positioning, probability of the target in each interest point is calculated, the highest probability is used as a positioning result, the positioning result is low in error, and positioning accuracy is much superior to that existing positioning methods.

Description

Based on the intelligent locating method of GPS/AGPS/GPSOne and goal behavior feature

Technical field

The present invention relates to a kind of localization method, specifically, relate to a kind of intelligent locating method based on GPS/AGPS/GPSOne and goal behavior feature.

Background technology

The openness improved constantly along with society, the floating population increasingly increased, also bring unsafe hidden danger while promoting social development.The safety problem of middle and primary schools, kindergarten student is on the rise, and the safety of each age level students becomes the focal issue that country, society, school and the head of a family pay close attention to.At present to the technological means of children's protection mainly through the construction of the technical precaution facilities such as the monitoring of inside, campus, antitheft, electronic visiting with should be used for improving campus security level.But these systems only can be protected the children in campus environment, due to the uncertainty of children's scope of activities outside campus, making to carry out effective protection to after-school activities children becomes a difficult point.

Along with the development of location technology and wireless communication technology, GPS/AGPS/GPSOne location technology penetrates into every field at present, is all widely used in military affairs, business, geography, transport with in communicating etc.Difference according to actual needs, also different to the requirement of gps system.The positioning principle of GPS is relatively simple: the time that receiver is transmitted by the gps satellite accurately measuring earth overhead determines its position.The message that every satellite transmits continuously comprises the time of transmission of messages, accurate orbit parameter (ephemeris) and system state and all gps satellites track roughly (yearbook).Satellite, owing to depending on the performance of terminal unduly, scans, catches, pseudo range signals receives and the work such as positions calculations combines in terminal all over the body by traditional GPS technology, thus causes the defect of the aspects such as the low and terminal power consumption amount of location sensitivity is large.APGS/GPSOne technology then simplifies the workload of terminal, by satellite scanner uni positions calculations etc. the most heavy task hand to network from terminal side and complete, thus in sensitivity, cold start speed, terminal power consumption amount etc., have received satisfied effect.

The application of GPS location technology has reached perfervid degree a few days ago.The support of GPS technology is all be unable to do without from the electronic chart of handheld terminal to on-vehicle navigation apparatus.But as weighing an important indicator of positioning system performance, the raising of positioning precision is still current large technological difficulties.Positioning precision is by the impact of various factors, and the impact for intrinsic factor (as atmospheric environment, systematic error) normally cannot directly be eliminated; And for the impact of enchancement factor, can be controlled by the means of technology.In order to improve positioning precision, and meet the application demand of range finding, differential type GPS (DifferentialGPS, DGPS) arise at the historic moment, its method arranges GPS on the object of reference of Accurate Measurement coordinate, and be no less than the same group of satellite of four with the GPS simultaneous observation on transfer table, try to achieve the differential corrections number (differential position in this moment, pseudo range difference, differential smooth pseudo distance of phase and phase differential grade and revise number), give these correction real-time broadcastings the transfer table (user) that works by wireless software download nearby or send transfer table (user) afterwards to, by transfer table (user) differential corrections received, its GPS locator data is revised in real time, its objective is and eliminate common error item, the impact of attenuation of correlation error effectively, and then obtain more accurate positioning result, improve positioning precision.DGPS brings the raising of the order of magnitude relative to the navigation and positioning accuracy that GPS can be user, is applied in Aeronautics and Astronautics, navigation and automotive field such as the landing of aircraft precision approach, unmanned plane, ballistic trajectory measurement, vehicle positioning and navigations.DGPS can provide accurate position and velocity information when using visible carrier phase, but due to the generation of dephasing between receiver and satellite, can cause cycle landing (cycleslip), and then affect measuring accuracy.This cycle landing greatly can reduce positioning precision in a dynamic mode.Such as, when vehicle enters tunnel, the easy conductively-closed of vehicle GPS receiver signal and interrupting.In order to improve DGPS performance in a dynamic mode, the mode that researchist proposes to adopt GPS/INS (Inertial Navigation System) to combine strengthens the positioning precision of GPS, the short-term relative positional accuracy of INS just can be used to the cycle landing detecting and correct in gps carrier phase place, meanwhile, the accurate measurement of receiver can constantly for INS provides renewal.INS can also be filled up the difference of satellite tracking by masking effect.This integrated form system by than independent system more accurately, more reliable, and operationally have more retractility.But random noise is existed for INS device, denoising method conventional at present has mean filter, medium filtering, Wavelet Denoising Method etc.Mean Filtering Algorithm is simple and practical, and filtering time is short, but is only applicable to denoising that is static and low Dynamic Signal; Median filtering algorithm is applicable to the denoising of the stronger signal of dynamic, but when there is high frequency noise in the signal of slowly change, its denoising effect is not as mean filter; Wavelet Denoising Method effect is better than mean filter and medium filtering, and static state, Dynamic Signal are all applicable, but owing to carrying out multi-level signal decomposition and reconstruct, required time is longer, is difficult to the requirement of real-time meeting system.

Up to now, developing rapidly and popularizing relative to GPS technology, location technology accordingly for the behavior pattern of specific crowd relatively lags behind, far can not meet specific crowd (as children, the elderly) positioning field and automatic classification and sterically defined requirement are carried out to crowd behaviour pattern.Only scheme of guarding for children's safety adopts the technological orientations such as GPS/AGPS/GPSOne at present, although this method has the advantages such as simple, directly perceived, but this method can not obtain higher positioning accuracy, the general area of children present position can only be determined, can not children or the elderly be located accurately according to the behavioural characteristic of children or the elderly, effective location and safe preservation cannot be carried out to specific crowd.Essentially, the main cause producing these drawbacks is, at present based on the resolution characteristic of the localization method shortage crowd behaviour feature of GPS/AGPS/GPSOne, not possess the identification advantage of specific crowd location.

Summary of the invention

The object of the invention is to overcome above-mentioned defect, a kind of intelligent locating method based on GPS/AGPS/GPSOne and goal behavior feature improving the identification capability put accurate as to position for specific crowd in existing location technology is provided.

To achieve these goals, the technical solution used in the present invention is as follows:

Based on an intelligent locating method for GPS/AGPS/GPSOne and goal behavior feature, comprise the following steps:

(1) all points of interest in electronic chart are carried out the hierarchical classification of X layer, be respectively first order classification, second level classification, third level classification ... X level is classified, wherein, containing dissimilar point of interest class in every first-level class, the first order is categorized as higher level's class of second level classification, the second level is categorized as higher level's class of third level classification, the like, only include certain concrete point of interest in each point of interest class of X level classification; Specifically, all points of interest in electronic chart are carried out the hierarchical classification of x layer, be some large classes (ground floor) by Partition for Interest Points, each large class is further subdivided into some subclasses, each subclass continues to be divided into some little subclasses, successively analogizes, until X layer, an object is only included, certain the concrete point of interest namely on electronic chart in each point of interest class of X layer; The division of point of interest can depending on actual situation, as be divided into 3 grades, 4 grades, 5 grades even more, if be divided into 3 grades, then there are first order point of interest class, second level point of interest class and third level point of interest class, wherein, first order point of interest class is higher level's type of second level point of interest class, and second level point of interest class is higher level's type of third level point of interest class, and third level point of interest class is then concrete point of interest;

(2) to point of interest classes different in every one-level, arrange corresponding initial weight, and the point of interest class setting weight sum belonging to same higher level's class in same layer equals 1, the setting weight sum of all point of interest classes in first order classification equals 1;

(3) shared by point of interest concrete in the classification of X level, area size carries out fence to concrete point of interest;

(4) according to from the S set of all concrete point of interest in the positioning result compute location center that GPS/AGPS/GPSONE obtains and error radius R, and the distance Dis of each concrete point of interest and the centre of location in set of computations S;

(5) for concrete point of interest each in S set, the type weight ω (i) of this point of interest is calculated according to formula (1):

$\mathrm{\ω}\left(i\right)=\underset{j=1}{\overset{x}{\mathrm{\Σ}}}{k}_j{{\mathrm{\ω}}^i}_j---\left(1\right)$

(6) the actual weight ω of each concrete point of interest in S is calculated according to formula (2) _i:

${\mathrm{\ω}}_i=\mathrm{\α}(1-\frac{\mathrm{Dis}}{R})+\mathrm{\β\ω}\left(i\right)---\backslash *\mathrm{MERGEFORMAT}\left(2\right)$

(7) calculate according to formula (3) probability that target is in each concrete point of interest in S set:

$p_i=\frac{{\mathrm{\ω}}_i}{\underset{i\∈S}{\mathrm{\Σ}}{\mathrm{\ω}}_i}---\left(3\right)$

Wherein, R is error radius, the weight of the jth level point of interest classification belonging to point of interest i, k _jfor constant, and meet with 1 > k _m> k _n> 0, (x>=m > n>=1), α, β are constant, represent the actual weight sum of all points of interest in S set;

(8) point of interest choosing maximum probability in S set is target location.

Step (8): after the point of interest choosing maximum probability in S set is target location, upgrades the weight of the points of interest classification at different levels belonging to point of interest i be selected (j=1..x):

${\mathrm{\ω}}_j^i=(1+k_j){\mathrm{\ω}}_j^i---\left(4\right)$

Weight after renewal, the weight after renewal is used as setting weight during each higher level's point of interest class location next time belonging to selected point of interest i.

Compared with prior art, the present invention has following beneficial effect:

The present invention is based on existing electronic chart, each region is carried out the division of point of interest, according to the behavioural characteristic of target, setting weight is divided to every grade, and then in the error range of location, calculate the probability that target is in each point of interest, choose probability the highest as positioning result, not only positioning result error is little, setting accuracy is much better than existing localization method, and effectively overcome the resolution characteristic of existing location technology shortage crowd behaviour feature, do not possess the problem of the identification advantage of specific crowd location.

Accompanying drawing explanation

Fig. 1 is the schematic diagram of certain locating area.

Embodiment

Below in conjunction with embodiment, the invention will be further described, and embodiments of the present invention include but not limited to the following example.

Embodiment

Present embodiments provide a kind of intelligent locating method based on GPS/AGPS/GPSOne and goal behavior feature, children/old man is configured with locating device with it, realize between this locating device with background server communicating, the opposing party reads locating information by background server.The design concept of the present embodiment is the behavioural characteristic (being drawn by statistics) according to target, corresponding weight is set to different point of interest, and then within the scope of positioning error, calculate the probability that target is in each point of interest, select probability large as final positioning result.

Step one: based on electronic chart, divides multistage point of interest:

By the mode classification of the POI (point of interest) on electronic chart according to high moral map, be divided into level Four: the first order is the large class of affiliated POI, have at present: physical culture and recreational facilities, medical related service, food and beverage sevice, accommodation service, scenic spot, office building, development area, residential quarters, government bodies, science and education and cultural facility, means of transportation, financial service facility, company and enterprise, automobile related facility, as sold, maintenance, high-speed service district viaduct etc. to drive the several types such as relevant POI, service for life, symbolization represent, k represents the type in first-level class catalogue, such as, belong to scenic spot class, in order to distinguish dissimilar point of interest, marks subscript below k.The second level is the further segmentation of the large class in first order POI place, is divided at present: the classifications such as refuelling station, other energy source station, automobile maintenance/decoration, washing bay, sale of automobile, Chinese Restaurant, foreign dining room, symbolization representing, in order to distinguish dissimilar point of interest, below k, marking subscript.The third level is the further segmentation of often kind of second level classification, and be divided into the type such as Sinopec, the sale of Toyota of Guangzhou Automobile Workshop, east wind sale, Zhengzhou daily output maintenance, Sichuan cuisine at present, we use symbol representing, in order to distinguish dissimilar point of interest, below k, marking subscript.Concrete classified information is shown in the high moral map POI classification table of comparisons.The fourth stage is categorized as certain concrete POI, be such as positioned at No. 9, Chenghua district Wanke road " Kai De square. glamour city " the rural base fast food restaurant of No. 04,05A, B1 floor, adopt represent.

Step 2: to point of interest classes different in every one-level, arranges corresponding initial weight, and the point of interest class setting weight sum belonging to same higher level's class in same layer equals 1, and the setting weight sum of all point of interest classes in first order classification equals 1:

The principle of weight setting is, the behavioural characteristic according to target sets.Such as: for child user, according to its behavioural characteristic, for the classification setting weight of often kind of different POI.If first-level class always total n kind is dissimilar, then it is right to need arrange different weights, weight coefficient is respectively α ₁~ α _n, and α ₁+ α ₂+ ... + α _n=1.Equally, need two, reclassify arranges corresponding weight coefficient, if the secondary classification type that always total m kind is different, then corresponding secondary classification weight is respectively β ₁~ β _m, and β ₁+ β ₂+ ... + β _m=1, if always total l kind is dissimilar for reclassify, then corresponding reclassify weight is respectively λ ₁~ λ _land λ ₁+ λ ₂+ ... + λ _l=1.

Step 3:

Shared by concrete point of interest, area size carries out fence to xth level point of interest.Fence technology is existing mature technology, and therefore not to repeat here.

Step 4:

According to the S set of all concrete point of interest in the positioning result compute location center obtained from GPS/AGPS/GPSONE and error radius R, and the distance Dis of each concrete point of interest and the centre of location in set of computations S; Concrete point of interest in S set to include in the circle that is contained in and formed with error radius or with its fence intersected (concrete point of interest), as shown in Figure 1;

Step 5:

The type weight ω (i) of higher level's point of interest class belonging to each concrete point of interest in set of computations S; Suppose to be divided into level Four point of interest, so, the type weight of point of interest i concrete in S set equals wherein, represent the weight of type belonging to this point of interest in the classification of jth level point of interest, k _jfor the level weights constant of the jth level point of interest classification preset meets and 1 > k _m> k _n> 0, (4>=m>=n>=1).Such as: the point of interest class in first order classification is food and beverage sevice, and its weight is 0.2, and level weights constant is 0.0625, point of interest class in the classification of the second level is Chinese Restaurant, level weights constant is 0.125, its weight is 0.3, point of interest class in third level classification is rural base fast food, level weights constant is 0.25, its weight is 0.2, the fourth stage classification in point of interest class be No. 9, Chenghua district Wanke road " Kai De square. glamour city " B1 floor 04, the rural base fast food restaurant of No. 05A, level weights constant is 0.5625, its weight is 0.3, then corresponding, for No. 9, point of interest Chenghua district Wanke road " Kai De square. glamour city " B1 floor 04, the type weight of the rural base fast food restaurant of No. 05A equals: 0.0625*0.2+0.125*0.3+0.25*0.2+0.5625*0.3=0.186

Step road six:

The actual weight ω of each concrete point of interest in S is calculated according to formula (2) _i:

${\mathrm{\ω}}_i=\mathrm{\α}(1-\frac{\mathrm{Dis}}{R})+\mathrm{\β\ω}\left(i\right)---\backslash *\mathrm{MERGEFORMAT}\left(2\right)$

Step 7:

The probability that target is in each concrete point of interest in S set is calculated according to formula (3):

$p_i=\frac{{\mathrm{\ω}}_i}{\underset{i\∈S}{\mathrm{\Σ}}{\mathrm{\ω}}_i}---\left(3\right)$

Wherein, R is error radius, and α, β are constant, represent the actual weight sum of all concrete points of interest in S set;

Step 8:

The point of interest choosing maximum probability in S set is target location.

After the point of interest choosing maximum probability in S set is target location, to the weight of the points of interest classification at different levels belonging to selected point of interest i upgrade the weight of the point of interest classification at different levels belonging to point of interest i be selected (j=1..x):

${\mathrm{\ω}}_j^i=(1+k_j){\mathrm{\ω}}_j^i---\left(4\right)$

Wherein, point of interest classification at different levels refers to first order classification, second level classification, third level classification ... X level is classified.

For in step 5 No. 9, point of interest Chenghua district Wanke road " Kai De square. glamour city " the rural base fast food restaurant of No. 04,05A, B1 floor, if this point of interest is selected, then the new weight of the point of interest class in the fourth stage classification belonging to it is: (1+0.5625) * 0.3=0.469; In third level classification, the new weight of rural base fast food class is (1+0.25) * 0.2=0.25; In the classification of the second level, the new weight of Chinese Restaurant class is (1+0.125) * 0.3=0.3375;

In first order classification, the new weight of food and beverage sevice is (1+0.0625) * 0.2=0.2125.

By said method, according to the behavioural characteristic of target, can realize the accurate location to target, it is located the degree of accuracy of giving and is far superior to existing location technology.

According to above-described embodiment, just the present invention can be realized well.What deserves to be explained is; under prerequisite based on above-mentioned design concept; for solving same technical matters; even if some making on architecture basics disclosed in this invention are without substantial change or polishing; the essence of the technical scheme adopted is still the same with the present invention, therefore it also should in protection scope of the present invention.

Claims (2)

1., based on an intelligent locating method for GPS/AGPS/GPSOne and goal behavior feature, it is characterized in that, comprise the following steps:

(1) all points of interest in electronic chart are carried out the hierarchical classification of X layer, be respectively first order classification, second level classification, third level classification ... X level is classified, wherein, containing dissimilar point of interest class in every first-level class, the first order is categorized as higher level's class of second level classification, the second level is categorized as higher level's class of third level classification, the like, only include certain concrete point of interest in each point of interest class of X level classification;

(2) to point of interest classes different in every one-level, arrange corresponding initial weight, and the point of interest class setting weight sum belonging to same higher level's class in same layer equals 1, the setting weight sum of all point of interest classes in first order classification equals 1;

(3) shared by point of interest concrete in the classification of X level, area size carries out fence to concrete point of interest;

(4) according to from the S set of all concrete point of interest in the positioning result compute location center that GPS/AGPS/GPSONE obtains and error radius R, and the distance Dis of each concrete point of interest and the centre of location in set of computations S;

(5) for concrete point of interest i each in S set, according to the type weight ω (i) of the weight calculation point of interest of affiliated higher level's point of interest classification;

$\mathrm{\ω}\left(i\right)=\underset{j=1}{\overset{x}{\mathrm{\Σ}}}{k}_j{\mathrm{\ω}}_j^i---\left(1\right)$

Wherein represent the weight of the jth level point of interest classification belonging to point of interest i, k _jfor constant, and have $\underset{j=1}{\overset{x}{\mathrm{\Σ}}}{k}_j=1;$

(6) according to the actual weight ω of concrete point of interest each in formula (2) set of computations S _i:

${\mathrm{\ω}}_i=\mathrm{\α}(1-\frac{\mathrm{Dis}}{R})+\mathrm{\β\ω}\left(s\right)$ \*MERGEFORMAT(2)

(7) calculate according to formula (3) probability that target is in each concrete point of interest in S set:

$p_i=\frac{{\mathrm{\ω}}_i}{\underset{i\∈S}{\mathrm{\Σ}}{\mathrm{\ω}}_i}---\left(3\right)$

Wherein, R is error radius, and α, β are constant, represent the actual weight sum of all xth level points of interest in S set;

(8) point of interest choosing maximum probability in S set is target location.

2. a kind of intelligent locating method based on GPS/AGPS/GPSOne and goal behavior feature according to claim 1, it is characterized in that, the point of interest that described step (8) chooses maximum probability in S set upgrades the weight of each higher level's point of interest class belonging to selected point of interest i after being target location according to formula (4) (j=1..x), the weight after renewal is used as setting weight during each higher level's point of interest class location next time belonging to selected point of interest i, and formula (4) is as follows:

${\mathrm{\ω}}_j^i=(1+k_j){\mathrm{\ω}}_j^i---\left(4\right).$