CN110362107A - The method for solving of multiple no-manned plane Task Allocation Problem based on immune optimization algorithm - Google Patents
The method for solving of multiple no-manned plane Task Allocation Problem based on immune optimization algorithm Download PDFInfo
- Publication number
- CN110362107A CN110362107A CN201910520241.1A CN201910520241A CN110362107A CN 110362107 A CN110362107 A CN 110362107A CN 201910520241 A CN201910520241 A CN 201910520241A CN 110362107 A CN110362107 A CN 110362107A
- Authority
- CN
- China
- Prior art keywords
- task
- formula
- unmanned plane
- airport
- immune
- 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.)
- Withdrawn
Links
- 238000005457 optimization Methods 0.000 title claims abstract description 21
- 238000000034 method Methods 0.000 title claims abstract description 18
- 230000009467 reduction Effects 0.000 claims abstract description 6
- 239000011159 matrix material Substances 0.000 claims description 16
- 230000006870 function Effects 0.000 claims description 14
- 238000013377 clone selection method Methods 0.000 claims description 13
- 230000000875 corresponding effect Effects 0.000 claims description 13
- 230000001850 reproductive effect Effects 0.000 claims description 12
- 230000007246 mechanism Effects 0.000 claims description 8
- 239000000427 antigen Substances 0.000 claims description 7
- 108091007433 antigens Proteins 0.000 claims description 7
- 102000036639 antigens Human genes 0.000 claims description 7
- NAWXUBYGYWOOIX-SFHVURJKSA-N (2s)-2-[[4-[2-(2,4-diaminoquinazolin-6-yl)ethyl]benzoyl]amino]-4-methylidenepentanedioic acid Chemical compound C1=CC2=NC(N)=NC(N)=C2C=C1CCC1=CC=C(C(=O)N[C@@H](CC(=C)C(O)=O)C(O)=O)C=C1 NAWXUBYGYWOOIX-SFHVURJKSA-N 0.000 claims description 5
- 230000035772 mutation Effects 0.000 claims description 5
- 238000010353 genetic engineering Methods 0.000 claims description 4
- 230000015654 memory Effects 0.000 claims description 4
- 230000008569 process Effects 0.000 claims description 4
- 230000005764 inhibitory process Effects 0.000 claims description 3
- 210000000349 chromosome Anatomy 0.000 claims description 2
- 230000002596 correlated effect Effects 0.000 claims description 2
- 230000006054 immunological memory Effects 0.000 claims description 2
- 230000002401 inhibitory effect Effects 0.000 claims description 2
- 238000013178 mathematical model Methods 0.000 claims description 2
- 108090000623 proteins and genes Proteins 0.000 claims description 2
- 238000000844 transformation Methods 0.000 claims description 2
- 230000009466 transformation Effects 0.000 claims description 2
- 210000004027 cell Anatomy 0.000 claims 1
- 230000008859 change Effects 0.000 claims 1
- 239000000203 mixture Substances 0.000 claims 1
- 210000000987 immune system Anatomy 0.000 description 5
- 238000010586 diagram Methods 0.000 description 4
- 230000002068 genetic effect Effects 0.000 description 4
- 230000003044 adaptive effect Effects 0.000 description 3
- 230000006399 behavior Effects 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 238000001027 hydrothermal synthesis Methods 0.000 description 2
- 230000013016 learning Effects 0.000 description 2
- PXFBZOLANLWPMH-UHFFFAOYSA-N 16-Epiaffinine Natural products C1C(C2=CC=CC=C2N2)=C2C(=O)CC2C(=CC)CN(C)C1C2CO PXFBZOLANLWPMH-UHFFFAOYSA-N 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000009395 breeding Methods 0.000 description 1
- 230000001488 breeding effect Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 201000010099 disease Diseases 0.000 description 1
- 208000037265 diseases, disorders, signs and symptoms Diseases 0.000 description 1
- 238000011156 evaluation Methods 0.000 description 1
- 230000036039 immunity Effects 0.000 description 1
- 230000003053 immunization Effects 0.000 description 1
- 238000002649 immunization Methods 0.000 description 1
- 230000007233 immunological mechanism Effects 0.000 description 1
- 230000010365 information processing Effects 0.000 description 1
- 230000009545 invasion Effects 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000003860 storage Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
- G05D1/00—Control of position, course, altitude or attitude of land, water, air or space vehicles, e.g. using automatic pilots
- G05D1/10—Simultaneous control of position or course in three dimensions
- G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The method for solving of the invention discloses a kind of multiple no-manned plane Task Allocation Problem based on immune optimization algorithm.Steps are as follows by the present invention: 1: the mathematical modeling of problem, it is assumed that have the consistent unmanned plane of a collection of specifications and models in airport, be now to distributing to whole tasks into whole unmanned planes in airport；One task can only be accessed by a unmanned plane, but a unmanned plane is able to access that multiple tasks；Under the premise of whole tasks are accessed, the smallest multiple no-manned plane allocation plan of path cost is provided；2: the dimensionality reduction of variable and distribution in particular problem；3: the optimization of objective function and constraint condition, due to giving the coordinate between airport and task and completing the distribution of variable, therefore the distance between airport and task, task and task can be calculated, to specifically give and be optimized objective function and constraint condition.The known variables that the present invention generates after modeling to multiple no-manned plane Task Allocation Problem are iterated optimization, to obtain optimal solution.
Description
Technical field
The present invention relates to air vehicle technique field more particularly to a kind of method for solving of multiple no-manned plane Task Allocation Problem.
Technical background
Unmanned plane is commonly referred to as the UAV manipulated using radio robot and presetting apparatus,
English abbreviation is UAV (Unmanned Aerial Vehicle).With it is traditional it is manned drive an airplane compared with, unmanned plane is often more
It is suitble to the higher task of those danger coefficients, it is ensured that the sacrifice of aircrew is avoided while high quality completes task.
In addition, unmanned plane relies on the advantages such as degree of safety height, compact, cheap, extensive welcome is received in numerous areas.
With the progress in epoch and the development of science and technology, unmanned plane problems faced becomes increasingly complex.But due to
Itself flying distance and the limitation for handling information capability, a complicated problem just need at multiple UAVs collaboration simultaneously
Reason, to improve the efficiency of task execution.
Inspiration of the scientists by Immune System, based on human immunity theory, herein on develop
A kind of intelligent algorithm come, the algorithm are known as Artificial Immune Algorithm.
The scientists in the U.S. were taken the lead in the dynamic model that immune system is constructed based on immune network theory in 1986, most
The research to immune system is early started.They generate and maintain machine according to the Biodiversity Characteristics and antibody of Immune System
System, solves the problems, such as often occur and annoying for a long time " precocity " of scientist in general searching process, finally asks
Obtain globally optimal solution.
Since Artificial Immune Algorithm can seek globally optimal solution rather than locally optimal solution, rely on this advantage, science
Family Artificial Immune Algorithm and genetic algorithm are merged so that fused algorithm had both remained the characteristic of genetic algorithm,
The adaptive characteristic that immune algorithm solves multi-objective problem is remained again, can largely can be showed to avoid " precocity "
As enabling algorithm rapidly to converge at global optimum.
Immune System has very strong learning and memory and adaptive adjustment capability, possesses the external cause of disease of perfect resistance
The mechanism of body invasion is a kind of system for capableing of adaptive learning.Its powerful information processing capability possessed can be parallel
Complicated calculations are completed in the case where distribution, therefore, multiple no-manned plane Task Allocation Problem is solved using Artificial Immune Algorithm has been
It is complete feasible.
Summary of the invention
The present invention proposes a kind of method for solving of multiple no-manned plane Task Allocation Problem based on immune algorithm, and this method is passing
The operation such as antibody concentration adjustment mechanism and selection, intersection and variation in genetic algorithm is added in the immune algorithm of system, it is right
Traditional immune algorithm is improved.This method can be used to solve the optimal solution of multiple no-manned plane Task Allocation Problem.
Multiple no-manned plane Task Allocation Problem proposed by the present invention can be described as having a collection of model, combat radius etc. in airport
Parameter unmanned plane all the same and multiple the needing to access of the tasks, are now to distributing to whole tasks into whole nothings in airport
It is man-machine.Assuming that only needing a unmanned plane that the task of its current accessed can be completed, i.e., a task can only be by a unmanned plane institute
Access, but the accessible multiple tasks of unmanned plane.Under the premise of whole tasks are accessed, flying distance minimum is provided
Unmanned plane task allocation plan.
This method specifically includes the following steps:
Step 1: the mathematical modeling of problem, it is assumed that have the consistent unmanned plane of a collection of specifications and models, these nothings in airport
It is man-machine to need to access multiple and different tasks.It is now to distributing to whole tasks into whole unmanned planes in airport.Assuming that only needing
The task of its current accessed can be completed in one unmanned plane, i.e., a task can only be accessed by a unmanned plane, but a nothing
It is man-machine to be able to access that multiple tasks.Under the premise of whole tasks are accessed, it is minimum (i.e. apart from optimal) to provide path cost
Multiple no-manned plane allocation plan.
Based on apart from optimal multiple no-manned plane Task Allocation Problem, mathematical modeling is considered as objective function and all constraints
Condition.The objective function of Task Allocation Problem is as follows:
Whole constraint conditions of Task Allocation Problem are as follows:
IfThen
Wherein, formula (1) is the objective function of multiple no-manned plane Task Allocation Problem, and formula (2)~formula (10) is all to constrain item
Part.A task can also be at most visited again after one task of access of formula (2) expression unmanned plane, formula (3) indicates the visit of unmanned plane
A task can only at most be accessed by asking before a task, therefore formula (2) and formula (3) have codetermined a unmanned plane at most only
Two tasks can be accessed；Formula (4) indicates that accessing for task is unable to repeated accesses, i.e., when a unmanned plane has accessed a task again
After accessing b task, a task cannot be visited again；Formula (5) indicates a case where unmanned plane only accesses a task, i.e. unmanned plane
After accessing a task and returning to airport, other tasks cannot be visited again；Formula (6) indicates that identical task cannot repeat to visit
It asks, i.e., cannot rest on predecessor's business after a unmanned plane has accessed a task, it is necessary to access next task or return
Airport；Formula (7) indicates that all unmanned planes must set out；Formula (8) indicates that all unmanned planes must return to；Formula (9) indicates institute
Some tasks must be all accessed；Formula (10) indicates variable xi jkCan only value 1 or 0, indicate unmanned plane whether access the task,
The x if unmanned plane accessi jk=1, otherwise xi jk=0.
Wherein, the correlated variables in mathematical model is described as follows: D indicates total distance；M indicates the number of units of unmanned plane, n-1 table
Show the number of task；Indicate the i-th frame unmanned plane whether from airport access k location；Indicate that the i-th frame unmanned plane is accessing
Whether continue to access k location behind the complete position j；Indicate whether the i-th frame unmanned plane returns to airport after having accessed the position j；d1k
Indicate the distance between airport and k location；djkIndicate the distance between the position j and k location；dj1It indicates between the position j and airport
Distance；
Step 2: the dimensionality reduction of variable and distribution in particular problem, the specific steps are as follows:
The it is proposed of 2-1. particular problem.In the algorithm, chromosome is usually the one-dimension array being made of 0 and 1, it is therefore desirable to
Dimension-reduction treatment is carried out to the variable in model.M=2, n=4 are taken now, that is, there are 2 unmanned planes to go 3 tasks of access.Simultaneously will
Airport and 3 tasks are marked on plane rectangular coordinates figure, and 2 unmanned planes are located at airport (0,0), task one, two, three
It is located at (10,20), (30,20) and (40,10) of rectangular coordinate system；
The matrix description of 2-2. solution.The task distribution feelings of 2 unmanned planes are described with the matrix T1 and T2 of 2 4*4 now
Condition, wherein corresponding line number is the departure place of unmanned plane if certain element is 1 in matrix, corresponding columns is destination, such as formula
(11) shown in.
Wherein, the element that the 1st row the 4th arranges in T1 matrix is 1, illustrates First unmanned plane from airport access task
Three；Similarly the element of the 4th row the 1st column is 1, illustrates that First unmanned plane accesses airport from task threes, therefore T1 matrix description
The 1st unmanned plane during flying track are as follows: airport-three-airport of task.And so on, the 2nd unmanned plane during flying track are as follows: machine
One-task of field-task, two-airport.I.e. 2 unmanned planes access whole task that is over；
The distribution of 2-3. variable.The solution of problem is converted to and asks matrix T1 and T2, comprehensively considers constraint equation (6), i.e.,
The elements in a main diagonal of matrix is all 0, then dimensionality reduction and distributes the element in matrix again, so that it be made to meet in algorithm
Individual is usually the requirement of one-dimension array.Shown in allocation result such as formula (12).
The one-dimension array that thus variables transformations in problem are made of at one 24 variables, while according to constraint
The requirement of conditional (10), i.e. variable can only in 0 and 1 value, this is also fully consistent with feature individual in algorithm.
Step 3: the optimization of objective function and constraint condition, due to giving the coordinate between airport and task and completion
The distribution of variable, therefore the distance between airport and task, task and task can be calculated, thus by objective function peace treaty
Beam condition specifically gives and is optimized.Formula (13) is the result after objective function optimization.
D=22.36* (x1+x4+x13+x16)+36.06* (x2+x7+x14+x19)+41.23* (x3+x10+x15+x22)+
20* (x5+x8+x17+x20)+31.62* (x6+x11+x18+x23)+14.14* (x9+x12+x21+x24) (13)
Formula (14)~formula (19) is the result after constraint condition optimization.
Wherein, a task can also at most be visited again after one task of access of formula (14) expression unmanned plane；Formula (15) table
A task can only at most be accessed by showing before one task of access of unmanned plane；Formula (16) indicates that accessing for task cannot repeat
Access；Formula (17) indicates that all unmanned planes must set out and return；Formula (18) indicates that a unmanned plane only accesses a task
The case where；Formula (19) indicates that all tasks must be all accessed.This completes become whole constraint condition with 24
Amount is indicated, convenient for being solved with algorithm to problem.
Step 4: antigen recognizing and initial antibodies generate, by required objective function and constraint condition as antigen into
Row identification, judges whether to solve the problems, such as this once with this.During generating initial antibodies, by unmanned plane model in x1
~x24This 24 variables are considered as an antibody, and the process that variable is initialized then corresponds to be solved in algorithm initial
Value.By the identification of antibody and antigen, if once solving the problems, such as this, so that it may corresponding solution is directly found, to generate
Initial antibodies.If not solving the problems, such as this, the initial antibodies population p that population scale is n is randomly generated, wherein p=
{p1, p2..., pn, for each antibody, all it is made of m genes, it can be expressed as Agi={ ag1, ag2..., agm}(i
∈ [1, n]).
Step 5: immunological memory.The high antibody of a part of affinity is saved in data base, specific step is to utilize
Immune ruleIt is saved in data base from δ * n outstanding antibody are chosen in antibody population.Generate the specific rule of data base
It is then as follows:
Step 6: calculating affinity and diversity.The tool for generally calculating affinity in the algorithm mainly has following three kinds:
Hamming distance from ( )
Euclidean distance
Manhattan distance
After affinity of antibody has been calculated, a set Aff being made of whole affinity is generated, wherein Aff=
{aff1, aff2... affn, affinity size is then generated into a new affinity set by sequence:
Rank (Aff)=rank ({ aff1, aff2... affn)={ Ab '1, Ab '2..., Ab 'ε*n}
Step 7: Immune Clone Selection.According to the size of each individual affinity, Immune Clone Selection rule is usedTo its carry out gram
Grand selection, Immune Clone Selection rule are as follows:
Meanwhile a select probability set p, i.e. p={ p are generated in Immune Clone Selection1, p2..., pn}.This selection is general
Rate set { p1, p2..., pnWith selection after population { Ab '1, Ab '2..., Ab 'nIt is one-to-one, that is to say, that it is affine
The probability that the bigger individual of power is selected is bigger.
Step 8: the promotion and inhibition of antibody.Since in traditional Immune Clone Selection operation, select probability and affinity are big
Small directly proportional, this diversity that this may result in population at individual is reduced, so that optimal solution be made to fall into local optimum rather than complete
Office is optimal.Therefore concentration adjustment mechanism is introduced in improved Artificial Immune Algorithm, the expectation breeding for calculating each individual is general
Rate.The formula for calculating expectation reproductive probability is as follows:
Wherein, α is constant.By above formula as it can be seen that the affinity of individual is higher, then its corresponding expectation reproductive probability is got over
Greatly；The concentration of individual is bigger, then it is expected reproductive probability with regard to smaller.Affinity height and concentration can with maximum probability be selected in this way
Low individual, so that it is guaranteed that the diversity of population.
Step 9: clonal expansion.For into the outstanding antibody crossing Immune Clone Selection step and promotion and generating after inhibiting,
Use clonal expansion ruleClonal expansion is carried out to them, wherein the clone sizes of each antibody are numerous by its corresponding expectation
Probability decision is grown, i.e. the expectation reproductive probability of antibody is bigger, then corresponding clone sizes are bigger.Specific clonal expansion ruleIt is as follows:
Wherein Ag 'i=jAg 'i, (i ∈ { 1,2 ..., n }, j ∈ { p1, p2..., pn*ε})。
Step 10: genetic manipulation.So-called genetic manipulation is exactly by selection, intersects, the operation of variation generation new population
Process.In the algorithm, individual is selected by calculating the expectation reproductive probability of individual, then passes through behaviour of intersecting and make a variation
Make to generate antibody of new generation, can guarantee that population is developed to the high direction of fitness in this way.
In selection operation, select probability p is as follows:
N outstanding antibody of ε *, i.e. { Ag ' are selected according to Immune Clone Selection rule1, Ag '2..., Ag 'ε * n}；
In mutation operation, high frequency closedown mechanism is used.N- ε * n antibody lower for affinity, uses height
Frequency makes a variation regularCarry out high frequency closedown to it, i.e. the probability that makes a variation of the lower antibody of affinity is bigger.It is specific high
Frequency makes a variation regularIt is as follows:
The present invention has the beneficial effect that:
The present invention is using the immunologic mechanism in Artificial Immune Algorithm, to what is generated after the modeling of multiple no-manned plane Task Allocation Problem
Known variables are iterated optimization, to obtain optimal solution.
The behaviour of antibody concentration adjustment mechanism and traditional genetic algorithm is added in the present invention in traditional Artificial Immune Algorithm
Make, in the case where obtaining identical fitness value, compared with traditional immune algorithm, can more be had using improved immune algorithm
The optimal solution for finding problem of effect.
By traditional algorithm and modified hydrothermal process respectively 2.40GHz, memory 4GB, 64 Windows operating systems meter
On calculation machine, emulation solution is carried out using MATLAB R2014a software.Traditional immunization algorithm and improved immune algorithm are carried out
150 iteration, wherein population scale is 50, and memory storage capacity is 4, crossover probability 0.75, mutation probability 0.01, diversity
Evaluation parameter is 0.95.After 150 iteration, improved Artificial Immune Algorithm has found most when the 4th iteration
Excellent solution, corresponding fitness value are 136.15.And the number that traditional immune algorithm obtains identical fitness value is 42.
Detailed description of the invention
Fig. 1 is unmanned plane task distribution schematic diagram；
Fig. 2 is the position view on airport and task in particular problem；
Fig. 3 is immune algorithm flow chart；
Fig. 4 is MATLAB result analogous diagram.
Specific embodiment
Below according to attached drawing, the present invention will be described in detail, and the objects and effects of the present invention will be more apparent.
Fig. 1 is unmanned plane task distribution schematic diagram.There are two unmanned planes UAV1 and UAV2 in airport, needs to access five and appoint
Business, same task can only be accessed by a unmanned plane, one or more accessible task of a unmanned plane, in whole
After task all accesses, all unmanned planes must return to airport.
Fig. 2 is the position view on airport and task in particular problem.Wherein 2 unmanned planes are located at airport (0,0), appoint
Business one, two, three is located at (10,20), (30,20) and (40,10) of rectangular coordinate system.
Fig. 3 is immune algorithm flow chart.Immune algorithm carries out 150 iteration, and wherein population scale is 50, remembers storage capacity
It is 4, crossover probability 0.75, mutation probability 0.01, Diversity parameter is 0.95.
Fig. 4 is MATLAB result analogous diagram.By traditional algorithm and modified hydrothermal process respectively in 2.40GHz, memory 4GB, 64
On the computer of the Windows operating system of position, emulation solution is carried out using MATLAB R2014a software.
Claims (5)
1. the method for solving of the multiple no-manned plane Task Allocation Problem based on immune optimization algorithm, it is characterised in that including walking as follows
It is rapid:
Step 1: the mathematical modeling of problem, it is assumed that have the consistent unmanned plane of a collection of specifications and models, these unmanned planes in airport
Need to access multiple and different tasks；It is now to distributing to whole tasks into whole unmanned planes in airport；Assuming that only needing one
The task of its current accessed can be completed in unmanned plane, i.e., a task can only be accessed by a unmanned plane, but a unmanned plane
It is able to access that multiple tasks；Under the premise of whole tasks are accessed, the smallest multiple no-manned plane distribution side of path cost is provided
Case；
Based on apart from optimal multiple no-manned plane Task Allocation Problem, mathematical modeling is considered as objective function and all constraint items
Part；The objective function of Task Allocation Problem is as follows:
Whole constraint conditions of Task Allocation Problem are as follows:
IfThen
Wherein, formula (1) is the objective function of multiple no-manned plane Task Allocation Problem, and formula (2)~formula (10) is whole constraint conditions；Formula
(2) task can also be at most visited again after indicating one task of access of unmanned plane, formula (3) indicates access one of unmanned plane
A task can only be at most accessed before task, therefore formula (2) and formula (3) have codetermined a unmanned plane and can only at most access
Two tasks；Formula (4) indicates that accessing for task is unable to repeated accesses, i.e., visits again b when a unmanned plane has accessed a task
After business, a task cannot be visited again；Formula (5) indicates a case where unmanned plane only accesses a task, i.e. unmanned plane accesses one
Task and after returning to airport, cannot visit again other tasks；Formula (6) indicates that identical task is unable to repeated accesses, i.e., when one
Platform unmanned plane cannot rest on predecessor's business after having accessed a task, it is necessary to access next task or return to airport；Formula (7)
Indicate that all unmanned planes must set out；Formula (8) indicates that all unmanned planes must return to；All tasks of formula (9) expression are all
It must be accessed；Formula (10) indicates variable xi jkCan only value 1 or 0, indicate unmanned plane whether access the task, if unmanned plane visit
Ask then xi jk=1, otherwise xi jk=0；
Wherein, the correlated variables in mathematical model is described as follows: D indicates total distance；M indicates the number of units of unmanned plane, and n-1 indicates to appoint
The number of business；Indicate the i-th frame unmanned plane whether from airport access k location；Indicate that the i-th frame unmanned plane is accessing j
It postpones and whether continues to access k location；Indicate whether the i-th frame unmanned plane returns to airport after having accessed the position j；d1kExpression machine
The distance between field and k location；djkIndicate the distance between the position j and k location；dj1Indicate between the position j and airport away from
From；
Step 2: the dimensionality reduction of variable and distribution in particular problem；
Step 3: the optimization of objective function and constraint condition, due to giving the coordinate between airport and task and completing change
The distribution of amount, therefore the distance between airport and task, task and task can be calculated, thus by objective function and constraint item
Part specifically gives and is optimized.
2. the method for solving of the multiple no-manned plane Task Allocation Problem according to claim 1 based on immune optimization algorithm,
It is characterized in that the dimensionality reduction of variable and distribution in particular problem described in step 2, is implemented as follows:
The it is proposed of 2-1. particular problem；In the algorithm, chromosome is usually the one-dimension array being made of 0 and 1, it is therefore desirable to mould
Variable in type carries out dimension-reduction treatment；M=2, n=4 are taken now, that is, there are 2 unmanned planes to go 3 tasks of access；Simultaneously by airport
It is marked on plane rectangular coordinates figure with 3 tasks, 2 unmanned planes are located at airport (0,0), and task one, two, three is distinguished
At (10,20), (30,20) and (40,10) of rectangular coordinate system；
The matrix description of 2-2. solution；The task distribution condition of 2 unmanned planes is described with the matrix T1 and T2 of 2 4*4 now,
If certain element is 1 in matrix in, and corresponding line number is the departure place of unmanned plane, and corresponding columns is destination, such as formula (11)
It is shown；
Wherein, the element that the 1st row the 4th arranges in T1 matrix is 1, illustrates First unmanned plane from airport access task three；Together
The element for managing the 4th row the 1st column is 1, illustrates that First unmanned plane accesses airport from task threes, therefore the 1st of T1 matrix description the
Platform unmanned plane during flying track are as follows: airport-three-airport of task；And so on, the 2nd unmanned plane during flying track are as follows: appoint on airport-
One-two-airport of task of business；I.e. 2 unmanned planes access whole task that is over；
The distribution of 2-3. variable；The solution of problem is converted to and asks matrix T1 and T2, comprehensively considers constraint equation (6), i.e. matrix
The elements in a main diagonal be all 0, then the element in matrix dimensionality reduction and is distributed again, so that it is individual in algorithm to meet it
The usually requirement of one-dimension array；Shown in allocation result such as formula (12)；
The one-dimension array that thus variables transformations in problem are made of at one 24 variables, while according to constraint condition
The requirement of formula (10), i.e. variable can only in 0 and 1 value, this is also fully consistent with feature individual in algorithm.
3. the method for solving of the multiple no-manned plane Task Allocation Problem according to claim 2 based on immune optimization algorithm,
It is characterized in that specifically optimization is accomplished by step 3
Formula (13) is the result after objective function optimization；
D=22.36* (x1+x4+x13+x16)+36.06* (x2+x7+x14+x19)+41.23* (x3+x10+x15+x22)+20* (x5
+x8+x17+x20)+31.62* (x6+x11+x18+x23)+14.14* (xIt pushes away+x12+x21+x24) (13)
Formula (14)~formula (19) is the result after constraint condition optimization；
Wherein, a task can also at most be visited again after one task of access of formula (14) expression unmanned plane；Formula (15) indicate without
A task can only be at most accessed before man-machine one task of access；Formula (16) indicates that accessing for task cannot repeat to visit
It asks；Formula (17) indicates that all unmanned planes must set out and return；Formula (18) indicates that a unmanned plane only accesses a task
Situation；Formula (19) indicates that all tasks must be all accessed；This completes by whole 24 variables of constraint condition
It is indicated, is applied to 24 variables as initial value based in immune optimization algorithm, it is regular to obtain high frequency closedown.
4. the method for solving of the multiple no-manned plane Task Allocation Problem according to claim 3 based on immune optimization algorithm,
Be characterized in that it is characterized in that based on immune optimization algorithm the specific implementation process is as follows:
Step (1): antigen recognizing and initial antibodies generate, and required objective function and constraint condition are carried out as antigen
Identification, judges whether to solve the problems, such as this once with this；
Step (2): immunological memory；The high antibody of a part of affinity is saved in data base, immune rule is utilizedFrom
δ * n outstanding antibody are chosen in antibody population to be saved in data base；
Step (3): affinity and diversity are calculated；
Step (4): Immune Clone Selection；According to the size of each individual affinity, Immune Clone Selection rule is usedIt is cloned
Selection；
Step (5): the promotion and inhibition of antibody, therefore concentration adjustment mechanism is introduced in improved Artificial Immune Algorithm, it counts
Calculate the expectation reproductive probability of each individual；
Step (6): clonal expansion makes the outstanding antibody generated by Immune Clone Selection step and promotion and after inhibiting
With clonal expansion ruleClonal expansion is carried out to them, wherein the clone sizes of each antibody are bred by its corresponding expectation
Probability determines that the expectation reproductive probability of i.e. antibody is bigger, then corresponding clone sizes are bigger；
Step (7): genetic manipulation, by calculate individual expectation reproductive probability individual is selected, then by intersect and
Mutation operation generates antibody of new generation, can guarantee that population is developed to the high direction of fitness in this way.
5. the method for solving of the multiple no-manned plane Task Allocation Problem according to claim 4 based on immune optimization algorithm,
It is characterized in that it is characterized in that being implemented as follows based on immune optimization algorithm:
Antigen recognizing described in step (1) and initial antibodies generate, specific as follows:
In initial antibodies generation step, using 24 variables as initial value, by the identification of antibody and antigen, if once solved
Certainly cross the problem, so that it may corresponding memory cell is directly found, to generate initial antibodies；If not solved this to ask
The initial antibodies population p that population scale is n is then randomly generated, wherein p={ p in topic1, p2..., pn, for each antibody, all by
M gene compositions, it can be expressed as Agi={ ag1, ag2..., agm(i ∈ [1, n])；
The generation specific rules of data base described in step (2) are as follows:
Calculating affinity and diversity described in step (3), tool mainly have following three kinds:
Hamming distance from
Euclidean distance
Manhattan distance
After affinity of antibody has been calculated, a set Aff being made of whole affinity is generated, wherein Aff={ aff1,
aff2... affn, affinity size is then generated into a new affinity set by sequence:
Rank (Aff)=rank ({ aff1, aff2... affn)={ Ab '1, Ab '2..., Ab 'ε*n}
The rule of Immune Clone Selection described in step (4) is as follows:
Meanwhile a select probability set p, i.e. p={ p are generated in Immune Clone Selection1, p2..., pn}；This select probability collection
Close { p1, p2..., pnWith selection after population { Ab '1, Ab '2..., Ab 'nIt is one-to-one, that is to say, that affinity is got over
The probability that big individual is selected is bigger；
The promotion and inhibition of antibody described in step (5) introduce concentration adjustment mechanism in Artificial Immune Algorithm, calculate each
The expectation reproductive probability of individual；The formula for calculating expectation reproductive probability is as follows:
Wherein, α is constant；By above formula as it can be seen that the affinity of individual is higher, then its corresponding expectation reproductive probability is bigger；It is a
The concentration of body is bigger, then it is expected reproductive probability with regard to smaller；Affinity height can with maximum probability be selected in this way and concentration is low
Individual, so that it is guaranteed that the diversity of population；
Clonal expansion described in step (6), specific clonal expansion ruleIt is as follows:
Wherein
Genetic manipulation described in step (7)；
In selection operation, select probability p is as follows:
N outstanding antibody of ε * are selected according to Immune Clone Selection rule, i.e.,
In mutation operation, high frequency closedown mechanism is used；N- ε * n antibody lower for affinity, is become using high frequency
Different ruleCarry out high frequency closedown to it, i.e. the probability that makes a variation of the lower antibody of affinity is bigger；Specific high frequency becomes
Different ruleIt is as follows:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910520241.1A CN110362107A (en) | 2019-06-17 | 2019-06-17 | The method for solving of multiple no-manned plane Task Allocation Problem based on immune optimization algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910520241.1A CN110362107A (en) | 2019-06-17 | 2019-06-17 | The method for solving of multiple no-manned plane Task Allocation Problem based on immune optimization algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110362107A true CN110362107A (en) | 2019-10-22 |
Family
ID=68217330
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910520241.1A Withdrawn CN110362107A (en) | 2019-06-17 | 2019-06-17 | The method for solving of multiple no-manned plane Task Allocation Problem based on immune optimization algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110362107A (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111337042A (en) * | 2020-03-13 | 2020-06-26 | 湖北大学 | Vehicle path planning method and system |
| CN111366169A (en) * | 2020-04-09 | 2020-07-03 | 湖南工学院 | Deterministic mobile robot path planning method |
| CN111414005A (en) * | 2020-03-11 | 2020-07-14 | 五邑大学 | Task dispatching decision method and system for unmanned aerial vehicle cluster |
| CN112633654A (en) * | 2020-12-15 | 2021-04-09 | 哈尔滨工程大学 | Multi-unmanned aerial vehicle task allocation method based on improved cluster expansion consistency bundle algorithm |
| CN113487221A (en) * | 2021-07-23 | 2021-10-08 | 中南大学 | Space-sky heterogeneous earth observation resource cooperative scheduling method for dynamic target observation |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105825293A (en) * | 2016-03-10 | 2016-08-03 | 桂林电子科技大学 | Fruit-fly-optimization-algorithm-based multi-station assembling sequence planning method |
| CN107168054A (en) * | 2017-05-10 | 2017-09-15 | 沈阳工业大学 | Multi-robotic task is distributed and paths planning method |
| CN109670689A (en) * | 2018-12-06 | 2019-04-23 | 杭州电子科技大学 | A kind of scientific workflow Multiobjective Scheduling method based on immunity particle cluster |
-
2019
- 2019-06-17 CN CN201910520241.1A patent/CN110362107A/en not_active Withdrawn
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105825293A (en) * | 2016-03-10 | 2016-08-03 | 桂林电子科技大学 | Fruit-fly-optimization-algorithm-based multi-station assembling sequence planning method |
| CN107168054A (en) * | 2017-05-10 | 2017-09-15 | 沈阳工业大学 | Multi-robotic task is distributed and paths planning method |
| CN109670689A (en) * | 2018-12-06 | 2019-04-23 | 杭州电子科技大学 | A kind of scientific workflow Multiobjective Scheduling method based on immunity particle cluster |
Non-Patent Citations (3)
| Title |
|---|
| 武装: "《几种改进的智能优化算法及其应用》", 31 August 2018, 北京：科学技术文献出版社 * |
| 沈小云: "基于人工免疫算法的多无人机任务分配研究", 《中国优秀博硕士学位论文全文数据库(硕士)工程科技Ⅱ辑》 * |
| 田玉玲，段富: "《免疫优化算法、模型及应用》", 28 February 2013 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111414005A (en) * | 2020-03-11 | 2020-07-14 | 五邑大学 | Task dispatching decision method and system for unmanned aerial vehicle cluster |
| CN111337042A (en) * | 2020-03-13 | 2020-06-26 | 湖北大学 | Vehicle path planning method and system |
| CN111337042B (en) * | 2020-03-13 | 2021-11-02 | 湖北大学 | Vehicle path planning method and system |
| CN111366169A (en) * | 2020-04-09 | 2020-07-03 | 湖南工学院 | Deterministic mobile robot path planning method |
| CN112633654A (en) * | 2020-12-15 | 2021-04-09 | 哈尔滨工程大学 | Multi-unmanned aerial vehicle task allocation method based on improved cluster expansion consistency bundle algorithm |
| CN113487221A (en) * | 2021-07-23 | 2021-10-08 | 中南大学 | Space-sky heterogeneous earth observation resource cooperative scheduling method for dynamic target observation |
| CN113487221B (en) * | 2021-07-23 | 2022-07-15 | 中南大学 | Space-sky heterogeneous earth observation resource cooperative scheduling method for dynamic target observation |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110362107A (en) | The method for solving of multiple no-manned plane Task Allocation Problem based on immune optimization algorithm | |
| Obayashi et al. | Multi-objective design exploration for aerodynamic configurations | |
| CA3085897A1 (en) | Evolutionary architectures for evolution of deep neural networks | |
| CA3151805A1 (en) | Automated path-based recommendation for risk mitigation | |
| Tawhid et al. | A hybrid social spider optimization and genetic algorithm for minimizing molecular potential energy function | |
| Hajipour et al. | A novel Pareto-based multi-objective vibration damping optimization algorithm to solve multi-objective optimization problems | |
| CN109583667B (en) | Cable path solving and optimizing method | |
| Pilania et al. | Mobile manipulator planning under uncertainty in unknown environments | |
| CN115755954B (en) | Routing inspection path planning method, system, computer equipment and storage medium | |
| US20210150085A1 (en) | Generative design techniques for multi-family housing projects | |
| Lin et al. | Object point cloud classification via poly-convolutional architecture search | |
| Zhao et al. | A new cellular manufacturing layout: Multi-floor linear cellular manufacturing layout | |
| Hani et al. | Simulation based optimization of a train maintenance facility | |
| CN108537253A (en) | A kind of adaptive semi-supervised dimension reduction method constrained in pairs based on probability | |
| Wu et al. | Heuristic position allocation methods for forming multiple UAV formations | |
| US11055655B1 (en) | Method and system for generating an alimentary combination help update | |
| US20220036234A1 (en) | Method of and system for determining physical transfer interchange nodes | |
| KR102614099B1 (en) | Method for distributing work points to plural task-performing robots | |
| CN110083674B (en) | Intellectual property information processing method and device | |
| CN106709572B (en) | A kind of data processing method and equipment | |
| CN114650321A (en) | Task scheduling method for edge computing and edge computing terminal | |
| CN116978450A (en) | Protein data processing method, device, electronic equipment and storage medium | |
| Feng et al. | A clustering algorithm based on emotional preference and migratory behavior | |
| Saini et al. | Automatic evolution of bi-clusters from microarray data using self-organized multi-objective evolutionary algorithm | |
| US20220390918A1 (en) | Methods and systems for selection of manufacturing orientation using machine learning |
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 | ||
| WW01 | Invention patent application withdrawn after publication | ||
| WW01 | Invention patent application withdrawn after publication |
Application publication date: 20191022 |