CN109840358A - A kind of track segmentation method based on track time-domain difference - Google Patents
A kind of track segmentation method based on track time-domain difference Download PDFInfo
- Publication number
- CN109840358A CN109840358A CN201910018341.4A CN201910018341A CN109840358A CN 109840358 A CN109840358 A CN 109840358A CN 201910018341 A CN201910018341 A CN 201910018341A CN 109840358 A CN109840358 A CN 109840358A
- Authority
- CN
- China
- Prior art keywords
- track
- difference
- speed
- distance
- ship
- 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.)
- Granted
Links
Landscapes
- Traffic Control Systems (AREA)
- Navigation (AREA)
Abstract
The invention discloses a kind of track segmentation methods based on track time-domain difference, track data can be used by including the following steps: 1) to extract the track points conduct in original track data comprising necessary information, it goes forward side by side row interpolation process of fitting treatment, reduces the influence because of shortage of data or sampling interval difference to sampled point Spatial Density Distribution；2) the voyage data of track points are obtained according to the calculation of longitude & latitude of sampled point, and multistage difference processing is done to track using calculus of finite differences, the equalization point for acquiring loss of voyage Yu speed of a ship or plane standard deviation by method for normalizing on the basis of multistage differential data, determines difference step size；3) average speed of original track data is calculated as fragmentation threshold, realizes track segmentation.The present invention carries out track segmentation using calculus of finite differences, solves the problems, such as that the segmentation divided based on rate-valve value is excessive, can be dynamically determined difference step size, guarantees the accurate division that track is realized under the premise of Modeling for Distance Calculation of Airline loss and stacking area Boundary Distortion are lesser.And the time efficiency of method is more excellent.
Description
Technical field
The present invention relates to track segmentation method fields, specifically, being related specifically to a kind of based on track time-domain difference
Track segmentation method.
Background technique
Track, which refers to, meets figure composed by the dynamic point of certain condition, and the one kind of track as track refers to ship
With the track when navigation such as overwater flight device.Track data is usually the continuous event sequence obtained according to certain frequency acquisition
Column, each track points have recorded mobile object and believe in the movement of the mobile objects such as the time of the point, position and direction, speed
Breath.Excavating achievement for track datas such as the behavioural analysis of marine mobile target, law discovery, feature detections can be maritime traffic
The practical application areas such as safety, biological Migratory Regularity, fishing fishing boat operation provide technical support, it has also become research hotspot.
Track data has the spies such as the period is long, information density is low and uneven, big, the sampling interval disunity of single data volume
Point brings difficult and challenge to the processing, analysis and information excavating of track data.
Track segmentation can distance length according to shipping agency speed or in the unit time track is divided into stop and traveling two
Point.Stop is that ship is stopped in a certain position more than certain time (such as freighter casts anchor after entering a port, handling goods etc.), or certain
It hovers in region, the states such as detour, and this process is also above certain time (such as marine surveys are investigated).Traveling is that ship exists
Moving process state between stacking area, normal speed are higher than resting state.Wherein dwell section can embody the weight such as shipping agency purpose
Action message is wanted, is the main target of gradation study.Reasonable track segmentation algorithm, which can be extracted effectively, stops segment information,
The information density for reducing track redundancy, improving track, reaches compressed data volume, reduces the purpose of processing cost.
The segmentation method based on rate, which is summed up, in the prior art can be mainly divided into following three classes:
(1) based on the segmentation of rate
Krumm et al. is directed to causes the problem for stopping identification mistake to make improvement by GPS positioning error, by defining the time
Threshold value and average speed threshold value and whether the average speed being calculated in the time threshold is lower than threshold speed, to divide track
Stop.Peng Xiangwen et al. carries out trajectory segment by the ship turning angle threshold value and percentage speed variation threshold value of setting, calculates
After to navigation difference and the rate variation of adjacent track point compared with preset value, meet one of them is then with the point
Breakpoint carries out track division.Such methods are suitble to moving direction to change little track, if target detour in stacking area is moved
The dynamic calculating that will affect target average speed is be easy to cause to influence division result, and when mobile target velocity fluctuation is frequent
Track segmentation is excessive, loses the meaning for extracting significant tracks section.
(2) track segmentation based on candidate stacking area.
Boundary and size of the Alvares et al. by pre-defined key area, then judge stop of the track in region
Whether the time reaches threshold value to obtain location information.It is similar, Qi Lingyan et al. definition extract that sub-trajectory to be considered three because
Element: time threshold, distance threshold, terrestrial reference pass through the moving distance after calculating target arrival landmark locations in stacking area, arrival
And time departure, judge whether to meet setting threshold value, to obtain the location information of track.Such method requires to predefine time
Stacking area or interest point information are selected, if user can not obtain or obtain imperfect these significant tracks that will directly lose and stop
Section causes to be segmented mistake.And Sea area boundary is indefinite, without the road network that land is stringent, therefore this method is uncomfortable
It is divided for marine track.
(3) based on the segmentation of cluster.
Du Shenglan et al. is handled a large amount of student tracks of Wuhan University in the school using DBSCAN algorithm, in conjunction with campus
The geography information such as supermarket, teaching building extract dwell characteristics, for analyzing the special group behavior of military university student.Such method can be because
The problem of causing spatially adjacent but practical residence time longer track to be divided into one section by mistake for uneven sampling density, and
The adjacent track section in the same area space not necessarily Time Continuous also has by the possibility of dislocation segmentation.Quan Yucheng et al. is proposed
With the time be cluster core away from DBSCAN innovatory algorithm, which solves the problems, such as to be mentioned above to a certain extent, but
It is not described determination method of the core away from size, and the time complexity of this method is higher.
Summary of the invention
It is an object of the invention to aiming at the shortcomings in the prior art, propose a kind of track based on track time-domain difference point
Phase method (TDD, Time Domain Difference) calculates the average boat in longer period by increasing difference step size
Speed is convenient for track segmentation, and dynamically solve difference step size in speed of a ship or plane difference mistake so that the change of speed of a ship or plane state is more obvious
Unstructured problem in journey.Finally on the basis of the difference speed of a ship or plane by compare in the difference and adjacent stacking area travel distance with
The difference of average rate travel distance judges whether two stacking areas of merging, realizes the accurate division of same stacking area.
Technical problem solved by the invention can be realized using following technical scheme:
A kind of track segmentation method based on track time-domain difference, includes the following steps:
1) track points in original track data comprising necessary information are extracted as track data can be used, row interpolation of going forward side by side is quasi-
Conjunction processing, reduces the influence because of shortage of data or sampling interval difference to sampled point Spatial Density Distribution；
2) the voyage data of track points are obtained according to the calculation of longitude & latitude of sampled point, and it is multistage to use calculus of finite differences to do track
Difference processing acquires the balance of loss of voyage Yu the speed of a ship or plane standard deviation on the basis of multistage differential data by method for normalizing
Point, determines difference step size；
3) average speed of original track data is finally calculated as fragmentation threshold, realizes track segmentation.
Further, the calculation formula of the voyage data of track points is as follows in the step 2):
Dis=R × (arccos (sin θlat1sinθlat2+cosθlat1cosθlat2cos(θlon1-θlon2))) (1)
Wherein dis is the distance between track points, and R is earth radius, the current track points that add up to distance value between starting point
Distance data D={ d can be obtained1,d2,d3...dn, diFor the distance of current track points；
Using present sample time t as minimum differential step-length, change step value h, with calculus of finite differences calculate track not commensurate when
Interior velocity amplitude, obtains differential data Vd={ v1,v2,v3...vn, viFor the difference speed of a ship or plane, formula is as follows:
vi=(di+1-di-1)/2h (2)
hi=n × t (3)
The condition that n should meet in formula (3) are as follows: n ∈ N, n > 0.
Further, the calculating process of the multistage difference processing in the step 2) is as follows:
Using original track as track points, the orderly point set constituted by sample time order is denoted as S={ T, V, θlon,
θlat}；Wherein T is the sampling time, and V is momentary rate, θlonFor sampled point longitude, θlatFor sampled point latitude；Due to the sampling time
Interval between point T is smaller, and momentary rate V is usually fluctuated frequently during ship running, and the speed of a ship or plane value of single sampled point is not
Current speed of a ship or plane state can be represented, it is therefore desirable to redefine suitable time interval to calculate current travel speed；It is described
Time interval is referred to as difference step size.
Further, the determination method of the difference step size is as follows:
Speed of a ship or plane standard deviation and distance loss to same track difference difference step size ask friendship after handling using method for normalizing
Point, and an equalization point is being sought in speed of a ship or plane standard deviation and distance loss bring influence, obtain relatively reasonable difference step
Long value；The standard deviation formula of the speed of a ship or plane are as follows:
Wherein veqlFor the speed average of original track；
Distance loss are as follows:
Sdis=Sp-Sn (5)
Wherein SpFor current distance value, SnFor actual distance value, SdisFor distance penalty values；
By the speed of a ship or plane standard deviation of multistage difference step size and distance penalty values normalized, i.e.,
Sta(hi)/Stamax=Sdis(hi)/Sdis max (6)
Wherein, StamaxFor speed of a ship or plane maximum standard deviation, Sdis maxFor track maximum distance penalty values；When formula (6) are set up,
H can be acquirediAs it is suitable for the difference step size value that current track divides.
Further, the track segmentation method further includes the merging method to stacking area, and formula is as follows:
Formula is as follows:
disi=di-di-1
disinter=di+1-di
diffi=disi-vequlTi
diffinter=disinter-vequlTinter
Wherein diFor track points voyage value, disiFor stacking area distance value, disinterFor the traveling between adjacent stacking area
Distance value, vequlFor average speed, TiFor residence time, TinterFor the running time between adjacent stacking area, diffiTo stop
The difference of area's travel distance and average speed travel distance, diffinterTravel distance and average speed between adjacent stacking area
Spend the difference of travel distance；Work as diffinterThe diff of respectively less than two adjacent stacking areasiWhen, merge two stacking areas.
Compared with prior art, the beneficial effects of the present invention are:
1. change the otherness that frequent phenomenon uses calculus of finite differences to improve speed of a ship or plane resting state and driving status for the speed of a ship or plane,
Convenient for the extraction of stay segment track.
2. being dynamically determined difference step size using method for normalizing balance track distance loss and speed of a ship or plane standard deviation and realizing track
Difference.
3. judging the distance difference of adjacent track section Yu current track section, merges track section and improve the accurate of track division
Property.
Detailed description of the invention
Fig. 1 is the flow chart of the track segmentation method of the present invention based on track time-domain difference.
Fig. 2 is multistage track difference result figure of the present invention.
Fig. 3 is distance of the present invention loss and the variation diagram that speed of a ship or plane standard deviation increases with difference step size.
Fig. 4 is the distance loss in data analysis and the variation diagram after the normalization of speed of a ship or plane standard deviation with difference step size.
The track division result figure for two methods that Fig. 5 is difference step size when being 0.8.
Fig. 6 is two methods run-time efficiency comparison diagram.
Specific embodiment
To be easy to understand the technical means, the creative features, the aims and the efficiencies achieved by the present invention, below with reference to
Specific embodiment, the present invention is further explained.
Referring to Fig. 1, the track segmentation method of the present invention based on track time-domain difference extracts original track number first
Track points in comprising necessary information are as can use track data, row interpolation process of fitting treatment of going forward side by side, reduce because of shortage of data or
Influence of the sampling interval difference to sampled point Spatial Density Distribution, then obtains track points according to the calculation of longitude & latitude of sampled point
Voyage data, and multistage difference processing is done to track using calculus of finite differences, pass through normalization side on the basis of multistage differential data
Method acquires the equalization point of loss of voyage Yu speed of a ship or plane standard deviation, determines suitable difference step size.Finally calculate original track data
Average speed realizes track segmentation as fragmentation threshold.
Multistage Difference Calculation: the orderly point set that original track is made of track points by sample time order is denoted as S=
{T,V,θlon,θlat}.Wherein T is the sampling time, and V is momentary rate, θlonFor sampled point longitude, θlatFor sampled point latitude.By
Interval between sampling time point T is smaller, and momentary rate V is usually fluctuated frequently during ship running, single sampled point
Speed of a ship or plane value can not represent current speed of a ship or plane state, it is therefore desirable to redefine suitable time interval to calculate current traveling
Speed.For corresponding segmentation method hereafter, which is referred to as difference step size by this method.
What track data recorded is the track points information of series of discrete, in order to calculate the average speed in continuous difference step size
Degree, this method do multistage difference processing to original track data using difference method.Calculus of finite differences (difference methods,
Abbreviation DM), it is a kind of differential equation numerical method for discrete data, is leading come Approximation Discrete variable by finite difference
Numerical value.The interval of discrete data is known as step-length, and same data can acquire different degrees of approximation with asynchronous long progress difference and lead
Number.Original track data does not include specific path length generally, but can obtain difference by the calculation of longitude & latitude of track points
The operating range of ship in step-length.Difference is carried out to track data using the time as independent variable on the basis of distance data, is passed through
The velocity amplitude in the continuous unit time can be obtained by defining difference step size, then realize track segmentation by defining threshold speed.
By the enforcement distance between calculation of longitude & latitude track points, formula is as follows:
Dis=R × (arccos (sin θlat1sinθlat2+cosθlat1cosθlat2cos(θlon1-θlon2))) (1)
Wherein dis is the distance between track points, and R is earth radius, the current track points that add up to distance value between starting point
Distance data D={ d can be obtained1,d2,d3...dn, diFor the distance of current track points.
Using present sample time t as minimum differential step-length, change step value h, with calculus of finite differences calculate track not commensurate when
Interior velocity amplitude, obtains differential data Vd={ v1,v2,v3...vn, viFor the difference speed of a ship or plane, formula is as follows:
vi=(di+1-di-1)/2h (2)
hi=n × t (3)
The condition that n should meet in formula (3) are as follows: n ∈ N, n > 0.
Multistage track difference result shown in Figure 2 is track initial data, difference step respectively from top to bottom in Fig. 2
A length of 0.2 (unit/day), 0.4 (unit/day), 0.8 (unit/day), the difference result of 4 (units/day), crude sampling time
Interval about 5 minutes, difference step size increases about 10,20,40,200 times compared with sampling time interval respectively.It can be from Fig. 2
Out in the case where speed of a ship or plane variation general trend is constant, crenellated phenomena gentle, and with the increase of difference step size, effect
It gradually increases.The difference speed of a ship or plane after increasing difference step size, more original speed of a ship or plane data are apparent in state change, are more suitable for carrying out
Track segmentation.Therefore calculus of finite differences can effectively reduce influence caused by sample frequency unevenness, speed of a ship or plane when highlighting different driving status
Otherness.
The determination of dynamic difference step-length: calculus of finite differences calculates the average boat in longer period by increasing difference step size
Speed is convenient for track segmentation so that the change of speed of a ship or plane state is more obvious, but when difference step size is excessive also bring along data distortion compared with
Big problem.It causes the characteristic points such as most speed of a ship or plane peak values and valley to lack since difference step size is excessive, may cause speed of a ship or plane state
Difference reduces, and adjacent sectional is merged and is segmented very few problem, as shown in the e in Fig. 2.And the freedom of marine traveling, so that
Different Flight has different stop distribution and feature, and difference step size when track being caused to divide is also different.Therefore dynamic
Determination difference step size to track it is accurate division have important influence.
Compare different difference step sizes difference result discovery the zigzag speed of a ship or plane amplitude it is excessive be to cause to be segmented excessive master
Reason is wanted, and the severity of crenellated phenomena can be measured by the standard deviation of the speed of a ship or plane, and with the increase of difference step size, boat
Fast standard deviation is gradually reduced.Meanwhile being to lose along with distance using the calculus of finite differences difference speed of a ship or plane obtained, and with difference
The increase of step-length, distance loss also increase with it.
Fig. 3 is the distance loss of track and the variation diagram that speed of a ship or plane standard deviation increases with difference step size, it can be seen that distance damage
Mistake increases with the increase of difference step size, is positively correlated with difference step size, and speed of a ship or plane standard deviation is then on the contrary, for negative correlation.
Speed of a ship or plane standard deviation and distance loss to same track difference difference step size ask friendship after handling using method for normalizing
Point, and an equalization point is being sought in speed of a ship or plane standard deviation and distance loss bring influence, obtain relatively reasonable difference step
Long value；The standard deviation formula of the speed of a ship or plane are as follows:
Wherein veqlFor the speed average of original track；
Distance loss are as follows:
Sdis=Sp-Sn (5)
Wherein SpFor current distance value, SnFor actual distance value, SdisFor distance penalty values；
By the speed of a ship or plane standard deviation of multistage difference step size and distance penalty values normalized, i.e.,
Sta(hi)/Stamax=Sdis(hi)/Sdis max (6)
Wherein, StamaxFor speed of a ship or plane maximum standard deviation, Sdis maxFor track maximum distance penalty values；When formula (6) are set up,
H can be acquirediAs it is suitable for the difference step size value that current track divides.
Consolidation strategy: the definition of track stacking area includes that the low speed of mobile target in certain area is hovered, when a certain stop
When the residence time in area is longer, there is the fluctuation more than speed of a ship or plane threshold value in the speed of a ship or plane, this is because moving in same stacking area
Caused by target is hovered between multiple destinations, as research ship is investigating multiple investigation points investigation in area, pleasure boat at scenic spot
Sight spot migration etc..The fluctuating of this speed of a ship or plane will affect final track division result, that is, a stacking area is divided into
The problem of multiple stacking areas.
For this problem, research finds crossing in division stacking area, travel distance and average rate between adjacent stacking area
The difference of travel distance is smaller, and time interval is not grown.It can be by comparing in the difference and adjacent stacking area travel distance and
The difference of fast travel distance judges whether two stacking areas of merging, realizes the accurate division of same stacking area.Formula is as follows:
disi=di-di-1
disinter=di+1-di
diffi=disi-veoulTi
diffinter=disinter-vequlTinter
Wherein diFor track points voyage value, disiFor stacking area distance value, disinterFor the traveling between adjacent stacking area
Distance value, vequlFor average speed, TiFor residence time, TinterFor the running time between adjacent stacking area, diffiTo stop
The difference of area's travel distance and average speed travel distance, diffinterTravel distance and average speed between adjacent stacking area
Spend the difference of travel distance；Work as diffinterThe diff of respectively less than two adjacent stacking areasiWhen, merge two stacking areas.
Analysis of experimental data
Data source is the track data of the 29th antarctic investigation in Chinese Polar central database.This time research ship is from upper
Sea is set out, and the South Pole is gone to carry out scientific investigation after stopping to stop at Australia, mainly investigates area by the gulf Pu Lizi, Ross Sea etc.
Domain finally returns that Shanghai.This navigation starts running that the time is on October 30th, 2012, and the end time is April 1 in 2013
Day, amount to about 153 days, in about 2.7 Wan-hai of total voyage.Track data includes track points ID number, data sampling time, track points sky
Between the information such as coordinate, the speed of a ship or plane, course.29th time scientific investigation track data in the South Pole is 191071 articles total, but most of data information
Missing is serious, it is contemplated that specific sampling time in this method research method, the speed of a ship or plane and spatial coordinated information importance, this method
The track data including this three information has been extracted, 6991 datas are obtained after screening.
Calculate difference step size
TDD division methods determine that suitable difference walks by calculating the equalization point of the loss of track distance and speed of a ship or plane standard deviation
It is long, difference processing is then done to track using calculus of finite differences, obtains the clearer difference speed of a ship or plane data of speed of a ship or plane state change.Calculate boat
Mark average speed realizes track segmentation as threshold value is divided.
Fig. 4 is the 29th antarctic investigation track distance loss and the variation after the normalization of speed of a ship or plane standard deviation with difference step size
Situation, two curves are respectively speed of a ship or plane standard deviation change curve and distance loss change curve, and intersection point is 0.8 day.D schemes in Fig. 2
For the division result of 0.8 difference step size, without obvious distortion, high low-velocity path section difference is more apparent, meets the pre- of further division
Phase.
Comparative analysis
(1) subsection efect compares
The comparison algorithm that this method is selected is the improvement DBSCAN clustering method based on time-domain.DBSCAN is one
It is different to be generally only applicable to the clustering method of convex sample set from K-means, BIRCH etc. for kind of Classic Clustering Algorithms based on density,
DBSCAN is not only suitable for convex sample set, is also applied for recessed sample set, is one of most common clustering method.After defining time gap
Whether (i.e. cluster core away from), traversal track data can get the core point for meeting density conditions, finally judge core point in phase
Realize that the merging of core point clusters in adjacent core neighborhood of a point, cluster result is the stacking area of track points composition.
The unit calculated value of two methods be all time value be difference step size and core respectively away from.Due in track segmentation
The optimal core of DBSCAN method is away from temporarily ununified standard is calculated, and this method is by the cluster core of DBSCAN method away from setting
It is consistent with the optimal difference step size of this method.It calculates under track average speed, core is away from corresponding maximum search distance, with this
Distance is the radius of neighbourhood, and calculating the track points in all core points and its neighborhood, whether to meet density reachable, traverses track points,
Merge the division result that the connected core point of density obtains improving DBSCAN clustering method, navigates with the difference of corresponding difference step size
Division result of the mark under identical track average speed threshold value compares, and comparing result is referring to Fig. 5.
The track division result figure for two methods that upper figure is difference step size when being 0.8 (unit: day), the height of shaded block in figure
Degree is speed of a ship or plane threshold value, and the track in wire frame is the track section i.e. stacking area for meeting division condition.Wherein difference division result figure is
The division result of this method method, and clustering result figure is to improve the division result of DBSCAN algorithm.
The travel route of 29th scientific investigation track is Shanghai, Guangzhou, Fremantle (Australia), middle mountain station, Ross
Sea, Hobart (Australia), middle mountain station, Fremantle (Australia), Shanghai.Practical stacking area in track has 7,
Following table is each stacking area at different difference step sizes (core away from), and two methods compare the division number of segment of same stacking area.
Table 1
From table 1 it follows that increase of the two methods with difference step size, each stacking area is divided number of segment and gradually subtracts
Few, but in the case of same difference step size (core away from), this method proposes number of segment that method is divided closer to true stacking area
Number divides effect and is better than clustering method.
(2) algorithm time efficiency compares
The programming tool of this method method is MATLAB, and version R2016a, hardware environment is CPU:Intel (R)
Core (TM) i5-6500, memory: 8G, operating system: 64 Win10.The operation time of two methods is as shown in Figure 5.
Fig. 6 is two methods run-time efficiency comparison diagram.With the required increase for dividing track points quantity, two methods
Runing time be stepped up.The substantially linear growth trend of two methods time loss, but the growth rate of clustering method is about
It is 3 times of difference method, and the runing time of clustering method increased when track points are 2000 to 3000, restore later
Former growth rate, this is because clustering method calculate meet the track points in stacking area when can repeatedly loop through, encounter compared with
Runing time can increased when the stacking area of long period.Being compared by efficiency of algorithm can obtain: the time of TDD division methods
Consumption and growth rate are respectively less than clustering method, and time efficiency is substantially better than clustering method, and more stable.
The above shows and describes the basic principles and main features of the present invention and the advantages of the present invention.The technology of the industry
Personnel are it should be appreciated that the present invention is not limited to the above embodiments, and the above embodiments and description only describe this
The principle of invention, without departing from the spirit and scope of the present invention, various changes and improvements may be made to the invention, these changes
Change and improvement all fall within the protetion scope of the claimed invention.The claimed scope of the invention by appended claims and its
Equivalent thereof.
Claims (5)
1. a kind of track segmentation method based on track time-domain difference, which comprises the steps of:
1) track points in original track data comprising necessary information are extracted as can use track data, at row interpolation fitting of going forward side by side
Reason reduces the influence because of shortage of data or sampling interval difference to sampled point Spatial Density Distribution；
2) the voyage data of track points are obtained according to the calculation of longitude & latitude of sampled point, and multistage difference is done to track using calculus of finite differences
Processing, acquires the equalization point of loss of voyage Yu speed of a ship or plane standard deviation, really on the basis of multistage differential data by method for normalizing
Determine difference step size；
3) average speed of original track data is calculated as fragmentation threshold, realizes track segmentation.
2. according to the claim track segmentation method according to claim 1 based on track time-domain difference, feature exists
In the calculation formula of the voyage data of track points is as follows in the step 2):
Dis=R × (arccos (sin θlat1sinθlat2+cosθlat1cosθlat2cos(θlon1-θlon2))) (1)
Wherein dis is the distance between track points, and R is earth radius, and the current track points that add up can be obtained to distance value between starting point
To distance data D={ d1,d2,d3...dn, diFor the distance of current track points；
Using present sample time t as minimum differential step-length, change step value h, is calculated in the track different unit time with calculus of finite differences
Velocity amplitude, obtain differential data Vd={ v1,v2,v3...vn, viFor the difference speed of a ship or plane, formula is as follows:
vi=(di+1-di-1)/2h (2)
hi=n × t (3)
The condition that n should meet in formula (3) are as follows: n ∈ N, n > 0.
3. the track segmentation method according to claim 2 based on track time-domain difference, which is characterized in that the step 2)
In multistage difference processing calculating process it is as follows:
Using original track as track points, the orderly point set constituted by sample time order is denoted as S={ T, V, θlon,θlat}；
Wherein T is the sampling time, and V is momentary rate, θlonFor sampled point longitude, θlatFor sampled point latitude；Due to sampling time point T it
Between interval it is smaller, momentary rate V is usually fluctuated frequently during ship running, and the speed of a ship or plane value of single sampled point can not represent
Current speed of a ship or plane state, it is therefore desirable to redefine suitable time interval to calculate current travel speed；Between the time
Every being referred to as difference step size.
4. the track segmentation method according to claim 3 based on track time-domain difference, which is characterized in that the difference step
Long determination method is as follows:
Speed of a ship or plane standard deviation and distance loss to same track difference difference step size are found intersection after being handled using method for normalizing, and
An equalization point is being sought in speed of a ship or plane standard deviation and distance loss bring influence, is obtaining relatively reasonable difference step size value；
The standard deviation formula of the speed of a ship or plane are as follows:
Wherein veqlFor the speed average of original track；
Distance loss are as follows:
Sdis=Sp-Sn (5)
Wherein SpFor current distance value, SnFor actual distance value, SdisFor distance penalty values；
By the speed of a ship or plane standard deviation of multistage difference step size and distance penalty values normalized, i.e.,
Sta(hi)/Stamax=Sdis(hi)/Sdismax (6)
Wherein, StamaxFor speed of a ship or plane maximum standard deviation, SdismaxFor track maximum distance penalty values；When formula (6) are set up, can acquire
hiAs it is suitable for the difference step size value that current track divides.
5. the track segmentation method according to claim 1 based on track time-domain difference, which is characterized in that the track point
Phase method further includes the merging method to stacking area, and formula is as follows:
Formula is as follows:
disi=di-di-1
disinter=di+1-di
diffi=disi-vequlTi
diffinter=disinter-vequlTinter
Wherein diFor track points voyage value, disiFor stacking area distance value, disinterFor the travel distance between adjacent stacking area
Value, vequlFor average speed, TiFor residence time, TinterFor the running time between adjacent stacking area, diffiFor stacking area row
Sail the difference of distance Yu average speed travel distance, diffinterTravel distance and average speed row between adjacent stacking area
Sail the difference of distance；Work as diffinterThe diff of respectively less than two adjacent stacking areasiWhen, merge two stacking areas.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910018341.4A CN109840358B (en) | 2019-01-09 | 2019-01-09 | Track time domain difference-based track segmentation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910018341.4A CN109840358B (en) | 2019-01-09 | 2019-01-09 | Track time domain difference-based track segmentation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109840358A true CN109840358A (en) | 2019-06-04 |
| CN109840358B CN109840358B (en) | 2023-04-07 |
Family
ID=66883608
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910018341.4A Active CN109840358B (en) | 2019-01-09 | 2019-01-09 | Track time domain difference-based track segmentation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109840358B (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110309383A (en) * | 2019-06-17 | 2019-10-08 | 武汉科技大学 | Ship trajectory clustering analysis method based on improved DBSCAN algorithm |
| CN110618290A (en) * | 2019-09-09 | 2019-12-27 | 中国船舶重工集团公司第七0七研究所九江分部 | Speed information fusion method |
| CN110921446A (en) * | 2019-12-10 | 2020-03-27 | 猫岐智能科技（上海）有限公司 | Equipment attribute acquisition system |
| CN110941671A (en) * | 2019-11-21 | 2020-03-31 | 中国人民解放军海军航空大学 | Offshore target track segmentation and description method, electronic device and storage medium |
| CN112164247A (en) * | 2020-09-03 | 2021-01-01 | 重庆大学 | Ship route prediction method based on ship track clustering |
| CN115294802A (en) * | 2022-07-25 | 2022-11-04 | 中远海运科技股份有限公司 | AIS data-based ship navigation state intelligent identification method and system |
| WO2024057997A1 (en) * | 2022-09-15 | 2024-03-21 | 株式会社メタシステム研究所 | Fishing ship activity estimation device and fishing ship activity estimation program |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017143631A1 (en) * | 2016-02-22 | 2017-08-31 | 江苏大学 | Automatic navigation route tracking method for aquaculture vessel |
| CN109000645A (en) * | 2018-04-26 | 2018-12-14 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Complex environment target classics track extracting method |
-
2019
- 2019-01-09 CN CN201910018341.4A patent/CN109840358B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2017143631A1 (en) * | 2016-02-22 | 2017-08-31 | 江苏大学 | Automatic navigation route tracking method for aquaculture vessel |
| CN109000645A (en) * | 2018-04-26 | 2018-12-14 | 西南电子技术研究所（中国电子科技集团公司第十研究所） | Complex environment target classics track extracting method |
Non-Patent Citations (1)
| Title |
|---|
| 王莉莉等: "基于LOFC时间窗分割算法的航迹聚类研究", 《南京航空航天大学学报》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110309383A (en) * | 2019-06-17 | 2019-10-08 | 武汉科技大学 | Ship trajectory clustering analysis method based on improved DBSCAN algorithm |
| CN110309383B (en) * | 2019-06-17 | 2021-07-13 | 武汉科技大学 | Ship track clustering analysis method based on improved DBSCAN algorithm |
| CN110618290A (en) * | 2019-09-09 | 2019-12-27 | 中国船舶重工集团公司第七0七研究所九江分部 | Speed information fusion method |
| CN110618290B (en) * | 2019-09-09 | 2021-09-28 | 中国船舶重工集团公司第七0七研究所九江分部 | Speed information fusion method |
| CN110941671A (en) * | 2019-11-21 | 2020-03-31 | 中国人民解放军海军航空大学 | Offshore target track segmentation and description method, electronic device and storage medium |
| CN110941671B (en) * | 2019-11-21 | 2022-10-25 | 中国人民解放军海军航空大学 | Offshore target track segmentation and description method, electronic device and storage medium |
| CN110921446A (en) * | 2019-12-10 | 2020-03-27 | 猫岐智能科技（上海）有限公司 | Equipment attribute acquisition system |
| CN112164247A (en) * | 2020-09-03 | 2021-01-01 | 重庆大学 | Ship route prediction method based on ship track clustering |
| CN115294802A (en) * | 2022-07-25 | 2022-11-04 | 中远海运科技股份有限公司 | AIS data-based ship navigation state intelligent identification method and system |
| WO2024057997A1 (en) * | 2022-09-15 | 2024-03-21 | 株式会社メタシステム研究所 | Fishing ship activity estimation device and fishing ship activity estimation program |
| JP7466129B2 (en) | 2022-09-15 | 2024-04-12 | 株式会社メタシステム研究所 | Fishing vessel activity estimation device and fishing vessel activity estimation program |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109840358B (en) | 2023-04-07 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109840358A (en) | A kind of track segmentation method based on track time-domain difference | |
| WO2022252398A1 (en) | Ship trajectory feature point extraction-based spatio-temporal dp method | |
| CN108802776B (en) | Bus GPS (global positioning system) deviation rectifying method based on abnormal point elimination and track compression algorithm | |
| CN104680187B (en) | Naval vessel track Hierarchical Cluster Analysis based on buffering area similarity measurement | |
| CN103531043B (en) | Based on the boats and ships point-to-point hours underway evaluation method of course line coupling | |
| CN107504972A (en) | A kind of aircraft's flight track method and device for planning based on dove group's algorithm | |
| CN106781479B (en) | A method of highway operating status is obtained based on mobile phone signaling data in real time | |
| CN104851286B (en) | A kind of method of utilization public transit vehicle gps data dynamic evaluation road traffic state | |
| CN109059944A (en) | Motion planning method based on driving habit study | |
| CN106710313A (en) | Method and system for ship in bridge area to actively avoid collision based on laser three-dimensional imaging technique | |
| CN107635204A (en) | A kind of indoor fusion and positioning method and device of motor behavior auxiliary, storage medium | |
| CN111341103B (en) | Lane information extraction method, device, equipment and storage medium | |
| CN106546244A (en) | It is a kind of to hinder the method that boat thing determines ship's navigation local optimum path for convex | |
| CN108171993A (en) | A kind of vehicle on highway speed calculation method based on mobile phone signaling big data | |
| CN109949593A (en) | A kind of traffic lights recognition methods and system based on crossing priori knowledge | |
| CN113032378A (en) | Ship behavior pattern mining method based on clustering algorithm and pattern mining | |
| CN106887012A (en) | A kind of quick self-adapted multiscale target tracking based on circular matrix | |
| CN110084491A (en) | Based on the optimal air route blockage percentage appraisal procedure for passing through path under the conditions of convection weather | |
| CN109520499A (en) | Region isochronal method in real time is realized based on vehicle GPS track data | |
| CN110444046A (en) | A kind of restricted waters non conflicting can meet ship cluster Situation analysis method | |
| Lin et al. | Noise filtering, trajectory compression and trajectory segmentation on GPS data | |
| CN110400113A (en) | Part real-time scheduling method and system are pulled in a kind of logistics | |
| CN113280818A (en) | Ship route automatic planning method based on adaptive triangulation network | |
| CN110081890A (en) | A kind of dynamic K arest neighbors map-matching method of combination depth network | |
| CN109000682A (en) | The method of aerial coarse alignment based on rectilinear path |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |