CN108710104A - The method and system of object for coal mine down-hole tunnel positioned in real time - Google Patents
The method and system of object for coal mine down-hole tunnel positioned in real time Download PDFInfo
- Publication number
- CN108710104A CN108710104A CN201810353671.4A CN201810353671A CN108710104A CN 108710104 A CN108710104 A CN 108710104A CN 201810353671 A CN201810353671 A CN 201810353671A CN 108710104 A CN108710104 A CN 108710104A
- Authority
- CN
- China
- Prior art keywords
- virtual reference
- signal strength
- received signal
- node
- strength indicator
- 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
- G01S5/02—Position-fixing by co-ordinating two or more direction or position line determinations; Position-fixing by co-ordinating two or more distance determinations using radio waves
- G01S5/0252—Radio frequency fingerprinting
Abstract
The present invention provides a kind of method and systems of the object for coal mine down-hole tunnel positioned in real time, wherein the method positioned in real time, including：It determines the position coordinates of the position coordinates and virtual reference node of the anchor node being arranged in tunnel, and determines the received signal strength indicator value of virtual reference node；Form offline fingerprint database；The proximity mapping model of the received signal strength indicator value and the received signal strength indicator value of the neighbouring anchor node of each virtual reference node of each virtual reference node of training；Form real time fingerprint database；The mapping relations for learning the received signal strength indicator value and position coordinates of virtual reference node, determine the position coordinates of object.Technical solution through the invention, using the fingerprint matching location algorithm based on non-ranging mode, avoid the range measurement error of distance measuring type location algorithm introducing, it is suppressed that multipath effect and shadow fading problem further improve the precision of underground coal mine positioned in real time.
Description
Technical field
The present invention relates to Mine Monitoring technical fields, in particular to a kind of object for coal mine down-hole tunnel
The method positioned in real time and a kind of object for coal mine down-hole tunnel real-time positioning system.
Background technology
Coal is the main body of the disposable energy-consuming in China, and the development relationship of coal industry economic life line of the country.However,
The most of coal mines in China belong to underground job, and coal geology construction is complicated, and gas density is high, these become the safe thing of China's coal-mine
Therefore incidence the reason of remaining high.When a fault occurs, coal mine downhole safety positioning system can position rapidly, and rescue is stranded
Personnel repair mechanical equipment, race against time for rescue.Meanwhile there is important meaning to personnel in the pit's management and running.Therefore, coal
Security positioning system is the important leverage of underground coal mine modernization safety in production under mine.Hundreds of meters of coal mine roadway depths underground is very
Supreme km, widely used GPS (Global Positioning System) positioning system can not in underground on ground at present
It uses.In recent years, wireless sensor network development is swift and violent, and is gradually applied to various fields of recent life so that wireless sensing node
Location technology is applied to underground coal mine and is possibly realized.
Currently, the location algorithm based on ranging includes being based on arrival time (Time in wireless sensor network location technology
Of Arrival, TOA), angle of arrival (Angle of arrival, AOA), received signal strength indicator value (Receive
Signal Strength Indicator, received signal strength indicator value) location algorithm.Underground coal mine environment is complicated, multipath
Transmission and shadow effect problem are serious, and error is larger under the conditions of TOA, AOA location technology non line of sight.Referred to based on received signal strength
The location model of indicating value, indoors using relatively broad in terms of ranging localization.However, underground coal mine environment is complicated and changeable, establish
It is suitble to the wireless channel model of underground coal mine environment more difficult.
Although in addition, received signal strength indicator value is protected from environmental to will produce fluctuation, but still reflect and should this moment
The state distribution of received signal strength indicator value at position.Fingerprint matching algorithm inhibits multipath effect and the moon to a certain extent
Shadow fading problem, but the fingerprint database that fingerprint matching algorithm is built offline cannot be solved with environment real-time update problems demand.
For this problem, many experts and scholars are studied.Lin Yiming, Luo Haiyong et al. exist《Grain based on dynamic Radio map
Son filtering indoor wireless location algorithm》The database update mode based on proximity relations model is proposed in one text.He Yanjun etc.
People is in article《Frequency modulation localization method based on dynamic radio-frequency fingerprint》It is middle to be fitted proximity relations using arithmetic of linearity regression
Model.It is non-to be presented in linear model fitting actual environment although multiple linear regression model can realize the update of database
Linear data can cause larger position error.The problem of linear fitting will produce large error, piece-wise linear is forced
Near-lying mode type and Nonlinear Learning model (such as Local Liner Prediction) are gradually applied to database update queue.Xia Linyuan exists
《Real-time and adaptive indoor orientation method research under more base station modes》In mention using BP neural network and distance reciprocal plus
Power interpolation method establishes proximity relations model.However this mode all virtual reference nodes in long and narrow tunnel are received as one
The neighbor point of the anchor node of signal will produce larger position error.
Invention content
The present invention is directed to solve at least one of the technical problems existing in the prior art or related technologies.
For this purpose, one aspect of the present invention is, a kind of the real-time fixed of the object for coal mine down-hole tunnel is provided
The method of position.
Another aspect of the present invention is, provides a kind of positioning in real time for the object for coal mine down-hole tunnel
System.
In view of this, the technical solution of first aspect present invention provides a kind of object for coal mine down-hole tunnel
The method positioned in real time, including：Determine in tunnel be arranged it is several can receiving and transmitting signal anchor node position coordinates and multiple void
The position coordinates of quasi- reference mode, and determine the received signal strength indicator value of each virtual reference node；According to each virtual
The position coordinates of the received signal strength indicator value of reference mode and each virtual reference node form offline fingerprint database；Instruction
The reception for practicing the received signal strength indicator value of each virtual reference node and the neighbouring anchor node of each virtual reference node is believed
The proximity mapping model of number strength indicator value；Real time fingerprint is formed according to the offline fingerprint database of proximity mapping model real-time update
Database；The mapping for learning the received signal strength indicator value of virtual reference node and position coordinates in real time fingerprint database is closed
System, according to the received signal strength indicator value of object, determines the position coordinates of object, wherein when one of anchor node
When emitting signal, virtual reference node, object and other anchor nodes receive signal, and the quantity of virtual reference node is more than anchor section
The quantity of point.
In the technical scheme, several anchor nodes and multiple virtual reference sections are set in the tunnel of underground coal mine first
Point, anchor node can receiving and transmitting signal, when one of anchor node emit signal when, virtual reference node, object and other anchor sections
Point receives signal, and the quantity of virtual reference node is more than the quantity of anchor node, received signal strength indicator value, the target of anchor node
The received signal strength indicator value of object can be detected to obtain by the equipment outside mine.Since anchor node, virtual reference node are advance
Setting, therefore the position coordinates of anchor node, the position coordinates of multiple virtual reference nodes and manual testing can be obtained ahead of time
The received signal strength indicator value of obtained virtual reference node, according to the received signal strength indicator of each virtual reference node
Value and position coordinates, obtain the finger print information of virtual reference node, and then obtain offline fingerprint database.
Since the variation of tunnel environment can influence the received signal strength indicator value of the anchor node of reception signal, and due to void
The position coordinates of quasi- reference mode are it is known that it is proximity relations, closer, the phase of distance to receive the anchor node of signal and virtual reference node
Like spend it is higher, linear relationship is more apparent, thus by the received signal strength indicator value of each virtual reference node of training with
The proximity mapping model of the received signal strength indicator value of the neighbouring anchor node of each virtual reference node, obtaining can real-time update
Virtual reference node real time fingerprint database, without when tunnel environment changes again by manually virtual into test
The received signal strength indicator value of reference mode, is effectively saved manual labor.
Finally, learn the received signal strength indicator value and position coordinates of virtual reference node in real time fingerprint database
Mapping relations, according to the received signal strength indicator value of object, it may be determined that the position coordinates of object.
In the technical program, using the fingerprint matching location algorithm based on non-ranging mode, avoids distance measuring type positioning and calculate
The range measurement error that method introduces, it is suppressed that multipath effect and shadow fading problem；Meanwhile training proximity mapping model is in real time more
New database avoids the problem that fingerprint matching algorithm establishes fingerprint database distortion offline, further improves underground coal mine
The precision positioned in real time.
It should be noted that the test mode of the received signal strength indicator value of each virtual reference node is：Each anchor
Node emits signal as transmitting signal node successively, and tester carries ZigBee node in each virtual reference node location
Place receives the received signal strength indicator value that all anchor nodes are sent.
In the above-mentioned technical solutions, further, the received signal strength indicator value of each virtual reference node of training refers to
The proximity mapping model of line and the received signal strength indicator value fingerprint of the neighbouring anchor node of each virtual reference node, it is specific to wrap
It includes：According to the principle of coordinate Euclidean distance minimum at least one virtual reference node is distributed to each anchor node for receiving signal；
The received signal strength indicator value of each virtual reference node and each virtual reference node are trained using BP neural network algorithm
Neighbouring anchor node received signal strength indicator value proximity mapping model.
In the technical scheme, an anchor node emits signal, other anchor nodes receive signal, determine the anchor for receiving signal
The method of the neighbouring virtual reference reference mode of node is：According to the principle of coordinate Euclidean distance minimum to each reception signal
Anchor node distributes at least one virtual reference node, i.e., the anchor node of one reception signal corresponds at least one neighbouring virtual ginseng
Examine reference mode, the anchor node of the corresponding neighbouring reception signal of each virtual reference reference mode is different.Training proximity mapping
The method of model is：Using BP neural network algorithm train the received signal strength indicator value of each virtual reference node with it is each
The proximity mapping model of the received signal strength indicator value of the neighbouring anchor node of virtual reference node.By establishing proximity mapping mould
Type, when the received signal strength indicator value of the anchor node of reception signal changes with tunnel environment, virtual reference node connects
Receiving signal strength indication value can also change therewith, i.e. the finger print information real-time update of real-time virtual reference mode, and then obtain reality
When fingerprint database so that the measurement of the position coordinates of object is more accurate.
In the above-mentioned technical solutions, further, learn the reception signal of virtual reference node in real time fingerprint database
The mapping relations of strength indicator value and position coordinates determine the position of object according to the received signal strength indicator value of object
Coordinate is set, is specifically included：Learn the reception of virtual reference node in real time fingerprint database according to PSO-BP neural network algorithms
The mapping relations of signal strength indication value and position coordinates；According to the received signal strength indicator value and mapping relations of object,
Determine the position coordinates of object.
In the technical scheme, virtual reference node in real time fingerprint database is learnt according to PSO-BP neural network algorithms
Received signal strength indicator value and position coordinates mapping relations, according to the received signal strength indicator value of object with mapping
Relationship, you can the position coordinates for determining object further improve the precision of the position coordinates of object.
The technical solution of the second aspect of the present invention provides a kind of the real-time fixed of the object for coal mine down-hole tunnel
Position system, including：Determination unit, for determining that be arranged in tunnel several can the position coordinates of anchor node of receiving and transmitting signal and more
The position coordinates of a virtual reference node, and determine the received signal strength indicator value of each virtual reference node；First data
Finishing unit is used for the position of the received signal strength indicator value and each virtual reference node according to each virtual reference node
Coordinate forms offline fingerprint database；Proximity mapping model training unit, for training the reception of each virtual reference node to believe
The proximity mapping model of number strength indicator value and the received signal strength indicator value of the neighbouring anchor node of each virtual reference node；
Second data preparation unit, for forming real time fingerprint data according to the offline fingerprint database of proximity mapping model real-time update
Library；Position coordinates determination unit, the received signal strength indicator value for learning virtual reference node in real time fingerprint database
The position coordinates of object are determined according to the received signal strength indicator value of object with the mapping relations of position coordinates,
In, when one of anchor node emits signal, virtual reference node, object and other anchor nodes receive signal, virtual to join
The quantity for examining node is more than the quantity of anchor node.
In the technical scheme, in the technical scheme, be arranged in the tunnel of underground coal mine first several anchor nodes and
Multiple virtual reference nodes, anchor node can receiving and transmitting signal, when one of anchor node emit signal when, virtual reference node, mesh
It marks object and other anchor nodes receives signal, the quantity of virtual reference node is more than the quantity of anchor node, the reception signal of anchor node
Strength indicator value, object received signal strength indicator value can detect to obtain by the equipment outside mine.Due to anchor node, virtually
Reference mode is the position seat for pre-setting, therefore the position coordinates of anchor node, multiple virtual reference nodes being obtained ahead of time
It is marked with and the received signal strength indicator value of virtual reference node that manual testing obtains, according to connecing for each virtual reference node
Signal strength indication value and position coordinates are received, obtain the finger print information of virtual reference node, and then obtain offline fingerprint database.
Since the variation of tunnel environment can influence the received signal strength indicator value of the anchor node of reception signal, and due to void
The position coordinates of quasi- reference mode are it is known that it is proximity relations, closer, the phase of distance to receive the anchor node of signal and virtual reference node
Like spend it is higher, linear relationship is more apparent, thus by the received signal strength indicator value of each virtual reference node of training with
The proximity mapping model of the received signal strength indicator value of the neighbouring anchor node of each virtual reference node, obtaining can real-time update
Virtual reference node real time fingerprint database, without when tunnel environment changes again by manually virtual into test
The received signal strength indicator value of reference mode, is effectively saved manual labor.
Finally, learn the received signal strength indicator value and position coordinates of virtual reference node in real time fingerprint database
Mapping relations, according to the received signal strength indicator value of object, it may be determined that the position coordinates of object.
In the technical program, the fingerprint matching location algorithm based on non-ranging mode avoids distance measuring type location algorithm and draws
The range measurement error entered, it is suppressed that multipath effect and shadow fading problem；Meanwhile training proximity mapping model real-time update number
According to library, avoids the problem that fingerprint matching algorithm establishes database distortion offline, further improve the real-time positioning of underground coal mine
Precision.
It should be noted that the test mode of the received signal strength indicator value of each virtual reference node is：Each anchor
Node emits signal as transmitting signal node successively, and tester carries ZigBee node in each virtual reference node location
Place receives the received signal strength indicator value that all anchor nodes are sent.
In the above-mentioned technical solutions, further, proximity mapping model training unit, specifically includes：Allocation unit is used for
According to the principle of coordinate Euclidean distance minimum at least one virtual reference node is distributed to each anchor node for receiving signal；First
Arithmetic element, for trained using BP neural network algorithm the received signal strength indicator value of each virtual reference node with it is each
The proximity mapping model of the received signal strength indicator value of the neighbouring anchor node of virtual reference node.
In the technical scheme, an anchor node emits signal, other anchor nodes receive signal, determine the anchor for receiving signal
The method of the neighbouring virtual reference reference mode of node is：According to the principle of coordinate Euclidean distance minimum to each reception signal
Anchor node distributes at least one virtual reference node, i.e., the anchor node of one reception signal corresponds at least one neighbouring virtual ginseng
Examine reference mode, the anchor node of the corresponding neighbouring reception signal of each virtual reference reference mode is different.Training proximity mapping
The method of model is：Using BP neural network algorithm train the received signal strength indicator value of each virtual reference node with it is each
The proximity mapping model of the received signal strength indicator value of the neighbouring anchor node of virtual reference node.By establishing proximity mapping mould
Type, when the received signal strength indicator value of the anchor node of reception signal changes with tunnel environment, virtual reference node connects
Receiving signal strength indication value can also change therewith, i.e. the finger print information real-time update of real-time virtual reference mode, and then obtain reality
When fingerprint database so that the measurement of the position coordinates of object is more accurate.
In the above-mentioned technical solutions, further, position coordinates determination unit specifically includes：Second arithmetic element, is used for
According to PSO-BP neural network algorithms learn real time fingerprint database in virtual reference node received signal strength indicator value with
The mapping relations of position coordinates；Third data processing unit is used for the received signal strength indicator value according to object and mapping
Relationship determines the position coordinates of object.
In the technical scheme, virtual reference node in real time fingerprint database is learnt according to PSO-BP neural network algorithms
Received signal strength indicator value and position coordinates mapping relations, according to the received signal strength indicator value of object with mapping
Relationship, you can the position coordinates for determining object further improve the precision of the position coordinates of object.
The additional aspect and advantage of the present invention will be set forth in part in the description, and will partly become from the following description
Obviously, or practice through the invention is recognized.
Description of the drawings
The above-mentioned and/or additional aspect and advantage of the present invention will become in the description from combination following accompanying drawings to embodiment
Obviously and it is readily appreciated that, wherein：
Fig. 1 shows the flow signal for the method for the object for coal mine down-hole tunnel of embodiment 1 positioned in real time
Figure；
Fig. 2 shows the signals of the flow for the method for the object for coal mine down-hole tunnel of embodiment 2 positioned in real time
Figure；
Fig. 3 shows the flow signal for the method for the object for coal mine down-hole tunnel of embodiment 3 positioned in real time
Figure；
Fig. 4 shows the structural schematic diagram of the real-time positioning system of the object for coal mine down-hole tunnel of embodiment 4；
Fig. 5 shows the position deployment diagram of anchor node and virtual reference node in the tunnel of embodiment 5；
Fig. 6 shows the BP neural network structure chart of embodiment 5；
Fig. 7 shows the updated auditory localization cues figure of the fingerprint database of embodiment 5.
Specific implementation mode
It is below in conjunction with the accompanying drawings and specific real in order to be more clearly understood that aforementioned aspect of the present invention, feature and advantage
Mode is applied the present invention is further described in detail.It should be noted that in the absence of conflict, the implementation of the application
Feature in example and embodiment can be combined with each other.
Many details are elaborated in the following description to facilitate a thorough understanding of the present invention, still, the present invention may be used also
To be implemented different from other modes described here using other, therefore, protection scope of the present invention is not by described below
Specific embodiment limitation.
Embodiment 1：
As shown in Figure 1, the real-time positioning of the object according to an embodiment of the invention for coal mine down-hole tunnel
Method, including：
Step S102, determine in tunnel be arranged it is several can receiving and transmitting signal anchor node position coordinates and multiple virtual ginsengs
The position coordinates of node are examined, and determine the received signal strength indicator value of each virtual reference node；
Step S104, according to the received signal strength indicator value of each virtual reference node and each virtual reference node
Position coordinates form offline fingerprint database；
Step S106, received signal strength indicator value and each virtual reference node of each virtual reference node of training
The proximity mapping model of the received signal strength indicator value of neighbouring anchor node；
Step S108 forms real time fingerprint database according to the offline fingerprint database of proximity mapping model real-time update；
Step S110 learns the received signal strength indicator value of virtual reference node in real time fingerprint database and is sat with position
Target mapping relations determine the position coordinates of object according to the received signal strength indicator value of object.
In the present embodiment, several anchor nodes and multiple virtual reference nodes, anchor are set in the tunnel of underground coal mine first
Node can receiving and transmitting signal, when one of anchor node emits signal, virtual reference node, object and other anchor nodes receive
Signal, the quantity of virtual reference node are more than the quantity of anchor node, and the received signal strength indicator value of anchor node, object connect
Receiving signal strength indication value can be detected to obtain by the equipment outside mine.Since anchor node, virtual reference node are to pre-set, because
The void that the position coordinates of anchor node, the position coordinates of multiple virtual reference nodes and manual testing obtain can be obtained ahead of time in this
The received signal strength indicator value of quasi- reference mode, according to the received signal strength indicator value of each virtual reference node and position
Coordinate obtains the finger print information of virtual reference node, and then obtains offline fingerprint database.
In addition, the received signal strength indicator value by each virtual reference node of training and each virtual reference node
The proximity mapping model of the received signal strength indicator value of neighbouring anchor node, obtain can real-time update virtual reference node reality
When fingerprint database, without when tunnel environment changes again by manually into test virtual reference node reception signal
Strength indicator value is effectively saved manual labor.
Finally, learn the received signal strength indicator value and position coordinates of virtual reference node in real time fingerprint database
Mapping relations, according to the received signal strength indicator value of object, it may be determined that the position coordinates of object.
In the technical program, the fingerprint matching location algorithm based on non-ranging mode avoids distance measuring type location algorithm and draws
The range measurement error entered, it is suppressed that multipath effect and shadow fading problem；Meanwhile training proximity mapping model real-time update number
According to library, avoids the problem that fingerprint matching algorithm establishes database distortion offline, further improve the real-time positioning of underground coal mine
Precision.
Specifically, anchor node is used as transmitting node to emit signal successively, and tester wears a zigbee section in loins
Point receives the received signal strength indicator value that all anchor nodes are sent, Mei Gexu successively at each virtual reference node location
Multi-group data is acquired at quasi- reference mode, the survey that human interference factor is brought in test process is reduced by way of averaging
Try error.By at each virtual reference node received signal strength indicator value and its corresponding position position coordinates be stored in database,
Complete offline fingerprint database structure, the basis as training proximity mapping model.
Structure real time fingerprint database is that the change of tunnel environment in order to prevent makes offline database be distorted, and causes to position
The process of error.In the structure real time fingerprint database stage, the barrier of some larger metal materials is arranged in tunnel at random
Underground coal mine mechanical equipment is imitated, meanwhile, increase tester's walking about in monitoring region, protrudes working environment in tunnel
Variation.Proximity mapping model needs receive the anchor node of signal and virtual reference node is proximity relations, closer, the similarity of distance
Higher, linear relationship is more apparent.Coal mine down-hole tunnel space structure is different from normal indoor structure, and tunnel is usually continuous several
If the gallery of kilometer connects using all virtual reference nodes in tunnel as the neighbor point of the anchor node of a reception signal
It is poor to receive signal strength indication value similitude.Meanwhile the shifting of work noise, staff and mechanical equipment that underground work generates
Dynamic path is random changeable, and the influence to signal transmission is non-uniform in tunnel.If by all void in whole tunnel
Neighbor point of the quasi- reference mode as each anchor node for receiving signal, will produce larger error of fitting, it is therefore desirable in real time
Update proximity mapping model.To avoid the appearance of these problems, the present invention from believing by the reception of each virtual reference node of training
The proximity mapping model of number strength indicator value and the received signal strength indicator value of the neighbouring anchor node of each virtual reference node
Obtain the real time fingerprint database of virtual reference node, without when tunnel environment changes again by manually empty into test
The received signal strength indicator value of quasi- reference mode, is effectively saved manual labor.
Embodiment 2：
As shown in Fig. 2, the real-time positioning of the object according to an embodiment of the invention for coal mine down-hole tunnel
Method, including：
Step S202, determine in tunnel be arranged it is several can receiving and transmitting signal anchor node position coordinates and multiple virtual ginsengs
The position coordinates of node are examined, and determine the received signal strength indicator value of each virtual reference node；
Step S204, according to the received signal strength indicator value of each virtual reference node and each virtual reference node
Position coordinates form offline fingerprint database；
Step S206, it is at least one to each anchor node distribution for receiving signal according to the principle of coordinate Euclidean distance minimum
Virtual reference node；
Step S208, using BP neural network algorithm train the received signal strength indicator value of each virtual reference node with
The proximity mapping model of the received signal strength indicator value of the neighbouring anchor node of each virtual reference node.
Step S210 forms real time fingerprint database according to the offline fingerprint database of proximity mapping model real-time update；
Step S212 learns the received signal strength indicator value of virtual reference node in real time fingerprint database and is sat with position
Target mapping relations determine the position coordinates of object according to the received signal strength indicator value of object.
One anchor node emits signal, other anchor nodes receive signal, determines the neighbouring virtual of the anchor node for receiving signal
Method with reference to reference mode is：It is distributed at least to each anchor node for receiving signal according to the principle of coordinate Euclidean distance minimum
The anchor node of one virtual reference node, i.e., one reception signal corresponds at least one neighbouring virtual reference reference mode, often
The anchor node of the corresponding neighbouring reception signal of a virtual reference reference mode is different.The method of training proximity mapping model is：
The received signal strength indicator value of each virtual reference node and each virtual reference node are trained using BP neural network algorithm
Neighbouring anchor node received signal strength indicator value proximity mapping model.By establishing proximity mapping model, believe receiving
Number the received signal strength indicator value of anchor node when changing with tunnel environment, the received signal strength of virtual reference node refers to
Indicating value can also change therewith, i.e. the finger print information real-time update of real-time virtual reference mode, and then obtain real time fingerprint database,
So that the measurement of the position coordinates of object is more accurate.
Closest principle distributes virtual reference node：
According to the deployed position of the anchor node of the inscribed collection of letters number in tunnel, the principle according to coordinate Euclidean distance minimum is to each
The anchor node for receiving signal distributes multiple virtual reference nodes.Shown in Euclidean distance d such as formula (1).
Wherein (Xr, Yr) be virtual reference node coordinate, (Xa, Ya) be receive signal anchor node coordinate.
Virtual reference node is virtual point known to position, builds proximity mapping model, leads to too small amount of reception signal
Anchor node maps out the received signal strength indicator value of each virtual reference node, realizes the update of fingerprint database.It avoids and works as
After underground communica tion environment changes, needs manually to carry test node in virtual reference node location measurement data, carry out
The update of fingerprint database saves a large amount of human resources.
The mode of segment processing is carried out to tunnel environment, subregion is fitted the environment in tunnel, avoids because of different zones
Database update error caused by environmental change unevenness.
According to apart from nearest principle, each anchor node for receiving signal is assigned with affiliated neighbouring virtual reference node,
Although environment changes in real time, the received signal strength indicator value phase between the adjacent virtual reference node of anchor node of signal is received
It keeps relative stability in time like property mapping relations.By segment processing, artificial real-time update proximity mapping relationship is avoided
Model, receives the anchor node of signal and receives signal strength indication value fingerprint indirectly with neighbouring virtual reference node and be similar to linearly close
System, to further decrease error, will influence virtual reference node with receive signal anchor node similitude distance factor as
Restrictive condition utilizes the stronger BP neural network model construction proximity mapping model of nonlinear fitting ability.
BP neural network is a kind of Multi-layered Feedforward Networks of error back propagation, according to the learning rules of steepest descent method
The mapping relations between input and output are practised, and correct and adjust the weights and threshold value of network by reverse propagated error, until
Network error quadratic sum is minimum.
Embodiment 3：
As shown in figure 3, the real-time positioning of the object according to an embodiment of the invention for coal mine down-hole tunnel
Method, including：
Step S302, determine in tunnel be arranged it is several can receiving and transmitting signal anchor node position coordinates and multiple virtual ginsengs
The position coordinates of node are examined, and determine the received signal strength indicator value of each virtual reference node；
Step S304, according to the received signal strength indicator value of each virtual reference node and each virtual reference node
Position coordinates form offline fingerprint database；
Step S306, it is at least one to each anchor node distribution for receiving signal according to the principle of coordinate Euclidean distance minimum
Virtual reference node；
Step S308, using BP neural network algorithm train the received signal strength indicator value of each virtual reference node with
The proximity mapping model of the received signal strength indicator value of the neighbouring anchor node of each virtual reference node.
Step S310 forms real time fingerprint database according to the offline fingerprint database of proximity mapping model real-time update；
Step S312 learns the reception of virtual reference node in real time fingerprint database according to PSO-BP neural network algorithms
The mapping relations of signal strength indication value and position coordinates；
Step S2314 determines that the position of object is sat according to the received signal strength indicator value and mapping relations of object
Mark.
Learn the received signal strength of virtual reference node in real time fingerprint database according to PSO-BP neural network algorithms
The mapping relations of indicated value and position coordinates, according to the received signal strength indicator value and mapping relations of object, you can determine
The position coordinates of object further improve the precision of the position coordinates of object.
Embodiment 4：
As shown in figure 4, the real-time positioning of the object according to an embodiment of the invention for coal mine down-hole tunnel
System, including：
Determination unit 402, for determining that be arranged in tunnel several can the position coordinates of anchor node of receiving and transmitting signal and more
The position coordinates of a virtual reference node, and determine the received signal strength indicator value of each virtual reference node；
First data preparation unit 404, for according to the received signal strength indicator value of each virtual reference node and every
The position coordinates of a virtual reference node form offline fingerprint database；
Proximity mapping model training unit 406, the received signal strength indicator value for training each virtual reference node
With the proximity mapping model of the received signal strength indicator value of the neighbouring anchor node of each virtual reference node；
Second data preparation unit 408, for forming reality according to the offline fingerprint database of proximity mapping model real-time update
When fingerprint database；
Position coordinates determination unit 410, the reception signal for learning virtual reference node in real time fingerprint database are strong
The mapping relations of degree indicated value and position coordinates determine the position of object according to the received signal strength indicator value of object
Coordinate,
Wherein, when one of anchor node emits signal, virtual reference node, object and other anchor nodes receive letter
Number, the quantity of virtual reference node is more than the quantity of anchor node.
In the present embodiment, several anchor nodes and multiple virtual reference nodes are set in the tunnel of underground coal mine first,
Anchor node can receiving and transmitting signal, when one of anchor node emit signal when, virtual reference node, object and other anchor nodes connect
It collects mail number, the quantity of virtual reference node is more than the quantity of anchor node, the received signal strength indicator value of anchor node, object
Received signal strength indicator value can be detected to obtain by the equipment outside mine.Since anchor node, virtual reference node are to pre-set,
Therefore it can be obtained ahead of time what the position coordinates of anchor node, the position coordinates of multiple virtual reference nodes and manual testing obtained
The received signal strength indicator value of virtual reference node, according to the received signal strength indicator value of each virtual reference node and position
Coordinate is set, the finger print information of virtual reference node is obtained, and then obtains offline fingerprint database.
Since the variation of tunnel environment can influence the received signal strength indicator value of the anchor node of reception signal, and due to void
The position coordinates of quasi- reference mode are it is known that it is proximity relations, closer, the phase of distance to receive the anchor node of signal and virtual reference node
Like spend it is higher, linear relationship is more apparent, thus by the received signal strength indicator value of each virtual reference node of training with
The proximity mapping model of the received signal strength indicator value of the neighbouring anchor node of each virtual reference node, obtaining can real-time update
Virtual reference node real time fingerprint database, without when tunnel environment changes again by manually virtual into test
The received signal strength indicator value of reference mode, is effectively saved manual labor.
Finally, learn the received signal strength indicator value and position coordinates of virtual reference node in real time fingerprint database
Mapping relations, according to the received signal strength indicator value of object, it may be determined that the position coordinates of object.
In the present embodiment, the fingerprint matching location algorithm based on non-ranging mode avoids the introducing of distance measuring type location algorithm
Range measurement error, it is suppressed that multipath effect and shadow fading problem；Meanwhile training proximity mapping model real-time update data
Library avoids the problem that fingerprint matching algorithm offline establishes database distortion, further improves positioning in real time for underground coal mine
Precision.
It should be noted that the test mode of the received signal strength indicator value of each virtual reference node is：Each anchor
Node emits signal as transmitting signal node successively, and tester carries ZigBee node in each virtual reference node location
Place receives the received signal strength indicator value that all anchor nodes are sent.
Further, proximity mapping model training unit 406, specifically includes：Allocation unit 4062, for according to coordinate Europe
The minimum principle of formula distance distributes at least one virtual reference node to each anchor node for receiving signal；First arithmetic element
4064, for trained using BP neural network algorithm the received signal strength indicator value of each virtual reference node with it is each virtual
The proximity mapping model of the received signal strength indicator value of the neighbouring anchor node of reference mode.
One anchor node emits signal, other anchor nodes receive signal, determines the neighbouring virtual of the anchor node for receiving signal
Method with reference to reference mode is：It is distributed at least to each anchor node for receiving signal according to the principle of coordinate Euclidean distance minimum
The anchor node of one virtual reference node, i.e., one reception signal corresponds at least one neighbouring virtual reference reference mode, often
The anchor node of the corresponding neighbouring reception signal of a virtual reference reference mode is different.The method of training proximity mapping model is：
The received signal strength indicator value of each virtual reference node and each virtual reference node are trained using BP neural network algorithm
Neighbouring anchor node received signal strength indicator value proximity mapping model.By establishing proximity mapping model, believe receiving
Number the received signal strength indicator value of anchor node when changing with tunnel environment, the received signal strength of virtual reference node refers to
Indicating value can also change therewith, i.e. the finger print information real-time update of real-time virtual reference mode, and then obtain real time fingerprint database,
So that the measurement of the position coordinates of object is more accurate.
Further, position coordinates determination unit 410, specifically includes：Second arithmetic element 4102, for according to PSO-BP
Neural network algorithm learns the received signal strength indicator value and position coordinates of virtual reference node in real time fingerprint database
Mapping relations；Third data processing unit 4104, for the received signal strength indicator value and mapping relations according to object, really
Set the goal the position coordinates of object.
Learn the received signal strength of virtual reference node in real time fingerprint database according to PSO-BP neural network algorithms
The mapping relations of indicated value and position coordinates, according to the received signal strength indicator value and mapping relations of object, you can determine
The position coordinates of object further improve the precision of the position coordinates of object.
Embodiment 5：
The present invention proposes a kind of location algorithm for being directed to underground coal mine particular surroundings and positioning in real time.Specific location algorithm
For experiment porch using IOT-NODE2530 as hardware platform, low-power consumption, the Z-stack communication protocol stacks of self-organizing are flat for software
Platform.
Tunnel deployment scheme is as shown in figure 5, the underpass (ignoring its height) that the present invention chooses 80*3m is surveyed as experiment
Region is tried, coal mine down-hole tunnel environment is simulated.For the communication overlay area for ensureing larger in channel, 4 anchor node cross-distributions
In channel both sides, determine that position coordinates are A (- 1.5,10), B (1.5,30), C (- 1.5,50), D (1.5,70) respectively.Ignore biography
System location algorithm additionally arranges that the mode for the anchor node for receiving signal, anchor node of the present invention as transmitting and receive signal simultaneously
Anchor node, the fluctuation that the anchor node by receiving signal receives signal strength indication value indirectly reflect personnel in the pit and equipment movement pair
Signal transmission generate block and shadow effect, caused by underground communica tion environment variation.1 is divided in channel every 1m to adopt
Sampling point divides 81 virtual reference nodes as virtual reference node altogether in channel, position coordinates are (0,0)-(0,80).
Step 1. builds the database stage offline.
Offline library of building is the stage for obtaining neighbour's mapping model training data.By tester one is worn in loins
Zigbee nodes receive the received signal strength indicator value of the anchor node of signal at each virtual reference node location,
50 groups of data are acquired at each virtual reference node, human interference factor band in test process is reduced by way of averaging
The test error come.By the received signal strength indicator value and its corresponding position coordinate deposit data at each virtual reference node
Library, the finger print information concrete form as virtual reference node are [RSSI1, RSSI2, RSSI3, RSSI4, (X, Y) ].Complete from
Line database sharing.
The step 2. database update stage.
The database update stage is that the change of tunnel environment in order to prevent makes offline database be distorted, and causes position error
Process.Experiment needs to arrange that the barrier of some larger metal materials imitates underground coal mine machinery and sets at random in tunnel
It is standby, meanwhile, increase tester's walking about in monitoring region, protrudes the variation of working environment in tunnel.
Nearest neighbouring rule distributes virtual reference node.
According to the deployed position of the anchor node of the inscribed collection of letters number in tunnel, the principle according to coordinate Euclidean distance minimum is to each
The anchor node for receiving signal distributes multiple virtual reference nodes.Shown in Euclidean distance d such as formula (1).
Wherein (Xr, Yr) be virtual reference node coordinate, (Xa, Ya) be receive signal anchor node coordinate.
If as shown in figure 5, when receiving the anchor node A transmitting signals of signal, anchor node B, C, D of the inscribed collection of letters number in tunnel connect
Received signal strength indicator value is received, the communication environment of underground different zones at this time is reflected.According to virtual reference node with connect
Virtual reference node is distributed to the anchor nodes of tri- reception signals of B, C, D, often by the anchor node of the collection of letters number apart from nearest principle
It is a receive signal anchor node obtain the neighbour's virtual reference node of itself, by build B, C, D receive signal anchor node with
The mapping model of its corresponding neighbour's virtual reference node updates connecing for the anchor node A of each virtual reference node reception reception signal
Receive signal strength indication value fingerprint.When B is as signal transmitting node, anchor node A, C, D the distribution neighbour for receiving signal virtually joins
Examine node.When C is signal transmitting node, anchor node A, B, D distribution neighbour's virtual reference node of signal is received.D sends out for signal
When penetrating node, anchor node A, B, C distribution neighbour's virtual reference node of signal is received.Virtual reference node is empty known to position
Quasi- point, builds neighbour's mapping model, and logical too small amount of anchor node for receiving signal maps out the reception letter of each virtual reference node
Number strength indicator value, realizes the update of fingerprint database.It avoids after underground communica tion environment changes, needs manually to take
Tape test node carries out the update of fingerprint database in virtual reference node location measurement data, saves a large amount of human resources.
However, when carrying out the distribution of virtual reference node arest neighbors, it may appear that receive the anchor node of signal to its neighbour's virtual reference node
Apart from identical situation.Meanwhile the received signal strength indicator value that the anchor node for receiving signal receives is also identical so that BP nerves
The training data the input phase that network occurs two virtual reference nodes in training neighbour's mapping model is same, exports different feelings
Condition generates larger training error.The present invention dock collect mail number anchor node to virtual reference node distance addition direction because
Element reduces the training error of BP neural network.For example, when receiving the anchor node A of signal as signal transmitting node, letter is received
Number anchor node C receive the received signal strength indicator value that A receives the anchor node of signal, in neighbour's virtual reference node of C
(0,40)-(0,49) is with the anchor node C of (0,51)-(0,60) to reception signal apart from identical.Therefore, the present invention will be located at and receive
The virtual reference node of the anchor node position coordinates negative direction of signal receives the distance of the anchor node of signal to it and is subject to negative sign table
Show direction.Reduce the training error of neighbour's mapping model.
(1) BP neural network trains neighbour's mapping model
Three layers of BP neural network that two input of present invention selection singly exports, input and are connect in real time to receive the anchor node of signal
Receive signal strength indication value and neighbouring virtual reference node distance d；Output is the received signal strength indicator value of reference mode.
BP neural network structure is as shown in Figure 6.
It is O in formula (2)iHidden layer exports formula.YkFor the output formula of output layer.Wherein fj、fkRespectively hidden layer
With the transfer function of output layer；ωijFor the weights of input layer to hidden layer, ωjkFor the weights of hidden layer to output layer.
Formula (3) is the reversed error correction formula of output layer and hidden layer.Wherein dkFor reality output, YkIt is defeated to predict
Go out.
Formula (4) is the formula that error back propagation corrects each layer weights.With real output value and prediction output valve error
Minimum principle preserves BP neural network weight matrix, as receives the adjacent virtual reference node of anchor node of signal
Mapping relations.
Step 3.PSO-BP neural network location models
After database update is completed, received signal strength indicator value fingerprint and the real time fingerprint database matching of object
Positioning, location algorithm select PSO-BP neural network models.Using the stronger None-linear approximation ability of BP neural network, learn number
According to the Nonlinear Mapping relationship of virtual reference node received signal strength indicator value fingerprint and its position coordinates in library, target is inputted
Object received signal strength indicator value finger print data, obtains the position coordinates of object.
(1) it determines neuron number, builds BP neural network topological structure.Positioning stage BP neural network input of the present invention
Neuron is 4, and output neuron is two.
(2) enter the PSO algorithm optimization stages.First, initialize BP neural network weights, including weights number, speed and
Position etc..
wnumber≈Inumber×Hnumber+Onumber×Hnumber (5)
Formula (5) is that weights number initializes formula, InumberTo input neuron number；HnumberFor hidden neuron
Number；OnumberFor output neuron number.
(3) fitness value for calculating each weights, with the training error of BP neural network for its fitness function.Wherein,
Local optimum particle (P in the first generationbest) it is initialization weights, global optimum particle (gbest) it is fitness value in all weights
Minimum weights.
Formula (6) is fitness function, wherein OiFor BP neural network reality output；TiFor BP neural network desired output.
(4) by setting the training error threshold value of BP neural network, as the condition for judging whether iteration terminates.If discontented
Sufficient end condition then updates the position and speed repetitive process (3) of weights, until meeting stopping criterion for iteration.
X (k+1)=x (k)+v (k+1) (7)
V (k+1)=v (k)+c1r1[pbest-w(k)]+c2r2[gbest-w(k)] (8)
Wherein formula (7) and (8) are weights and speed more new formula.Wherein w (k) is that the k moment, (fingerprint database updated
Before) weights, w (k+1) is the weights at k+1 moment (fingerprint database update after).c1, c2For Studying factors, r1, r2Wei [0,1]
Interior uniform random number.
(5) after meeting stopping criterion for iteration, best initial weights matrix is preserved.Meanwhile the BP neural network for calling training to complete,
The received signal strength indicator value fingerprint of object obtains the position coordinates of object as input.
Positioning stage is using PSO-BP neural network models matching positioning destination node.Pass through PSO algorithm optimizations BP nerves
Network improves convergence rate, avoids being absorbed in local optimum.According to experimental result as shown in fig. 7, positioning after finger print data update
Positioning accuracy when precision does not update higher than fingerprint database.Wherein, the positioning accuracy of object when fingerprint database does not update
Positioning accuracy is 2.6310m after being completed for the update of 3.2072m fingerprint databases, and positioning accuracy improves 18.53%, positioning accurate
Degree is according to shown in table 1.
1 object elements of a fix RMSE of table
| Fingerprint database | Fingerprint database does not update | Real time fingerprint database |
| Location structure/m | 3.2072 | 2.6310 |
Above in association with technical scheme of the present invention has been illustrated, technical solution through the invention is used based on non-ranging
The fingerprint matching location algorithm of mode avoids the range measurement error of distance measuring type location algorithm introducing, it is suppressed that multipath effect
With shadow fading problem；Meanwhile training proximity mapping model real-time update database, avoid fingerprint matching algorithm from establishing number offline
The problem of being distorted according to library further improves the precision of underground coal mine positioned in real time.
In the description of this specification, the description of term " one embodiment ", " some embodiments ", " specific embodiment " etc.
Mean that particular features, structures, materials, or characteristics described in conjunction with this embodiment or example are contained at least one reality of the present invention
It applies in example or example.In the present specification, schematic expression of the above terms are not necessarily referring to identical embodiment or reality
Example.Moreover, description particular features, structures, materials, or characteristics can in any one or more of the embodiments or examples with
Suitable mode combines.
The foregoing is only a preferred embodiment of the present invention, is not intended to restrict the invention, for the skill of this field
For art personnel, the invention may be variously modified and varied.All within the spirits and principles of the present invention, any made by repair
Change, equivalent replacement, improvement etc., should all be included in the protection scope of the present invention.
Claims (6)
1. a kind of method of object for coal mine down-hole tunnel positioned in real time, which is characterized in that including：
Determine be arranged in the tunnel it is multiple can receiving and transmitting signal anchor node position coordinates and multiple virtual reference nodes
Position coordinates, and determine the received signal strength indicator value of each virtual reference node；
According to the received signal strength indicator value of each virtual reference node and each virtual reference node
Position coordinates form offline fingerprint database；
The received signal strength indicator value of each virtual reference node of training is neighbouring with each virtual reference node
The proximity mapping model of the received signal strength indicator value of the anchor node；
Real time fingerprint database is formed according to offline fingerprint database described in the proximity mapping model real-time update；
Learn the received signal strength indicator value of virtual reference node in the real time fingerprint database to sit with the position
Target mapping relations determine the position coordinates of the object according to the received signal strength indicator value of the object,
Wherein, when one of them described anchor node emits signal, the virtual reference node, the object and other described in
Anchor node receives the signal, and the quantity of the virtual reference node is more than the quantity of the anchor node.
2. the method for object according to claim 1 positioned in real time, which is characterized in that each void of the training
The received signal strength indicator value fingerprint of quasi- reference mode connects with the neighbouring anchor node of each virtual reference node
The proximity mapping model for receiving signal strength indication value fingerprint, specifically includes：
It is at least one described virtual to each anchor node distribution for receiving signal according to the principle of coordinate Euclidean distance minimum
Reference mode；
Using BP neural network algorithm train the received signal strength indicator value of each virtual reference node with it is each described
The proximity mapping model of the received signal strength indicator value of the neighbouring anchor node of virtual reference node.
3. the method for object according to claim 2 positioned in real time, which is characterized in that the study real-time finger
The mapping relations of the received signal strength indicator value and the position coordinates of virtual reference node in line database, according to institute
The received signal strength indicator value for stating object, determines the position coordinates of the object, specifically includes：
Learn the received signal strength of virtual reference node described in real time fingerprint database according to PSO-BP neural network algorithms
The mapping relations of indicated value and position coordinates；
According to the received signal strength indicator value of the object and the mapping relations, determine that the position of the object is sat
Mark.
4. a kind of real-time positioning system of object for coal mine down-hole tunnel, which is characterized in that including：
Determination unit, for determine be arranged in tunnel it is multiple can receiving and transmitting signal anchor node position coordinates and multiple virtual ginsengs
The position coordinates of node are examined, and determine the received signal strength indicator value of each virtual reference node；
First data preparation unit, for according to each received signal strength indicator value of the virtual reference node and often
The position coordinates of a virtual reference node form offline fingerprint database；
Proximity mapping model training unit, for each virtual reference node of training received signal strength indicator value with it is every
The proximity mapping model of the received signal strength indicator value of the neighbouring anchor node of a virtual reference node；
Second data preparation unit, for forming reality according to offline fingerprint database described in the proximity mapping model real-time update
When fingerprint database；
Position coordinates determination unit, the reception signal for learning virtual reference node in the real time fingerprint database are strong
The mapping relations for spending indicated value and the position coordinates, according to the received signal strength indicator value of the object, described in determination
The position coordinates of object,
Wherein, when one of them described anchor node emits signal, the virtual reference node, the object and other described in
Anchor node receives the signal, and the quantity of the virtual reference node is more than the quantity of the anchor node.
5. real-time positioning system according to claim 4, which is characterized in that the proximity mapping model training unit, tool
Body includes：
Allocation unit, for being distributed at least to each anchor node for receiving signal according to the principle of coordinate Euclidean distance minimum
One virtual reference node；
First arithmetic element, for the received signal strength using each virtual reference node of BP neural network algorithm training
The proximity mapping mould of indicated value and the received signal strength indicator value of the neighbouring anchor node of each virtual reference node
Type.
6. real-time positioning system according to claim 5, which is characterized in that position coordinates determination unit specifically includes：
Second arithmetic element, for learning virtual reference section described in real time fingerprint database according to PSO-BP neural network algorithms
The mapping relations of the received signal strength indicator value and position coordinates of point；
Third data processing unit is used for the received signal strength indicator value according to the object and the mapping relations, really
The position coordinates of the fixed object.
Applications Claiming Priority (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810213820 | 2018-03-15 | ||
| CN2018102138207 | 2018-03-15 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108710104A true CN108710104A (en) | 2018-10-26 |
Family
ID=63867333
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810353671.4A Pending CN108710104A (en) | 2018-03-15 | 2018-04-19 | The method and system of object for coal mine down-hole tunnel positioned in real time |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108710104A (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110530362A (en) * | 2019-09-05 | 2019-12-03 | 北京航空航天大学 | A kind of fireman's indoor orientation method based on single reference mode/inertia combination |
| CN112698268A (en) * | 2020-12-10 | 2021-04-23 | 青岛海信网络科技股份有限公司 | Target equipment positioning method and positioning terminal |
| CN113822419A (en) * | 2021-09-26 | 2021-12-21 | 广东技术师范大学 | Self-supervision graph representation learning operation method based on structural information |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102131290A (en) * | 2011-04-26 | 2011-07-20 | 哈尔滨工业大学 | WLAN (Wireless Local Area Network) indoor neighbourhood matching positioning method based on autocorrelation filtering |
| CN102413564A (en) * | 2011-11-25 | 2012-04-11 | 北京工业大学 | Indoor positioning method based on BP neural network and improved centroid algorithm |
| CN105682224A (en) * | 2016-03-22 | 2016-06-15 | 浙江大学城市学院 | Distributed wireless fingerprint positioning method free of off-line training |
| CN106686547A (en) * | 2016-12-23 | 2017-05-17 | 南京邮电大学 | Indoor fingerprint positioning improvement method based on area division and network topology |
-
2018
- 2018-04-19 CN CN201810353671.4A patent/CN108710104A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102131290A (en) * | 2011-04-26 | 2011-07-20 | 哈尔滨工业大学 | WLAN (Wireless Local Area Network) indoor neighbourhood matching positioning method based on autocorrelation filtering |
| CN102413564A (en) * | 2011-11-25 | 2012-04-11 | 北京工业大学 | Indoor positioning method based on BP neural network and improved centroid algorithm |
| CN105682224A (en) * | 2016-03-22 | 2016-06-15 | 浙江大学城市学院 | Distributed wireless fingerprint positioning method free of off-line training |
| CN106686547A (en) * | 2016-12-23 | 2017-05-17 | 南京邮电大学 | Indoor fingerprint positioning improvement method based on area division and network topology |
Non-Patent Citations (1)
| Title |
|---|
| 崔丽珍等: "基于PSO-BP神经网络的煤矿井下自适应定位算法", 《工矿自动化》 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110530362A (en) * | 2019-09-05 | 2019-12-03 | 北京航空航天大学 | A kind of fireman's indoor orientation method based on single reference mode/inertia combination |
| CN112698268A (en) * | 2020-12-10 | 2021-04-23 | 青岛海信网络科技股份有限公司 | Target equipment positioning method and positioning terminal |
| CN113822419A (en) * | 2021-09-26 | 2021-12-21 | 广东技术师范大学 | Self-supervision graph representation learning operation method based on structural information |
| CN113822419B (en) * | 2021-09-26 | 2023-08-01 | 广东技术师范大学 | Self-supervision graph representation learning operation method based on structural information |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shah et al. | 3D weighted centroid algorithm & RSSI ranging model strategy for node localization in WSN based on smart devices | |
| CN103401922B (en) | Distributed localization apparatus and method based on game method in wireless sensor network | |
| CN101191832B (en) | Wireless sensor network node position finding process based on range measurement | |
| CN104853317B (en) | The structure and update method of fingerprint base in a kind of WiFi indoor positionings | |
| CN106488548B (en) | A kind of determination method and device of indoor multipath error | |
| CN108710104A (en) | The method and system of object for coal mine down-hole tunnel positioned in real time | |
| CN106353725A (en) | RSSI (Received Signal Strength Indication) based indoor moving target location method | |
| CN106912105A (en) | 3-D positioning method based on PSO_BP neutral nets | |
| CN102883428B (en) | Based on the node positioning method of ZigBee wireless sensor network | |
| CN109413575A (en) | A kind of adaptive AP layout method of indoor positioning | |
| CN108459298A (en) | A kind of outdoor positioning method based on LoRa technologies | |
| CN110398320A (en) | A kind of gas leakage detection localization method and system being easy to Continuous optimization | |
| CN102830402A (en) | Target tracking system and method for underwater sensor network | |
| CN105704652A (en) | Method for building and optimizing fingerprint database in WLAN/Bluetooth positioning processes | |
| CN113554849A (en) | Air-ground monitoring slope system and instability risk assessment method thereof | |
| CN106525044A (en) | Person positioning navigation system and method of large naval ship based on ship structure chart | |
| CN109151750A (en) | A kind of LTE indoor positioning floor method of discrimination based on Recognition with Recurrent Neural Network model | |
| CN109041210A (en) | A kind of wireless sensor network locating method | |
| CN107110652A (en) | Handle space characteristics | |
| CN107580294A (en) | A kind of indoor hybrid locating method | |
| CN106341887A (en) | Positioning method and device of indoor robot | |
| JP4491162B2 (en) | Position calculation apparatus and position calculation method | |
| CN110260855A (en) | A kind of indoor pedestrian navigation localization method merging pedestrian's dead reckoning, Geomagnetism Information and indoor map information | |
| CN103249144A (en) | C-type-based wireless sensor network node location method | |
| CN107884746B (en) | Automatic station selection method for multi-station passive time difference positioning technology in wireless gridding |
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 |
Application publication date: 20181026 |
|
| RJ01 | Rejection of invention patent application after publication |