CN106646366A - Visible light positioning method and system based on particle filter algorithm and intelligent equipment - Google Patents
Visible light positioning method and system based on particle filter algorithm and intelligent equipment Download PDFInfo
- Publication number
- CN106646366A CN106646366A CN201611105449.XA CN201611105449A CN106646366A CN 106646366 A CN106646366 A CN 106646366A CN 201611105449 A CN201611105449 A CN 201611105449A CN 106646366 A CN106646366 A CN 106646366A
- Authority
- CN
- China
- Prior art keywords
- led
- target device
- filter algorithm
- particle filter
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01S—RADIO DIRECTION-FINDING; RADIO NAVIGATION; DETERMINING DISTANCE OR VELOCITY BY USE OF RADIO WAVES; LOCATING OR PRESENCE-DETECTING BY USE OF THE REFLECTION OR RERADIATION OF RADIO WAVES; ANALOGOUS ARRANGEMENTS USING OTHER WAVES
- G01S5/00—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations
- G01S5/16—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using electromagnetic waves other than radio waves
Landscapes
- Physics & Mathematics (AREA)
- Electromagnetism (AREA)
- Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Navigation (AREA)
- Position Fixing By Use Of Radio Waves (AREA)
- Traffic Control Systems (AREA)
Abstract
The invention discloses a visible light positioning method based on a particle filter algorithm and intelligent equipment, and the method comprises the steps: converting the intensities of optical signals, detected by a photodiode, of indoor LED anchor nodes and an LED reference point into distances, and carrying out the positioning of target equipment; carrying out the positioning of the target equipment; enabling a system to collect the sensor information of the target equipment; receiving the positioning information and sensor information of the target equipment, simulating a pedestrian track reckoning model, and determining the track of the target equipment; building an offline position database of the LED reference point; measuring the positioning information of the LED reference point, the target equipment and all LED anchor nodes in a system online environment, and finally determining the positioning tracking information of the target equipment. The visible light positioning method and system based on the particle filter algorithm and intelligent equipment improves the anti-sheltering capability of a tracking system, reduces the cost of the reference point, is high in tracking precision, is good in economic and social effects, and can be widely used in the field of indoor positioning and tracking.
Description
Technical field
The present invention relates to indoor positioning follows the trail of field, it is specially a kind of based on the visible of particle filter algorithm and smart machine
Light projection method and system.
Background technology
RSSI：Received Signal Strength Indication, the signal intensity of reception is indicated, is by connecing
The signal strength for receiving determines the distance of signaling point and receiving point, and then carries out a kind of positioning of location Calculation according to corresponding data
Technology.
In recent years, research worker starts to further investigate indoor positioning tracing system.Indoor positioning tracing system
Realization is required for real-time and accurate information.Many positioning and tracer technique are all based on WiFi technology, radio-frequency technique, bluetooth
Technology, GPS technology etc..However, these technologies position indoors and defect are individually present in tracing system.Thus people are constantly
Pursue a kind of cheap, it is high-precision, and can widely used method.
Indoor positioning tracking is carried out using visible ray technology, only LED just need to can be received using photodiode and be sent
Modulated signal, without extra equipment and instrument is increased, mitigate the burden required for experiment.In signal transmission environment, utilize
Triangulation location, by the received signal strength (RSS) that photodiode is detected distance is converted into, and is then carried out by geometric algorithm
Coordinate setting.
Now, smart machine becomes the necessary of people's life, and people obtain various information using smart machine.With it
Meanwhile, substantial amounts of sensor such as acceleration transducer, geomagnetic sensor etc. is integrated on smart machine, is allowed to function increasingly
It is complete.Corresponding information is obtained using these sensors, model can be set up to estimate the movement locus of pedestrian.It is this to carry
Can be referred to as inertial navigation system method for the method for consecutive tracking.Iterative process by this method, was only capable of in the short time
Inside obtain the location tracking of degree of precision, it means that As time goes on, the site error of estimation can be increasing, therefore
It is necessary to be improved.
The content of the invention
In order to solve above-mentioned technical problem, it is an object of the invention to provide a kind of strong antijamming capability, system cost is low, fixed
Position high precision based on particle filter algorithm and the visible ray localization method and system of smart machine.
The technical solution adopted in the present invention is：
The present invention provides a kind of based on particle filter algorithm and the visible ray localization method of smart machine, including following step
Suddenly：
S1, the LED anchor nodes of the indoor setting that system detects on photodiode and the light signal strength of LED reference points
Distance is converted to, and target device is positioned；
S2, the sensor information of system acquisition target device；
S3, system receives the location information and sensor information of target device, simulates pedestrian's reckoning model, and
Track is optimized using particle filter algorithm, optimization processing is iterated to the track of target device and target device is determined
Track；
S4, system measures the location information between the LED reference points and each LED anchor nodes under offline environment,
And set up the offline location database of LED reference points；
S5, system is being measured between the LED reference points, target device and each LED anchor nodes under thread environment
Location information, and by comparing the location information and the offline position data of the LED reference points and each LED anchor nodes
Storehouse, the final location tracking information for determining target device.
Further, in step S4 and S5, respectively repeatedly measurement LED anchor nodes, LED reference points and target device it
Between location information, and gaussian filtering process and average value processing are taken turns doing to location information measurement result.
Further, the particle filter algorithm includes particle initialization, sequential importance sampling and resampling.
Further, the sequential importance sampling stage is calculated using following equation and each particle that standardizes is in K
Weight when secondary,
Xt=ft-1(Xt-1)+nt-1
Zt=ht(Xt)+et
p(xt|Zt-1)=∫ (p (xt|xt-1)p(xt-1|Zt-1))dxt-1
Wherein t be discrete time, XtFor system mode, ZtFor observer state, f (x) is state transition equation, and h (x) is sight
Survey equation, nt-1For process noise, etFor the noise of Gaussian distributed, p (xt|Zt) represent Posterior distrbutionp；
And weight is redistributed using formula below,
Wherein, Wt (i)For i-th particle t weight,Represent reference distribution.
Further, the resampling stage, remove the too low particle of weight using equation below and to replicate weight higher
Particle,
Wherein, MthrFor sample degeneracy degree.
On the other hand, the present invention also provides a kind of based on particle filter algorithm and the visible ray alignment system of smart machine,
Including：
LED anchor nodes, LED reference points and target device, wherein, the LED anchor nodes and LED reference points quantity is equal
At least 3；
Visible ray locating module, for measuring between the LED reference points, target device and each LED anchor nodes
Location information；
Pedestrian's reckoning module, for obtaining corresponding information using target device sensor, and estimates the fortune of pedestrian
Dynamic rail mark；
Track correct module, for by the location information between the target device for obtaining and each LED anchor nodes and work as
The sensor information obtained under front environment carries out track correct, so that it is determined that the track of target device；
Offline location tracking module, for measure under offline environment the LED reference points and each LED anchor nodes it
Between location information, set up offline location database；
Location tracking correcting module, for carrying out Data Matching to target device track and offline fingerprint database data, so as to
Obtain accurate location tracking positional information.
Further, wireless communication module, the radio communication mold are provided with the LED reference points, target device
Block is optical communications module, for optical signal demodulation after coding to be converted into into the signal of telecommunication.
Further, there is acceleration transducer, direction sensor, magnetic field sensor on the target device.
Further, the LED anchor nodes adopt luminous model for the white light LEDs of lambert's mode.
The invention has the beneficial effects as follows：The visible light projection based on particle filter algorithm and smart machine that the present invention is provided
Method and system, using smart machine and wireless technology location tracking is carried out, and just positioning is carried out initially with visible ray, while matching somebody with somebody
Conjunction carries out consecutive tracking tracking using sensor；Then trace information is carried out into track to target device using particle filter algorithm
Amendment, improves the anti-ability of blocking of tracing system, and reducing increases the cost of reference point, and follows the trail of high precision, with good
Good economic and social benefit.
In addition, the present invention is by wireless location market demand Gaussian filter algorithm and Mean Filtering Algorithm, so as to reduce
The impact of some small probabilities, the unexpected incidents of big interference to overall measured value, increased the capacity of resisting disturbance of location tracking.
Description of the drawings
The specific embodiment of the present invention is described further below in conjunction with the accompanying drawings：
Fig. 1 is based on the implementation principle figure of triangulation methods location technology；
Fig. 2 is based on the visible light communication system figure of one embodiment of the invention；
Fig. 3 is the Organization Chart of one embodiment of the invention；
Fig. 4 is the system model figure of one embodiment of the invention；
Fig. 5 is kinestate graph of a relation between sensor；
Fig. 6 is each axle acceleration of motion curve in walking process；
Fig. 7 is particle filter algorithm flow chart；
Fig. 8 is path smooth factor moving direction schematic diagram；
Fig. 9 is the result schematic diagram of an embodiment；
Figure 10 is the schematic diagram of an embodiment；
Figure 11 is the schematic diagram of another embodiment；
Figure 12 is the comparison diagram positioned using various technologies.
Specific embodiment
It should be noted that in the case where not conflicting, the feature in embodiment and embodiment in the application can phase
Mutually combination.
The present invention provides a kind of based on particle filter algorithm and the visible ray localization method of smart machine, including following step
Suddenly：
S1, the LED anchor nodes of the indoor setting that system detects on photodiode and the light signal strength of LED reference points
Distance is converted to, and target device is positioned；
S2, the sensor information of system acquisition target device；
S3, system receives the location information and sensor information of target device, simulates pedestrian's reckoning model, and
Track is optimized using particle filter algorithm, optimization processing is iterated to the track of target device and target device is determined
Track；
S4, system measures the location information between the LED reference points and each LED anchor nodes under offline environment,
And set up the offline location database of LED reference points；
S5, system is being measured between the LED reference points, target device and each LED anchor nodes under thread environment
Location information, and by comparing the location information and the offline position data of the LED reference points and each LED anchor nodes
Storehouse, the final location tracking information for determining target device.
Further, in step S4 and S5, respectively repeatedly measurement LED anchor nodes, LED reference points and target device it
Between location information, and gaussian filtering process and average value processing are taken turns doing to location information measurement result.
Further, the particle filter algorithm includes particle initialization, sequential importance sampling and resampling.
Further, the sequential importance sampling stage is calculated using following equation and each particle that standardizes is in K
Weight when secondary,
Xt=ft-1(Xt-1)+nt-1
Zt=ht(Xt)+et
p(xt|Zt-1)=∫ (p (xt|xt-1)p(xt-1|Zt-1))dxt-1
Wherein t be discrete time, XtFor system mode, ZtFor observer state, f (x) is state transition equation, and h (x) is sight
Survey equation, nt-1For process noise, etFor the noise of Gaussian distributed, p (xt|Zt) represent Posterior distrbutionp；
And weight is redistributed using formula below,
Wherein, Wt (i)For i-th particle t weight,Represent reference distribution.
Further, the resampling stage, remove the too low particle of weight using equation below and to replicate weight higher
Particle,
Wherein, MthrFor sample degeneracy degree.
On the other hand, the present invention also provides a kind of based on particle filter algorithm and the visible ray alignment system of smart machine,
Including：
LED anchor nodes, LED reference points and target device, wherein, the LED anchor nodes and LED reference points quantity is equal
At least 3；
Visible ray locating module, for measuring between the LED reference points, target device and each LED anchor nodes
Location information；
Pedestrian's reckoning module, for obtaining corresponding information using target device sensor, and estimates the fortune of pedestrian
Dynamic rail mark；
Track correct module, for by the location information between the target device for obtaining and each LED anchor nodes and work as
The sensor information obtained under front environment carries out track correct, so that it is determined that the track of target device；
Offline location tracking module, for measure under offline environment the LED reference points and each LED anchor nodes it
Between location information, set up offline location database；
Location tracking correcting module, for carrying out Data Matching to target device track and offline fingerprint database data, so as to
Obtain accurate location tracking positional information.
Further, wireless communication module, the radio communication mold are provided with the LED reference points, target device
Block is optical communications module, for optical signal demodulation after coding to be converted into into the signal of telecommunication.
Further, there is acceleration transducer, direction sensor, magnetic field sensor on the target device.
Further, the LED anchor nodes adopt luminous model for the white light LEDs of lambert's mode.
It is the implementation principle based on triangulation methods location technology with reference to Fig. 1, is with anchor node known to three positions
The center of circle, circle is done with each anchor node to the distance of target to be measured as radius, and three circles meet at the target to be measured.If described treat
Survey target coordinate position be (x, y), it is known that the coordinate of the anchor node 1,2,3 be respectively (x1, y1), (x2, y2) and (x3,
Y3), its distance for arriving the target to be measured is d1, d2, d3, then solve the position that following equations group is obtained the target to be measured
Put coordinate：
(x-x1)2+(y-y1)2=d12
(x-x2)2+(y-y2)2=d22
(x-x3)2+(y-y3)2=d32
With reference to the visible light communication system figure that Fig. 2 is system.Due to there is interference during signal transmission and declining,
Signal communication easily goes wrong, so needing first signal to be carried out into pretreatment.Can be reasonably resistant to respectively by chnnel coding
Interference is planted, so as to reach fidelity.By information-driven LED after modulation.Telecommunications is converted optical signal into using photoelectric switching circuit
Number, then the signal of telecommunication is demodulated, the signal after demodulation is sent to into Intelligent mobile equipment MD, then carry out system location tracking.
Generally, it is believed that the light-emitting mode of LED is lambert's mode, with light source normal into any φ side
Luminous intensity upwards is：
So, in the horizontal plane the illuminance on arbitrary coordinate (X, Y) is：
Wherein I (0) is LED center luminous intensity,For radiation angle, D is transmitter centre distance of the receiver to LED；m
For Lambertian radiation series, its transmitter half-power angle Φ with LED1/2It is relevant.
M=-ln2/ln (cos Φ1/2) formula (3)
In visible ray alignment system, we adopt light-emitting mode for the White LED of lambert's mode.Its receiving light work(
Rate with transmitting luminous power formula be：
Pr=PtH (0) formula (4)
In formula H (0) for optical communication channel DC current gain.
In formula：0≤ψ≤ψFOV, D is the distance between signal transmitting and receiving machine, ψFOVFor the angle of visual field of photoreceiver.ArFor optical detection
The detection area of device,Radiation angle and angle of incidence are respectively with ψ.Ts(ψ) it is the optically filtering gain of receptor, g (ψ) is poly- for optics
The focusing gain of light device.WhereinN is dioptric system coefficient.
With reference to Fig. 3 and 4, the model buildings of the present invention are in the room of a height of 5*5*3 of length and width.Using 4 groups of LED matrixes, its
It is respectively arranged in the surface in room.
First, fingerprint base is set up.The location mechanism of fingerprint base mainly includes two stages：The offline gathered data stage and
Lines matching information phase.In the offline gathered data stage, it is usually required mainly for collect the location information of each reference point (RP).Setting
Reference point RP is placed every 5cm, the location information of 4 groups of LED is collected respectively on each RP, and is stored in fingerprint base.
Repeat related work on each RP, finish until collecting all information.
It is the Organization Chart of one embodiment of the invention with reference to Fig. 3, wherein carrying out On-line matching information, system is according to target MD
The location information collected, by the location information of MD real-time tracking is carried out.In order to reduce the error that online tracing MD is brought, profit
The location tracking information fed back is optimized with the various motion sensors of Intelligent mobile equipment.According to pedestrian's reckoning mould
Type (PRD), can estimate the fact of pedestrian's movement, including speed and direction.
Pedestrian's reckoning model is a kind of method that can be provided based on Newtonian mechanics direct measurement multidate information.In the past
Some researchs show that inertial navigation system structure can be that the mobile device estimated based on Walking Mode and walking distance carries out determining
Position, but high-precision measurement data can not be obtained.In this patent, we utilize the various sensors and weight of mobile device
Factor pair human motion state is optimized estimation.
It is kinestate relation between sensor with reference to Fig. 5.There are close ties between kinestate and sensor.Cause
This, is studied by the sensor to general smart machine, generally has 5 kinds of sensors to have an impact kinestate.
The sensor of the Intelligent mobile equipment of table 1
The x of acceleration transducer, y and z-axis three axles corresponding with the earth are not the same, there is an angle between them
Degree.It is believed that speed is not only relevant with acceleration, and also there is relation with azimuth and acceleration of gravity.Can pass through following
Formula calculate speed：
R (ψ, ρ, γ)=C-1(ψ, ρ, γ)=CT(ψ, ρ, γ) formula (10)
HereIt is the acceleration obtained from acceleration transducer,It is the acceleration in the relative earth's axis
Value, and R (ψ, ρ, γ) is transfer matrix,It is the velocity amplitude at earth's axis moment.
Human motion direction can be tried to achieve using equation below:
HereIt is the direction of motion at certain moment.Although high-precision direction can not be obtained from direction sensor
Value, but general direction can be obtained.Deflection error is corrected using the weight of particle filter.In order to obtain more
Plus accurate direction and velocity amplitude, using reasonable sensor sample frequency.In this programme, the frequency for adopting is for 50HZ.
It is each axle acceleration of motion curve in user acceleration walking process with reference to Fig. 6.
Location information and movable information are combined the further path optimizing of particle filter algorithm by next step.
Particle filter, is based on the approximate Bayesian filter algorithm of Monte Carlo simulation.Its core concept be with some from
Scattered stochastical sampling point (particle) carrys out the probability density function of approximation system stochastic variable, and with sample average integral operation is replaced, from
And obtain the minimum variance estimate of state.
Bayesian filter algorithm is used for the Probability estimate from observation and goes out current state.Particle filter method is a kind of online
Posterior probability estimating algorithm, by with reference to importance sampling and Monte carlo algorithm can be used for estimate posterior probability density.When
Can solve the problem that the decay of particle filter, and computer, when can have sufficiently large computing capability, particle filter algorithm is just
Non-gaussian and non-linear system status estimation can be advantageously applied to.In orientation problem, target location Jing commonly uses state space
Model is represented, including state equation (12) and observational equation (13).
Xt=ft-1(Xt-1)+nt-1Formula (12)
Zt=ht(Xt)+etFormula (13)
Wherein t be discrete time, XtFor system mode, ZtFor observer state, f (x) is state transition equation, and h (x) is sight
Survey equation, nt-1For process noise, etFor the noise of Gaussian distributed.
Bayesian filter algorithm includes prediction and corrects.We assume that initial probability density P (x0) and probability density P
(x0) time be known, the formula in forecast period and on-line amending stage is as follows：
p(xt|Zt-1)=∫ (p (xt|xt-1)p(xt-1|Zt-1))dxt-1Formula (14)
Wherein ZtFor observation vector, XtFor state value.Formula (11) is difficult a large amount of calculating.Particle filter algorithm can be effective
Ground solves the limitation of Bayesian filter algorithm.Particle filter algorithm can be considered an approximation of Bayesian filter algorithm.Grain
Sub- filtering algorithm includes three steps：Particle initialization, sequential importance sampling and resampling.With reference to Fig. 7.
Particle is initialized：Initialization particle.Particle cloud of the initial sample typically from the Gauss distribution in physical location
Meansigma methodss.All of particle has identical weight in the starting stage.
Sequential importance sampling：In this stage, calculated using formula (12) (13) (14) (15) and each grain that standardizes
Weight of the son in kth.Next weight is redistributed using formula (16).
Resampling：The resampling stage is an important stage of particle filter algorithm.The ultimate principle of resampling is to remove
The too low particle of weight.In this stage, the too low particle of weight is removed using equation below, replicate the higher particle of weight.
It is resulting due to the measurement noise that there is visible ray and occlusion issue in the system that visible ray-fingerprint base is built
The degree of accuracy of location tracking information is not high.Using the advantage of particle filter algorithm, this set system can be optimized, so as to reduce positioning
Tracking error.
By the way that with reference to inertial navigation system information, particle filter algorithm can be used for the measurement of visible light signal.Particle filter
Algorithm can be conveniently incorporated into the movable information of the sensor from inertial navigation system.Because particle is used with mobile device
The motor behavior of person has direct relation, and these movable informations can guide particle.
Formula (12) and (13) are embodied, formula (18) and (19) can be just obtained
WhereinRepresent the state vector of each particle in t, TsIt is the run time twice between measurement,
nt-1It is process noise, etIt is the noise of Gaussian distributed.
In visible ray location tracking system, with reference to particle filter algorithm and Intelligent mobile equipment location tracking process such as
Lower step：
1st, fingerprint base is set up
Visible ray-fingerprint base system includes off-line phase and on-line stage.Off-line phase needs to measure determining for each point
Position information includes coordinate and corresponding RSS, and on-line stage carries out real-time positioning tracking to target device.
2nd, kinestate is calculated
System can collect in smart machine the various data of sensor and calculate correlated characteristic vector, then according to public affairs
Formula (7) and (11) calculate azimuth and speed.
3rd, particle filter
After particle initialization, each particle can be shifted as a new particle through state transition equation (11).
All of particle can all be transferred to new position.All of particle can all move to a new representative position.Root
According to formula (20) and (21), power difference is calculated using Euclidean algorithm, at the same time by recalculating weight.
In addition, when motion vector is added in pedestrian's track deception, we are by the particle weights factorAs path
Smoothing factor, the different particle weights factor has different paths, and path smooth factor moving direction schematic diagram is shown in Fig. 8.
Wherein, θtFor calculating the direction of motion in t sensing data,It is to introduce road in i-th particle of t
The direction of motion after the smoothing factor of footpath.
4. position estimation
We utilize weighting method, and by formula (23) the new position of smart machine is obtained.Then repeat step 2 is until stopping
Only.
In order to verify the feasibility of proposed method, we are described with reference to emulation experiment.Hereinafter experiment is to be based on
Smart machine and Matlab emulation are obtained：
The experiment porch parameter of table 2
Mobile phone application is developed in this experiment, for the initial data of collecting sensor, and calculates target intelligence
The movable information of energy equipment includes speed and direction.Motion information transmission carries out Matlab emulation experiments on computer.
In order to inquire into the feasibility of proposed method, following emulation experiment is carried out to model by Matlab.Its
It is (0.5,0.5) that the parameter and room model of middle LED is shown in Table the distribution coordinate of (2) LED, (4.5,0.5), (4.5,4.5),
(0.5,4.5).Emulation experiment is simulated in the room of 100*100 reference point, and each reference point is separated by 0.05m.
Table 3
200 points are selected in this experiment as original path (solid-line paths in Fig. 9), and in Matlab the system is built
Emulation is simulated, the simulaed path for obtaining is the dotted path of Fig. 9, as it can be seen in figure 9 that wherein regular curve is real
Border path；Two paths are most of consistent, including the path for turning to.Therefore the method proposed with this programme is demonstrated to mesh
It is feasible that mark smart machine carries out location tracking.
In order to further verify the feasibility of proposed method, the analog simulation experiment of one multipath.Reference picture
Shown in 10, dashed path is sensor-based PDR paths, it can be observed that PDR paths are not fine, pushing away over time
Move, original path is increasingly deviateed in path.Mainly due at the beginning when, sensor can obtain preferable data, but
That error occurs, with error accumulation will deviation it is very big.Dotted path is that particle filter is based in VISIBLE LIGHT SYSTEM
The estimation path of algorithm and sensor, it is seen that it coincide substantially with original path.Direction can be rapidly captured using sensor
Change, can reduce error using particle filter algorithm.
The displacement problem that VISIBLE LIGHT SYSTEM based on fingerprint base is present is when location tracking is carried out, if barrier hides
Gear, then resulting location tracking error will be very big.So to be checked using the system blocked to this when the method
Problem makes moderate progress.We simulate the scene for having shelter in emulation experiment, have the position of arrow with reference to Figure 11, it can be seen that
Original path is all largely deviates from that when following the trail of path and reaching and exist where shelter, but is calculated based on particle filter
Method of the visible ray of the method and smart machine-fingerprint base location tracking method substantially than being visible ray-fingerprint base is good, and it can be very
It revert to soon in original path.
Figure 12 is contrasted to several positioning tracking technologies.Various filtering modes are to the accumulation of error nor the same.
If simply using sensor, As time goes on, error can be increasing.Also may not be certain to positioning using Kalman filtering
Tracking error has clear improvement.The other methods of contrast, the location tracking method that this programme is proposed increases to precision.
Visible ray-fingerprint base the system based on particle filter algorithm and smart machine is proposed in this programme, is contributed to
Improve the degree of accuracy of indoor positioning tracer technique.Meanwhile, our Binding experiments and emulation can come demonstrate proposed method
By property, the location tracking error obtained in experiment and emulation is 8cm.Even in complex environment, based on what is proposed
The positioning result of method also can Fast Convergent.Simultaneously, it is seen that the occlusion issue in photosystem is also improved well.
The present invention provide based on particle filter algorithm and the visible ray localization method and system of smart machine, using intelligence
Equipment and wireless technology carry out location tracking, just positioning are carried out initially with visible ray, while being connected with the use of sensor
Continuous location tracking；Then trace information is carried out into track correct to target device using particle filter algorithm, improves tracking system
The anti-ability of blocking of system, reducing increases the cost of reference point, and follows the trail of high precision, with good economy and society's effect
Benefit.
In addition, the present invention is by wireless location market demand Gaussian filter algorithm and Mean Filtering Algorithm, so as to reduce
The impact of some small probabilities, the unexpected incidents of big interference to overall measured value, increased the capacity of resisting disturbance of location tracking.
It is more than that the preferable enforcement to the present invention is illustrated, but the invention is not limited to the enforcement
Example, those of ordinary skill in the art can also make a variety of equivalent variations on the premise of without prejudice to spirit of the invention or replace
Change, the deformation or replacement of these equivalents are all contained in the application claim limited range.
Claims (9)
1. it is a kind of based on particle filter algorithm and the visible ray localization method of smart machine, it is characterised in that to comprise the following steps：
S1, the LED anchor nodes of the indoor setting that system detects on photodiode and the light signal strength conversion of LED reference points
It is distance, and target device is positioned；
S2, the sensor information of system acquisition target device；
S3, system receives the location information and sensor information of target device, simulates pedestrian's reckoning model, and adopts
Particle filter algorithm is optimized to track, is iterated optimization processing to the track of target device and determines the rail of target device
Mark；
S4, system measures the location information between the LED reference points and each LED anchor nodes under offline environment, and builds
The offline location database of vertical LED reference points；
S5, system is measuring the positioning between the LED reference points, target device and each LED anchor nodes under thread environment
Information, and by comparing the location information and the offline location database of the LED reference points and each LED anchor nodes,
The final location tracking information for determining target device.
2. according to claim 1 based on particle filter algorithm and the visible ray localization method of smart machine, its feature exists
In, in step S4 and S5, LED anchor nodes, the location information between LED reference points and target device are repeatedly measured respectively,
And gaussian filtering process and average value processing are taken turns doing to location information measurement result.
3. according to claim 1 based on particle filter algorithm and the visible ray localization method of smart machine, its feature exists
In the particle filter algorithm includes particle initialization, sequential importance sampling and resampling.
4. according to claim 3 based on particle filter algorithm and the visible ray localization method of smart machine, its feature exists
In, the sequential importance sampling stage is calculated and weight of each particle in kth of standardizing using following equation,
Xt=ft-1(Xt-1)+nt-1
Zt=ht(Xt)+et
p(xt|Zt-1)=∫ (p (xt|xt-1)p(xt-1|Zt-1))dxt-1
Wherein t be discrete time, XtFor system mode, ZtFor observer state, f (x) is state transition equation, and h (x) is observation side
Journey, nt-1For process noise, etFor the noise of Gaussian distributed, p (xt|Zt) represent Posterior distrbutionp；
And weight is redistributed using formula below,
Wherein, Wt (i)For i-th particle t weight,
Represent reference distribution.
5. according to claim 4 based on particle filter algorithm and the visible ray localization method of smart machine, its feature exists
In, the resampling stage, remove the too low particle of weight using equation below and replicate the higher particle of weight,
Wherein, MthrFor sample degeneracy degree.
6. it is a kind of based on particle filter algorithm and the visible ray alignment system of smart machine, it is characterised in that to include：
LED anchor nodes, LED reference points and target device, wherein, the LED anchor nodes and LED reference points quantity is at least
For 3；
Visible ray locating module, for measuring the positioning between the LED reference points, target device and each LED anchor nodes
Information；
Pedestrian's reckoning module, for obtaining corresponding information using target device sensor, and estimates the motion rail of pedestrian
Mark；
Track correct module, for by the location information between the target device for obtaining and each LED anchor nodes and working as front ring
The sensor information obtained under border carries out track correct,
So that it is determined that the track of target device；
Offline location tracking module, for measuring under offline environment between the LED reference points and each LED anchor nodes
Location information, sets up offline location database；
Location tracking correcting module, for carrying out Data Matching to target device track and offline fingerprint database data, so as to obtain
Accurate location tracking positional information.
7. according to claim 6 a kind of based on particle filter algorithm and the visible ray alignment system of smart machine, it is special
Levy and be, wireless communication module is provided with the LED reference points, target device, the wireless communication module is light and leads to
Letter module, for optical signal demodulation after coding to be converted into into the signal of telecommunication.
8. according to claim 7 a kind of based on particle filter algorithm and the visible ray alignment system of smart machine, it is special
Levy and be, there is acceleration transducer, direction sensor, magnetic field sensor on the target device.
9. according to claim 8 a kind of based on particle filter algorithm and the visible ray alignment system of smart machine, it is special
Levy and be, the LED anchor nodes adopt luminous model for the white light LEDs of lambert's mode.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611105449.XA CN106646366A (en) | 2016-12-05 | 2016-12-05 | Visible light positioning method and system based on particle filter algorithm and intelligent equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201611105449.XA CN106646366A (en) | 2016-12-05 | 2016-12-05 | Visible light positioning method and system based on particle filter algorithm and intelligent equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106646366A true CN106646366A (en) | 2017-05-10 |
Family
ID=58819136
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201611105449.XA Pending CN106646366A (en) | 2016-12-05 | 2016-12-05 | Visible light positioning method and system based on particle filter algorithm and intelligent equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106646366A (en) |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107613459A (en) * | 2017-09-08 | 2018-01-19 | 华南理工大学 | A kind of high-precision visible ray localization method and system based on particle group optimizing |
| CN108680160A (en) * | 2018-03-30 | 2018-10-19 | 深圳清创新科技有限公司 | Indoor positioning, air navigation aid, device, storage medium and computer equipment |
| CN109001677A (en) * | 2018-06-05 | 2018-12-14 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of mobile robot visible light localization method based on reckoning |
| CN109039458A (en) * | 2018-08-06 | 2018-12-18 | 杭州电子科技大学 | A kind of indoor locating system and method |
| CN109116298A (en) * | 2017-06-22 | 2019-01-01 | 中兴通讯股份有限公司 | A kind of localization method, storage medium and positioning system |
| CN110677813A (en) * | 2018-07-02 | 2020-01-10 | 北京信息科技大学 | Visible light indoor positioning method of DV-hop based on fruit fly correction |
| CN111198365A (en) * | 2020-01-16 | 2020-05-26 | 东方红卫星移动通信有限公司 | Indoor positioning method based on radio frequency signal |
| CN111860613A (en) * | 2020-06-30 | 2020-10-30 | 江苏科技大学 | Multi-target tracking and state prediction method based on multi-source heterogeneous signals |
| CN112865864A (en) * | 2020-12-31 | 2021-05-28 | 华南师范大学 | Transmitting signal simulation method and signal simulator of visible light communication system |
| CN113132006A (en) * | 2021-04-23 | 2021-07-16 | 湖南大学 | High-precision visible light positioning method for moving vehicle based on image sensor |
| CN113790728A (en) * | 2021-09-29 | 2021-12-14 | 佛山市南海区广工大数控装备协同创新研究院 | Loosely-coupled multi-sensor fusion positioning algorithm based on visual odometer |
| CN114088095A (en) * | 2021-10-29 | 2022-02-25 | 鹏城实验室 | Three-dimensional indoor positioning method based on photodiode |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105376855A (en) * | 2015-09-09 | 2016-03-02 | 华南师范大学 | Indoor positioning method and system for adaptive obstacle judgment based on wireless technology |
| CN105676175A (en) * | 2016-01-29 | 2016-06-15 | 中国矿业大学(北京) | Monitoring and positioning and personnel help system having high downhole disaster tolerance |
| CN105933975A (en) * | 2016-04-11 | 2016-09-07 | 南京邮电大学 | WiFi fingerprint-based accuracy improved indoor positioning method |
| CN106153049A (en) * | 2016-08-19 | 2016-11-23 | 北京羲和科技有限公司 | A kind of indoor orientation method and device |
-
2016
- 2016-12-05 CN CN201611105449.XA patent/CN106646366A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105376855A (en) * | 2015-09-09 | 2016-03-02 | 华南师范大学 | Indoor positioning method and system for adaptive obstacle judgment based on wireless technology |
| CN105676175A (en) * | 2016-01-29 | 2016-06-15 | 中国矿业大学(北京) | Monitoring and positioning and personnel help system having high downhole disaster tolerance |
| CN105933975A (en) * | 2016-04-11 | 2016-09-07 | 南京邮电大学 | WiFi fingerprint-based accuracy improved indoor positioning method |
| CN106153049A (en) * | 2016-08-19 | 2016-11-23 | 北京羲和科技有限公司 | A kind of indoor orientation method and device |
Non-Patent Citations (3)
| Title |
|---|
| 唐钱胜: "基于LED通信的室内定位方法研究", 《万方学位论文数据库》 * |
| 董景新等: "《现代控制理论与方法概论》", 30 June 2016, 北京：清华大学出版社 * |
| 陈明剑等: "《星载GNSS卫星编队相对导航技术》", 31 January 2016, 武汉：中国地质大学出版社 * |
Cited By (19)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109116298B (en) * | 2017-06-22 | 2023-12-22 | 中兴通讯股份有限公司 | Positioning method, storage medium and positioning system |
| CN109116298A (en) * | 2017-06-22 | 2019-01-01 | 中兴通讯股份有限公司 | A kind of localization method, storage medium and positioning system |
| CN107613459B (en) * | 2017-09-08 | 2019-11-15 | 华南理工大学 | A kind of high-precision visible light localization method and system based on particle group optimizing |
| CN107613459A (en) * | 2017-09-08 | 2018-01-19 | 华南理工大学 | A kind of high-precision visible ray localization method and system based on particle group optimizing |
| CN108680160B (en) * | 2018-03-30 | 2021-09-03 | 深圳一清创新科技有限公司 | Indoor positioning and navigation method and device, storage medium and computer equipment |
| CN108680160A (en) * | 2018-03-30 | 2018-10-19 | 深圳清创新科技有限公司 | Indoor positioning, air navigation aid, device, storage medium and computer equipment |
| CN109001677A (en) * | 2018-06-05 | 2018-12-14 | 佛山市南海区广工大数控装备协同创新研究院 | A kind of mobile robot visible light localization method based on reckoning |
| CN110677813B (en) * | 2018-07-02 | 2021-10-22 | 北京信息科技大学 | Visible light indoor positioning method of DV-hop based on fruit fly correction |
| CN110677813A (en) * | 2018-07-02 | 2020-01-10 | 北京信息科技大学 | Visible light indoor positioning method of DV-hop based on fruit fly correction |
| CN109039458A (en) * | 2018-08-06 | 2018-12-18 | 杭州电子科技大学 | A kind of indoor locating system and method |
| CN111198365A (en) * | 2020-01-16 | 2020-05-26 | 东方红卫星移动通信有限公司 | Indoor positioning method based on radio frequency signal |
| CN111860613A (en) * | 2020-06-30 | 2020-10-30 | 江苏科技大学 | Multi-target tracking and state prediction method based on multi-source heterogeneous signals |
| CN111860613B (en) * | 2020-06-30 | 2023-10-27 | 江苏科技大学 | Multi-target tracking and state predicting method based on multi-source heterogeneous signals |
| CN112865864A (en) * | 2020-12-31 | 2021-05-28 | 华南师范大学 | Transmitting signal simulation method and signal simulator of visible light communication system |
| CN112865864B (en) * | 2020-12-31 | 2022-03-15 | 华南师范大学 | Transmitting signal simulation method and signal simulator of visible light communication system |
| CN113132006A (en) * | 2021-04-23 | 2021-07-16 | 湖南大学 | High-precision visible light positioning method for moving vehicle based on image sensor |
| CN113790728A (en) * | 2021-09-29 | 2021-12-14 | 佛山市南海区广工大数控装备协同创新研究院 | Loosely-coupled multi-sensor fusion positioning algorithm based on visual odometer |
| CN114088095A (en) * | 2021-10-29 | 2022-02-25 | 鹏城实验室 | Three-dimensional indoor positioning method based on photodiode |
| CN114088095B (en) * | 2021-10-29 | 2023-07-25 | 鹏城实验室 | Three-dimensional indoor positioning method based on photodiode |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106646366A (en) | Visible light positioning method and system based on particle filter algorithm and intelligent equipment | |
| Li et al. | Toward location-enabled IoT (LE-IoT): IoT positioning techniques, error sources, and error mitigation | |
| CN104655137B (en) | The Wi Fi received signals fingerprint location algorithms of pedestrian's flying track conjecture auxiliary | |
| CN112533163B (en) | Indoor positioning method based on NB-IoT (NB-IoT) improved fusion ultra-wideband and Bluetooth | |
| US9292936B2 (en) | Method and apparatus for determining location | |
| CN103884337B (en) | Multi-attitude step correction positioning system and method | |
| CN105043380A (en) | Indoor navigation method based on a micro electro mechanical system, WiFi (Wireless Fidelity) positioning and magnetic field matching | |
| Ali et al. | Tightly coupling fusion of UWB ranging and IMU pedestrian dead reckoning for indoor localization | |
| CN109725338A (en) | Indoor and outdoor integration positioning system based on UWB and GPS | |
| CN106031263A (en) | Method and system for location estimation | |
| CN106412826A (en) | Indoor positioning method and positioning device based on multi-source information fusion | |
| CN105704652A (en) | Method for building and optimizing fingerprint database in WLAN/Bluetooth positioning processes | |
| US10557711B2 (en) | Apparatus for inferring pedestrian position based on pedestrian movement detection, and method therefor | |
| CN107179079A (en) | The indoor orientation method merged based on PDR with earth magnetism | |
| CN107270889B (en) | Indoor positioning method and positioning system based on geomagnetic map | |
| CN105004340A (en) | Inertial navigation-fingerprint location-combined positioning error correction method | |
| CN108120436A (en) | Real scene navigation method in a kind of iBeacon auxiliary earth magnetism room | |
| CN103561462A (en) | Indoor positioning system and method totally based on smart mobile terminal platform | |
| CN106248107A (en) | A kind of flight path based on indoor earth magnetism path matching infers calibration steps and device | |
| CN111698774B (en) | Indoor positioning method and device based on multi-source information fusion | |
| CN103997717A (en) | Real-time indoor positioning system and method | |
| CN108413965A (en) | A kind of indoor and outdoor crusing robot integrated system and crusing robot air navigation aid | |
| CN109506653A (en) | Based on the indoor positioning pedestrian's method for improving particle filter under NLOS environment | |
| CN109141427A (en) | EKF localization method under nlos environment based on distance and angle probabilistic model | |
| CN103152826A (en) | Moving target tracking method based on NLOS (non line of sight) state inspection compensation |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170510 |