CN106803361A - A kind of navigation method of control based on Rolling Planning strategy - Google Patents
A kind of navigation method of control based on Rolling Planning strategy Download PDFInfo
- Publication number
- CN106803361A CN106803361A CN201710141983.4A CN201710141983A CN106803361A CN 106803361 A CN106803361 A CN 106803361A CN 201710141983 A CN201710141983 A CN 201710141983A CN 106803361 A CN106803361 A CN 106803361A
- Authority
- CN
- China
- Prior art keywords
- ship
- omega
- delta
- wind field
- marine site
- 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
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G3/00—Traffic control systems for marine craft
-
- G—PHYSICS
- G08—SIGNALLING
- G08G—TRAFFIC CONTROL SYSTEMS
- G08G3/00—Traffic control systems for marine craft
- G08G3/02—Anti-collision systems
Abstract
The present invention relates to a kind of navigation method of control based on Rolling Planning strategy, including following several steps, the real-time and historical position information of ship is obtained by sea radar first;Then in each sampling instant, the track for speculating ship in future time period is rolled according to the real-time and historical position information of ship;The current running status of ship and historical position observation sequence are based on again, obtain the numerical value of marine site wind field variable;The safety regulation collection that the ship of the running status and setting that are based on each ship again need to be met when being run in marine site, dynamic behaviour implementing monitoring to ship and for maritime traffic control centre provides timely warning information;When warning information occurs, on the premise of ship physical property and marine site traffic rules is met, by setting optimizing index function and incorporating wind field variable value, Rolling Planning is carried out to ship collision avoidance track using Adaptive Control Theory method, and program results is transferred to each ship to perform.
Description
The application is Application No.:201410849264.4, invention and created name is《A kind of navigation traffic control method》,
The applying date is:The divisional application of the application for a patent for invention on December 30 in 2014.
Technical field
The present invention relates to a kind of marine site traffic control method, more particularly to a kind of marine site traffic based on Rolling Planning strategy
Method of control.
Background technology
With the fast development of global shipping business, the traffic in the busy marine site in part is further crowded.It is close in vessel traffic flow
Collection complexity marine site, the regulation model allocated at artificial interval has still been combined for the collision scenario between ship not using sail plan
Adapt to the fast development of shipping business.To ensure the personal distance between ship, implement effective conflict allotment and just handed over as marine site
The emphasis of siphunculus system work.Ship conflict Resolution is a key technology in navigational field, safely and efficiently frees scheme pair
It is significant in increasing marine site ship flow and ensuring that sea-freight is safe.
In order to improve the efficiency of navigation of ship, marine radar automatic plotter has been widely applied to ship monitor at present
In collision prevention, the equipment is by extracting ship relevant information for the judgement of collision scenario between ship provides reference frame.Although this
Kind equipment greatly reduces the load of manual monitoring, but it does not have the automatic conflict Resolution function of ship.For ship conflict
Problem is freed, current processing mode mainly includes geometric deterministic algorithm and the major class scheme of Heuristic Intelligent Algorithm two, phase
Close literature research and be concentrated mainly on conflict avoiding planning algorithm under unconfined condition between two ships and many with " off-line form "
Be to free track in the presence of the ship planning of conflict, thereby result in each ship free the dynamic adaptable and robustness of track compared with
Difference.Additionally, in ship real navigation, influenceed by various factors such as meteorological condition, navigation equipment and driver operations, it
Running status often not exclusively belong to a certain specific motion state, needed during ship trajectory predictions consider it is various with
The influence of machine factor, implements rolling forecast to its Future Trajectory and strengthens its rail by the newest characteristic for obtaining all kinds of enchancement factors
The robustness of mark prediction.
The content of the invention
The technical problem to be solved in the present invention is to provide a kind of preferably navigation based on Rolling Planning strategy of robustness and hands over
Logical method of control, the ship trajectory predictions precision of the method is higher and can effectively prevent vessel motion conflict.
Realize that the technical scheme of the object of the invention is to provide a kind of navigation method of control based on Rolling Planning strategy, including
Following several steps:
1. the real-time and historical position information of ship is obtained by sea radar, the positional information of each ship is discrete two-dimensional
Position sequence x'=[x1',x2',...,xn'] and y'=[y1',y2',...,yn'], by application wavelet transformation theory to original
Discrete two-dimensional position sequence x'=[x1',x2',...,xn'] and y'=[y1',y2',...,yn'] preliminary treatment is carried out, so as to obtain
Take the denoising discrete two-dimensional position sequence x=[x of ship1,x2,...,xn] and y=[y1,y2,...,yn];
2. in each sampling instant, the real-time and historical position information of the ship 1. obtained according to step is rolled and speculates future
The track of ship in period, its detailed process is as follows:
2.1) ship track data pretreatment, the original discrete two-dimensional position sequence x=[x of ship acquired in1,
x2,...,xn] and y=[y1,y2,...,yn], treatment is carried out to it using first-order difference method and obtains new ship discrete location
Sequence △ x=[△ x1,△x2,...,△xn-1] and △ y=[△ y1,△y2,...,△yn-1], wherein △ xi=xi+1-xi,△
yi=yi+1-yi(i=1,2 ..., n-1);
2.2) ship track data is clustered, to ship discrete two-dimensional position sequence △ x and △ y new after treatment, is passed through
Setting cluster number M', is clustered to it respectively using K-means clustering algorithms;
2.3) parameter training is carried out using HMM to ship track data in each sampling instant, by inciting somebody to action
Vessel motion track data △ x and △ y after treatment is considered as the aobvious observation of hidden Markov models, by setting hidden status number
Mesh N and parameter update period τ ', roll and obtain newest hidden Ma Erke according to T' nearest position detection value and use B-W algorithms
Husband's model parameter λ ';
2.4) according to HMM parameter, obtained corresponding to current time observation using Viterbi algorithm
Hidden state q;
2.5) in each sampling instant, by setting prediction time domain W, the hidden state q based on ship current time, obtain not
Carry out the position prediction value O of period ship;
3. in each sampling instant, based on the current running status of ship and historical position observation sequence, marine site wind is obtained
The numerical value of field variable;
4. in each sampling instant, the ship of running status and setting based on each ship needs to meet when being run in marine site
Safety regulation collection, when being possible to occur violating the situation of safety regulation between ship, to its dynamic behaviour implementing monitoring and be
Maritime traffic control centre provides timely warning information;
5. when warning information occurs, on the premise of ship physical property and marine site traffic rules is met, by setting
Optimizing index function and wind field variable value is incorporated, ship collision avoidance track is rolled using Adaptive Control Theory method
Planning, and program results is transferred to each ship execution, its detailed process is as follows:
5.1) termination reference point locations P, collision avoidance policy control time domain Θ, the trajectory predictions of ship collision avoidance trajectory planning are set
Time domain W;
5.2) on the premise of being set in given optimizing index function, based on cooperative collision avoidance trajectory planning thought, by giving
Each ship assigns different weight and incorporates real-time wind field variable filtering numerical value, obtains the collision avoidance track of each ship and keeps away
Hit control strategy and program results is transferred to each ship and perform, and each ship only implements its first in Rolling Planning interval
Optimal Control Strategy;
5.3) in next sampling instant, repeat step 5.2) until each ship reaches it and frees terminal.
Further, the step 1. in, by application wavelet transformation theory to original discrete two-dimensional position sequence x'=
[x1',x2',...,xn'] and y'=[y1',y2',...,yn'] preliminary treatment is carried out, so as to obtain the denoising discrete two-dimensional of ship
Position sequence x=[x1,x2,...,xn] and y=[y1,y2,...,yn]:For the original two dimensional sequence data x'=for giving
[x1',x2',...,xn'], it is carried out approximately respectively using the linear representation of following form:
Wherein:
F'(x' the function expression to being obtained after data smoothing processing) is represented, ψ (x') represents female ripple, and δ, J and K are small
Wave conversion constant, ψJ,K(x') transition form of female ripple, c are representedJ,KThe function coefficients that expression is obtained by wavelet transform procedure, its body
Wavelet ψ is showedJ,K(x') to the weight size of whole approximation to function, if this coefficient very little, then it means wavelet ψJ,K(x')
Weight it is also smaller, thus can be on the premise of not influence function key property, from during approximation to function by wavelet ψJ,K
(x') remove;In real data processing procedure, " threshold transition " is implemented by given threshold χ, work as cJ,K<During χ, c is setJ,K
=0;The selection of threshold function table uses the following two kinds mode:
With
For y'=[y1',y2',...,yn'], being also adopted by the above method carries out denoising;
The detailed process that 3. step obtains the numerical value of marine site wind field variable is as follows:
3.1) stop position of ship is set as track reference coordinate origin and is set up axis of abscissas in the horizontal plane and is indulged
Reference axis;
3.2) when ship is in straight running condition and at the uniform velocity turning running status, marine site wind field linear filtering mould is built
Type x1(t+ △ t)=F (t) x1(t)+w (t) and z (t)=H (t) x1T ()+v (t) obtains wind field variable value, wherein △ t are represented
Sampling interval, x1T () represents the state vector of t, z (t) represents the observation vector of t, and x1(t)=[x (t), y (t),
vx(t),vy(t),wx(t),wy(t)]T, wherein x (t) and y (t) represent t vessel position in axis of abscissas and ordinate respectively
Component on axle, vx(t) and vyT () represents component of the t speed of the ship in metres per second on axis of abscissas and axis of ordinates, w respectivelyx(t)
And wyT () represents component of the t wind field numerical value on axis of abscissas and axis of ordinates respectively, F (t) and H (t) represent shape respectively
State transfer matrix and output calculation matrix, w (t) and v (t) represent system noise vector sum measurement noise vector respectively:
When ship is in speed change turning running status, marine site wind field nonlinear filtering wave pattern x is built1(t+ △ t)=Ψ
(t,x1(t), u (t))+w (t), z (t)=Ω (t, x1(t))+v (t) and u (t)=[ωa(t),γa(t)]T, wherein Ψ () and
Ω () represents state-transition matrix and output calculation matrix, ω respectivelya(t) and γaT () represents turning rate and rate of acceleration respectively:
Wherein:△ t represent sampling time interval,
3.3) Filtering Model according to constructed by obtains the numerical value of wind field variable.
Further, the step 2. in, step 2.3) in determine flight path HMM parameter lambda '=(π, A, B)
Process it is as follows:
2.3.1) variable assigns initial value:Variable π is given using being uniformly distributedi, aijAnd bj(ok) assign initial value WithAnd
It is set to meet constraints:WithThus
To λ0=(π0,A0,B0), wherein okRepresent a certain aobvious observation, π0、A0And B0It is respectively by elementWithThe square of composition
Battle array, makes parameter l=0, o=(ot-T'+1,...,ot-1,ot) it is T' historical position observation before current time t;
2.3.2 E-M algorithms) are performed:
2.3.2.1) E- steps:By λlCalculate ξe(i, j) and γe(si);
VariableSo
Wherein s represents a certain hidden state;
2.3.2.2) M- steps:WithRespectively
Estimate πi, aijAnd bj(ok) and thus obtain λl+1;
2.3.2.3) circulate:L=l+1, repeats E- steps and M- steps, until πi、aijAnd bj(ok) convergence, i.e.,
|P(o|λl+1)-P(o|λl)|<ε, wherein parameter ε=0.00001, return to step 2.3.2.4);
2.3.2.4):Make λ '=λl+1, algorithm terminates.
Further, the step 2. in, step 2.4) determine the iterative process of the optimal hidden status switch of ship track
It is as follows:
2.4.1) variable assigns initial value:Make g=2, βT'(si(the s of)=1i∈ S), δ1(si)=πibi(o1), ψ1(si)=0, its
In,
, wherein variable ψg(sj) represent make variable δg-1(si)aijTake the hidden state s of ship track of maximumi, parameter S represents
The set of hidden state;
2.4.2) recursive process:
2.4.3) moment renewal:G=g+1 is made, if g≤T', return to step 2.4.2), otherwise iteration ends and go to step
2.4.4);
2.4.4)Go to step 2.4.5);
2.4.5) optimal hidden status switch is obtained:
2.4.5.1) variable assigns initial value:Make g=T'-1;
2.4.5.2) backward recursion:
2.4.5.3) moment renewal:G=g-1 is made, if g >=1, return to step 2.4.5.2), otherwise terminate.
Further, the step 2. in, the value of cluster number M' is 4, and the value of hidden state number N is 3, when parameter updates
Section τ ' is 30 seconds, and T' is 10, and prediction time domain W is 300 seconds.
Further, the step 4. in the dynamic behaviour implementing monitoring of each ship and for maritime traffic control centre carries
Detailed process for timely warning information is as follows:
4.1) the safety regulation collection D that need to be met when construction ship runs in marine sitemr(t)≥Dmin, wherein DmrT () represents
Any two ship m and ship r t distance, DminRepresent the minimum safe distance between ship;
4.2) according to the sampling time, set up by the observer Λ of the continuous running status of ship to discrete sampling state:Γ→
Ξ, wherein Γ represent the continuous running status of ship, and Ξ represents the discrete sampling state of ship;
4.3) as the observer Λ of ship m and rmAnd ΛrDiscrete observation numerical value ΞmAnd ΞrShow the vector not in t
When safety regulation is concentrated, i.e. relational expression Dmr(t)≥DminWhen invalid, alarm letter is sent to maritime traffic control centre at once
Breath.
Further, step 5. in, step 5.2) detailed process be:Order
WhereinRepresent distance between t ship R present positions and next navigation channel point square, PR(t)=
(xRt,yRt),The priority index of so t ship R may be set to:
Wherein ZtRepresent the ship number for existing in t marine site and conflicting, from the implication of priority index, ship away from
From its next navigation channel point more close to, its priority is higher;
Setting optimizing index
, wherein R ∈ I (t) expressions ship code and I (t)={ 1,2 ..., Zt, PR(t+h △ t) represents ship at the moment
The position vector of (t+h △ t),Represent that ship R's frees terminating point, uRThe optimal control sequence of ship R to be optimized is represented,
QRtIt is positive definite diagonal matrix, its diagonal element is priority index Ls of the ship R in tRt, and
Further, 5. middle termination reference point locations P is set as next navigation channel point of vessel motion, collision avoidance to the step
Policy control time domain Θ is 300 seconds;Trajectory predictions time domain W is 300 seconds.
The present invention has positive effect:(1) present invention during the real-time estimate of ship track, incorporated it is random because
The influence of element, the rolling track prediction scheme for being used can in time extract the changing condition of extraneous enchancement factor, improve ship
The accuracy of oceangoing ship trajectory predictions.
(2) present invention has incorporated the influence of wind field in marine site during ship conflict Resolution, and the rolling for being used is freed
Trajectory planning scheme can track be freed in adjustment in time according to the change of wind field in marine site, improves the robust of ship conflict Resolution
Property.
(3) present invention is based on different performance index, can free trajectory planning side for the multiple ships in the presence of conflict are provided
Case, improves the economy of vessel motion and the utilization rate of sea area resources.
Brief description of the drawings
Fig. 1 is the short-term Track Pick-up schematic flow sheet of vessel motion in the present invention;
Fig. 2 is the Wind filter method flow schematic diagram in the present invention;
Fig. 3 is the vessel motion situation monitoring schematic flow sheet in the present invention;
Fig. 4 is the ship collision avoidance track optimizing method schematic flow sheet in the present invention.
Specific embodiment
(embodiment 1)
A kind of navigation method of control based on Rolling Planning strategy of the present embodiment includes following several steps:
1. the real-time and historical position information of ship is obtained by sea radar, the positional information of each ship is discrete two-dimensional
Position sequence x'=[x1',x2',...,xn'] and y'=[y1',y2',...,yn'], by application wavelet transformation theory to original
Discrete two-dimensional position sequence x'=[x1',x2',...,xn'] and y'=[y1',y2',...,yn'] preliminary treatment is carried out, so as to obtain
Take the denoising discrete two-dimensional position sequence x=[x of ship1,x2,...,xn] and y=[y1,y2,...,yn]:It is original for what is given
Two-dimensional sequence data x'=[x1',x2',...,xn'], it is carried out approximately respectively using the linear representation of following form:
Wherein:
F'(x' the function expression to being obtained after data smoothing processing) is represented, ψ (x') represents female ripple, and δ, J and K are small
Wave conversion constant, ψJ,K(x') transition form of female ripple, c are representedJ,KThe function coefficients that expression is obtained by wavelet transform procedure, its body
Wavelet ψ is showedJ,K(x') to the weight size of whole approximation to function, if this coefficient very little, then it means wavelet ψJ,K(x')
Weight it is also smaller, thus can be on the premise of not influence function key property, from during approximation to function by wavelet ψJ,K
(x') remove;In real data processing procedure, " threshold transition " is implemented by given threshold χ, work as cJ,K<During χ, c is setJ,K
=0;The selection of threshold function table uses the following two kinds mode:
With
For y'=[y1',y2',...,yn'], being also adopted by the above method carries out denoising.
2. in each sampling instant, the real-time and historical position information of the ship 1. obtained according to step is rolled and speculates future
The track of ship, sees Fig. 1 in period, and its detailed process is as follows:
2.1) ship track data pretreatment, the original discrete two-dimensional position sequence x=[x of ship acquired in1,
x2,...,xn] and y=[y1,y2,...,yn], treatment is carried out to it using first-order difference method and obtains new ship discrete location
Sequence △ x=[△ x1,△x2,...,△xn-1] and △ y=[△ y1,△y2,...,△yn-1], wherein △ xi=xi+1-xi,△
yi=yi+1-yi(i=1,2 ..., n-1);
2.2) ship track data is clustered, to ship discrete two-dimensional position sequence △ x and △ y new after treatment, is passed through
Setting cluster number M', is clustered to it respectively using K-means clustering algorithms;
2.3) parameter training is carried out using HMM to ship track data in each sampling instant, by inciting somebody to action
Vessel motion track data △ x and △ y after treatment is considered as the aobvious observation of hidden Markov models, by setting hidden status number
Mesh N and parameter update period τ ', roll and obtain newest hidden Ma Erke according to T' nearest position detection value and use B-W algorithms
Husband's model parameter λ ';Determine flight path HMM parameter lambda '=the process of (π, A, B) is as follows:
2.3.1) variable assigns initial value:Variable π is given using being uniformly distributedi, aijAnd bj(ok) assign initial value WithAnd
It is set to meet constraints:WithThus
To λ0=(π0,A0,B0), wherein okRepresent a certain aobvious observation, π0、A0And B0It is respectively by elementWithThe square of composition
Battle array, makes parameter l=0, o=(ot-T'+1,...,ot-1,ot) it is T' historical position observation before current time t;
2.3.2 E-M algorithms) are performed:
2.3.2.1) E- steps:By λlCalculate ξe(i, j) and γe(si);
VariableSo
Wherein s represents a certain hidden state;
2.3.2.2) M- steps:WithRespectively
Estimate πi, aijAnd bj(ok) and thus obtain λl+1;
2.3.2.3) circulate:L=l+1, repeats E- steps and M- steps, until πi、aijAnd bj(ok) convergence, i.e.,
|P(o|λl+1)-P(o|λl)|<ε, wherein parameter ε=0.00001, return to step 2.3.2.4);
2.3.2.4):Make λ '=λl+1, algorithm terminates.
2.4) according to HMM parameter, obtained corresponding to current time observation using Viterbi algorithm
Hidden state q;Determine that the iterative process of the optimal hidden status switch of ship track is as follows:
2.4.1) variable assigns initial value:Make g=2, βT'(si(the s of)=1i∈ S), δ1(si)=πibi(o1), ψ1(si)=0, its
In,
, wherein variable ψg(sj) represent make variable δg-1(si)aijTake the hidden state s of ship track of maximumi, parameter S represents
The set of hidden state;
2.4.2) recursive process:
2.4.3) moment renewal:G=g+1 is made, if g≤T', return to step 2.4.2), otherwise iteration ends and go to step
2.4.4);
2.4.4)Go to step 2.4.5);
2.4.5) optimal hidden status switch is obtained:
2.4.5.1) variable assigns initial value:Make g=T'-1;
2.4.5.2) backward recursion:
2.4.5.3) moment renewal:G=g-1 is made, if g >=1, return to step 2.4.5.2), otherwise terminate.
2.5) in each sampling instant, by setting prediction time domain W, the hidden state q based on ship current time, obtain not
Carry out the position prediction value O of period ship.
The value of above-mentioned cluster number M' is 4, and the value of hidden state number N is 3, and parameter updated period τ ' for 30 seconds, and T' is 10,
Prediction time domain W is 300 seconds.
3. in each sampling instant, based on the current running status of ship and historical position observation sequence, marine site wind is obtained
The numerical value of field variable, is shown in Fig. 2, and its detailed process is as follows:
3.1) stop position of ship is set as track reference coordinate origin and is set up axis of abscissas in the horizontal plane and is indulged
Reference axis;
3.2) when ship is in straight running condition and at the uniform velocity turning running status, marine site wind field linear filtering mould is built
Type x1(t+ △ t)=F (t) x1(t)+w (t) and z (t)=H (t) x1T ()+v (t) obtains wind field variable value, wherein △ t are represented
Sampling interval, x1T () represents the state vector of t, z (t) represents the observation vector of t, and x1(t)=[x (t), y (t),
vx(t),vy(t),wx(t),wy(t)]T, wherein x (t) and y (t) represent t vessel position in axis of abscissas and ordinate respectively
Component on axle, vx(t) and vyT () represents component of the t speed of the ship in metres per second on axis of abscissas and axis of ordinates, w respectivelyx(t)
And wyT () represents component of the t wind field numerical value on axis of abscissas and axis of ordinates respectively, F (t) and H (t) represent shape respectively
State transfer matrix and output calculation matrix, w (t) and v (t) represent system noise vector sum measurement noise vector respectively:
When ship is in speed change turning running status, marine site wind field nonlinear filtering wave pattern x is built1(t+ △ t)=Ψ
(t,x1(t), u (t))+w (t), z (t)=Ω (t, x1(t))+v (t) and u (t)=[ωa(t),γa(t)]T, wherein Ψ () and
Ω () represents state-transition matrix and output calculation matrix, ω respectivelya(t) and γaT () represents turning rate and rate of acceleration respectively:
Wherein:△ t represent sampling time interval,
3.3) Filtering Model according to constructed by obtains the numerical value of wind field variable.
4. in each sampling instant, the ship of running status and setting based on each ship needs to meet when being run in marine site
Safety regulation collection, when being possible to occur violating the situation of safety regulation between ship, to its dynamic behaviour implementing monitoring and be
Maritime traffic control centre provides timely warning information, sees Fig. 3, and its detailed process is as follows:
4.1) the safety regulation collection D that need to be met when construction ship runs in marine sitemr(t)≥Dmin, wherein DmrT () represents
Any two ship m and ship r t distance, DminRepresent the minimum safe distance between ship;
4.2) according to the sampling time, set up by the observer Λ of the continuous running status of ship to discrete sampling state:Γ→
Ξ, wherein Γ represent the continuous running status of ship, and Ξ represents the discrete sampling state of ship;
4.3) as the observer Λ of ship m and rmAnd ΛrDiscrete observation numerical value ΞmAnd ΞrShow the vector not in t
When safety regulation is concentrated, i.e. relational expression Dmr(t)≥DminWhen invalid, alarm letter is sent to maritime traffic control centre at once
Breath.
5. when warning information occurs, on the premise of ship physical property and marine site traffic rules is met, by setting
Optimizing index function and wind field variable value is incorporated, ship collision avoidance track is rolled using Adaptive Control Theory method
Planning, and program results is transferred to each ship execution, see Fig. 4, its detailed process is as follows:
5.1) termination reference point locations P, collision avoidance policy control time domain Θ, the trajectory predictions of ship collision avoidance trajectory planning are set
Time domain W;
5.2) on the premise of being set in given optimizing index function, based on cooperative collision avoidance trajectory planning thought, by giving
Each ship assigns different weight and incorporates real-time wind field variable filtering numerical value, obtains the collision avoidance track of each ship and keeps away
Hit control strategy and program results is transferred to each ship and perform, and each ship only implements its first in Rolling Planning interval
Optimal Control Strategy:Order
WhereinRepresent distance between t ship R present positions and next navigation channel point square, PR(t)=
(xRt,yRt),The priority index of so t ship R may be set to:
Wherein ZtRepresent the ship number for existing in t marine site and conflicting, from the implication of priority index, ship away from
From its next navigation channel point more close to, its priority is higher;
Setting optimizing index
, wherein R ∈ I (t) expressions ship code and I (t)={ 1,2 ..., Zt, PR(t+h △ t) represents ship at the moment
The position vector of (t+h △ t),Represent that ship R's frees terminating point, uRThe optimal control sequence of ship R to be optimized is represented,
QRtIt is positive definite diagonal matrix, its diagonal element is priority index Ls of the ship R in tRt, and
5.3) in next sampling instant, repeat step 5.2 is until each ship reaches it and frees terminal.
Above-mentioned termination reference point locations P is set as next navigation channel point of vessel motion, and collision avoidance policy control time domain Θ is
300 seconds;Trajectory predictions time domain W is 300 seconds.
Obviously, above-described embodiment is only intended to clearly illustrate example of the present invention, and is not to of the invention
The restriction of implementation method.For those of ordinary skill in the field, it can also be made on the basis of the above description
The change or variation of its multi-form.There is no need and unable to be exhaustive to all of implementation method.And these belong to this hair
Obvious change that bright spirit is extended out or among changing still in protection scope of the present invention.
Claims (1)
1. a kind of navigation method of control based on Rolling Planning strategy, it is characterised in that including following several steps:
1. the real-time and historical position information of ship is obtained by sea radar, the positional information of each ship is discrete two-dimensional position
Sequence x'=[x1',x2',...,xn'] and y'=[y1',y2',...,yn'], by application wavelet transformation theory to original discrete
Two-dimensional position sequence x'=[x1',x2',...,xn'] and y'=[y1',y2',...,yn'] preliminary treatment is carried out, so as to obtain ship
The denoising discrete two-dimensional position sequence x=[x of oceangoing ship1,x2,...,xn] and y=[y1,y2,...,yn];
2. in each sampling instant, the real-time and historical position information of the ship 1. obtained according to step is rolled and speculates future time period
The track of interior ship, its detailed process is as follows:
2.1) ship track data pretreatment, the original discrete two-dimensional position sequence x=[x of ship acquired in1,x2,...,
xn] and y=[y1,y2,...,yn], treatment is carried out to it using first-order difference method and obtains new ship discrete location sequence △ x
=[△ x1,△x2,...,△xn-1] and △ y=[△ y1,△y2,...,△yn-1], wherein △ xi=xi+1-xi,△yi=yi+1-
yi(i=1,2 ..., n-1);
2.2) ship track data is clustered, to ship discrete two-dimensional position sequence △ x and △ y new after treatment, by setting
Cluster number M', is clustered to it respectively using K-means clustering algorithms;
2.3) parameter training is carried out using HMM to ship track data in each sampling instant, by that will process
Vessel motion track data △ x and △ y afterwards is considered as the aobvious observation of hidden Markov models, by setting hidden state number N
Period τ ' is updated with parameter, is rolled and is obtained newest Hidden Markov according to T' nearest position detection value and use B-W algorithms
Model parameter λ ';
2.4) according to HMM parameter, the hidden shape corresponding to current time observation is obtained using Viterbi algorithm
State q;
2.5) in each sampling instant, by setting prediction time domain W, the hidden state q based on ship current time, when obtaining following
The position prediction value O of section ship;
3. in each sampling instant, based on the current running status of ship and historical position observation sequence, obtain marine site wind field and become
The numerical value of amount;
4. in each sampling instant, the peace that the ship of running status and setting based on each ship need to meet when being run in marine site
Full rule set, when being possible to occur violating the situation of safety regulation between ship, to its dynamic behaviour implementing monitoring and for marine
Traffic control center provides timely warning information;
5. when warning information occurs, on the premise of ship physical property and marine site traffic rules is met, optimized by setting
Target function and wind field variable value is incorporated, rolling rule are carried out to ship collision avoidance track using Adaptive Control Theory method
Draw, and program results is transferred to each ship and perform, its detailed process is as follows:
5.1) termination reference point locations P, collision avoidance policy control time domain Θ, the trajectory predictions time domain of ship collision avoidance trajectory planning are set
W;
5.2) on the premise of being set in given optimizing index function, based on cooperative collision avoidance trajectory planning thought, by each
Ship assigns different weight and incorporates real-time wind field variable filtering numerical value, obtains collision avoidance track and the collision avoidance control of each ship
Program results is simultaneously transferred to each ship execution, and each ship only implements its first optimization in Rolling Planning interval by system strategy
Control strategy;
5.3) in next sampling instant, repeat step 5.2) until each ship reaches it and frees terminal;
The detailed process that 3. step obtains the numerical value of marine site wind field variable is as follows:
3.1) stop position of ship is set as track reference coordinate origin and sets up axis of abscissas and ordinate in the horizontal plane
Axle;
3.2) when ship is in straight running condition and at the uniform velocity turning running status, structure marine site wind field linear filtering model x1
(t+ △ t)=F (t) x1(t)+w (t) and z (t)=H (t) x1T ()+v (t) obtains wind field variable value, wherein △ t represent sampling
Interval, x1T () represents the state vector of t, z (t) represents the observation vector of t, and x1(t)=[x (t), y (t), vx
(t),vy(t),wx(t),wy(t)]T, wherein x (t) and y (t) represent t vessel position in axis of abscissas and ordinate respectively
Component on axle, vx(t) and vyT () represents component of the t speed of the ship in metres per second on axis of abscissas and axis of ordinates, w respectivelyx(t)
And wyT () represents component of the t wind field numerical value on axis of abscissas and axis of ordinates respectively, F (t) and H (t) represent shape respectively
State transfer matrix and output calculation matrix, w (t) and v (t) represent system noise vector sum measurement noise vector respectively:
When ship is in speed change turning running status, marine site wind field nonlinear filtering wave pattern x is built1(t+ △ t)=Ψ (t, x1
(t), u (t))+w (t), z (t)=Ω (t, x1(t))+v (t) and u (t)=[ωa(t),γa(t)]T, wherein Ψ () and Ω
() represents state-transition matrix and output calculation matrix, ω respectivelya(t) and γaT () represents turning rate and rate of acceleration respectively:
Wherein:△ t represent sampling time interval,
3.3) Filtering Model according to constructed by obtains the numerical value of wind field variable.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710141983.4A CN106803361A (en) | 2014-12-30 | 2014-12-30 | A kind of navigation method of control based on Rolling Planning strategy |
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201410849264.4A CN104504934B (en) | 2014-12-30 | 2014-12-30 | A kind of navigation traffic control method |
CN201710141983.4A CN106803361A (en) | 2014-12-30 | 2014-12-30 | A kind of navigation method of control based on Rolling Planning strategy |
Related Parent Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410849264.4A Division CN104504934B (en) | 2014-12-30 | 2014-12-30 | A kind of navigation traffic control method |
Publications (1)
Publication Number | Publication Date |
---|---|
CN106803361A true CN106803361A (en) | 2017-06-06 |
Family
ID=52946550
Family Applications (4)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410849264.4A Active CN104504934B (en) | 2014-12-30 | 2014-12-30 | A kind of navigation traffic control method |
CN201710141982.XA Pending CN106803360A (en) | 2014-12-30 | 2014-12-30 | A kind of navigation traffic control method based on Rolling Planning strategy |
CN201710141983.4A Pending CN106803361A (en) | 2014-12-30 | 2014-12-30 | A kind of navigation method of control based on Rolling Planning strategy |
CN201710141985.3A Pending CN106846916A (en) | 2014-12-30 | 2014-12-30 | A kind of navigation method of control |
Family Applications Before (2)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201410849264.4A Active CN104504934B (en) | 2014-12-30 | 2014-12-30 | A kind of navigation traffic control method |
CN201710141982.XA Pending CN106803360A (en) | 2014-12-30 | 2014-12-30 | A kind of navigation traffic control method based on Rolling Planning strategy |
Family Applications After (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710141985.3A Pending CN106846916A (en) | 2014-12-30 | 2014-12-30 | A kind of navigation method of control |
Country Status (1)
Country | Link |
---|---|
CN (4) | CN104504934B (en) |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106154296A (en) * | 2015-06-26 | 2016-11-23 | 安徽华米信息科技有限公司 | The method of adjustment of a kind of path locus and device |
CN106971628A (en) * | 2017-05-11 | 2017-07-21 | 厦门卫星定位应用股份有限公司 | A kind of ship track monitoring system and method |
CN107577230B (en) * | 2017-08-16 | 2020-01-14 | 武汉理工大学 | Intelligent collision avoidance system for unmanned ship |
CN109657863B (en) * | 2018-12-20 | 2021-06-25 | 智慧航海(青岛)科技有限公司 | Firefly algorithm-based unmanned ship global path dynamic optimization method |
CN113221449B (en) * | 2021-04-27 | 2024-03-15 | 中国科学院国家空间科学中心 | Ship track real-time prediction method and system based on optimal strategy learning |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030028293A1 (en) * | 1999-12-21 | 2003-02-06 | Christian Jankowiak | Sea surveillance method |
US20050124291A1 (en) * | 2001-11-26 | 2005-06-09 | Hart Nicholas R. | Satellite system for vessel identification |
CN101866157A (en) * | 2009-04-20 | 2010-10-20 | 中国水产科学研究院东海水产研究所 | Ship track recording and monitoring method |
CN201732479U (en) * | 2010-07-12 | 2011-02-02 | 郑硕钧 | Vehicle and vessel collision early warning control system |
US20110210865A1 (en) * | 2010-02-26 | 2011-09-01 | Electronics And Telecommunications Research Institute | Traffic control apparatus and method using navigation route information of vehicle for each navigational situation |
CN102194332A (en) * | 2011-03-24 | 2011-09-21 | 中国船舶重工集团公司第七○九研究所 | Self-adaptation flight path data correlation method |
CN103106812A (en) * | 2013-01-17 | 2013-05-15 | 中华人民共和国深圳海事局 | Method obtaining sea ship system average collision risks |
CN104899263A (en) * | 2015-05-22 | 2015-09-09 | 华中师范大学 | Ship trajectory mining, analysis and monitoring method based on specific region |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
DE19941635A1 (en) * | 1999-09-01 | 2001-03-22 | Reinhard Mueller | Process for resolving traffic conflicts through the use of master-slave structures in localized areas in shipping |
KR101104964B1 (en) * | 2008-12-04 | 2012-01-12 | 한국전자통신연구원 | apparatus and method for avoiding collision |
CN202615628U (en) * | 2012-05-25 | 2012-12-19 | 南通航运职业技术学院 | AIS-based offshore water intelligent traffic management system |
CN103336863B (en) * | 2013-06-24 | 2016-06-01 | 北京航空航天大学 | The flight intent recognition methods of flight path observed data of flying based on radar |
-
2014
- 2014-12-30 CN CN201410849264.4A patent/CN104504934B/en active Active
- 2014-12-30 CN CN201710141982.XA patent/CN106803360A/en active Pending
- 2014-12-30 CN CN201710141983.4A patent/CN106803361A/en active Pending
- 2014-12-30 CN CN201710141985.3A patent/CN106846916A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US20030028293A1 (en) * | 1999-12-21 | 2003-02-06 | Christian Jankowiak | Sea surveillance method |
US20050124291A1 (en) * | 2001-11-26 | 2005-06-09 | Hart Nicholas R. | Satellite system for vessel identification |
CN101866157A (en) * | 2009-04-20 | 2010-10-20 | 中国水产科学研究院东海水产研究所 | Ship track recording and monitoring method |
US20110210865A1 (en) * | 2010-02-26 | 2011-09-01 | Electronics And Telecommunications Research Institute | Traffic control apparatus and method using navigation route information of vehicle for each navigational situation |
CN201732479U (en) * | 2010-07-12 | 2011-02-02 | 郑硕钧 | Vehicle and vessel collision early warning control system |
CN102194332A (en) * | 2011-03-24 | 2011-09-21 | 中国船舶重工集团公司第七○九研究所 | Self-adaptation flight path data correlation method |
CN103106812A (en) * | 2013-01-17 | 2013-05-15 | 中华人民共和国深圳海事局 | Method obtaining sea ship system average collision risks |
CN104899263A (en) * | 2015-05-22 | 2015-09-09 | 华中师范大学 | Ship trajectory mining, analysis and monitoring method based on specific region |
Non-Patent Citations (5)
Title |
---|
D.DELAHAYE ET AL.: "Windfield estimation by radar track Kalman filtering and vector spline extrapolation", 《THE 22ND DIGITAL AVIONICS SYSTEMS CONFERENCE》 * |
吴勇: "基于小波的信号去噪方法研究", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
杨君兰: "基于复杂度建模的船舶碰撞预警研究", 《中国优秀硕士学位论文全文库 工程科技Ⅱ辑》 * |
柳长昕: "基于二维隐马尔科夫模型的离心泵故障诊断方法研究", 《中国优秀硕士学位论文全文数据库 工程科技II辑》 * |
桑凌志等: "基于航迹预测的实时船舶安全预警方法", 《中国安全科学学报》 * |
Also Published As
Publication number | Publication date |
---|---|
CN106846916A (en) | 2017-06-13 |
CN104504934B (en) | 2017-03-08 |
CN106803360A (en) | 2017-06-06 |
CN104504934A (en) | 2015-04-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104462856B (en) | Ship conflict method for early warning | |
CN104484726B (en) | Ship track real-time predicting method | |
CN104537891B (en) | A kind of boats and ships track real-time predicting method | |
CN104504277B (en) | A kind of ship conflict method for early warning | |
CN104504934B (en) | A kind of navigation traffic control method | |
CN104504935B (en) | Maritime traffic control method | |
CN113221449B (en) | Ship track real-time prediction method and system based on optimal strategy learning | |
CN104050506A (en) | Aircraft conflict detection method based on Spiking neural network | |
CN104485023B (en) | Planning method for ship conflict release | |
CN112907970A (en) | Variable lane steering control method based on vehicle queuing length change rate | |
CN110723152B (en) | Artificial intelligence detection method, device, equipment or storage medium for guaranteeing automatic driving safety | |
CN113887789B (en) | Improved ship track prediction method and device based on cyclic neural network | |
Wei et al. | Big multi-step ship motion forecasting using a novel hybrid model based on real-time decomposition, boosting algorithm and error correction framework | |
Yang et al. | Improved reinforcement learning for collision-free local path planning of dynamic obstacle | |
Zhang et al. | Time-delay wavelet network predictor based on sensitivity analysis with application to predictive ship course control | |
CN109002860B (en) | Intelligent adaptive matching prediction method for sudden change wind speed along high-speed railway | |
He et al. | Towards the Deep Learning-Based Autonomous Collision Avoidance | |
CN117131940A (en) | Situation reasoning method for unmanned surface vehicle | |
CN117474168A (en) | Ship track prediction method and device of CNN-BiGRU based on attention mechanism | |
CN117115731A (en) | Dust shielding system and method for building construction | |
CN116118772A (en) | Uncertainty-considered automatic driving reinforcement learning movement planning method and system | |
Liu et al. | Real-time online forecasting model of ship rolling motion based on chaotic online LSSVM | |
CN116994442A (en) | Self-adaptive traffic signal control method based on CQL off-line reinforcement learning | |
CN116382285A (en) | Multi-unmanned-ship-area patrol method and system based on situation information under countermeasure environment | |
CN111914982A (en) | Variational self-coding method combined with randomization technology |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170606 |