CN109269502A - A kind of no-manned plane three-dimensional Route planner based on more stragetic innovation particle swarm algorithms - Google Patents
A kind of no-manned plane three-dimensional Route planner based on more stragetic innovation particle swarm algorithms Download PDFInfo
- Publication number
- CN109269502A CN109269502A CN201811091779.7A CN201811091779A CN109269502A CN 109269502 A CN109269502 A CN 109269502A CN 201811091779 A CN201811091779 A CN 201811091779A CN 109269502 A CN109269502 A CN 109269502A
- Authority
- CN
- China
- Prior art keywords
- algorithm
- particle
- particle swarm
- population
- compared
- 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- Withdrawn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
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/20—Instruments for performing navigational calculations
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01C—MEASURING DISTANCES, LEVELS OR BEARINGS; SURVEYING; NAVIGATION; GYROSCOPIC INSTRUMENTS; PHOTOGRAMMETRY OR VIDEOGRAMMETRY
- G01C21/00—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00
- G01C21/10—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration
- G01C21/12—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning
- G01C21/16—Navigation; Navigational instruments not provided for in groups G01C1/00 - G01C19/00 by using measurements of speed or acceleration executed aboard the object being navigated; Dead reckoning by integrating acceleration or speed, i.e. inertial navigation
Landscapes
- Engineering & Computer Science (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Automation & Control Theory (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The present invention relates to a kind of Path Planning for UAV, and in particular to a kind of no-manned plane three-dimensional Route planner based on more stragetic innovation particle swarm algorithms.Simulation modeling is carried out for two kinds of real terrains of undulating topography and alpine terrain, inertia weight based on linear decrease is introduced to particle swarm algorithm, with the Logistic mapping function in chaos strategy, and a kind of new improved IPSO algorithm is proposed using survival of the fittest strategy to the low individual of the adaptive value in searching process.The experimental results showed that the planning path length of the IPSO algorithm proposed shortens 5.2% compared with AWPIO under undulating topography, shorten 5.9% compared with PSO;The time loss of IPSO algorithm shortens 5.25% compared with AWPIO, shortens 4.15% compared with PSO;Under alpine terrain, the planning path length of the IPSO algorithm proposed shortens 1.3% compared with AWPIO, shortens 2.69% compared with PSO;The time loss of IPSO algorithm shortens 24.3% compared with AWPIO, shortens 26.3% compared with PSO.Therefore, IPSO is a kind of intelligent algorithm for being more suitable for routeing.
Description
Technical field
The invention belongs to field of intelligent control technology, are related to a kind of Path Planning for UAV, and in particular to Yi Zhongji
In the no-manned plane three-dimensional Route planner of more stragetic innovation particle swarm algorithms.
Background technique
Path Planning for Unmanned Aircraft Vehicle is mainly used in two kinds of environment: two-dimensional environment and three-dimensional environment.Boat in two-dimensional environment
Circuit planning is mainly used in small area, and there is no larger changes for flying height during executing entire task for unmanned plane
Change;Routeing in three-dimensional environment is mainly used in relatively large regional scope, and unmanned plane is in the process for executing entire task
In to consider height above sea level problem, exist largely climb, step-down operation, need to consider the flying height of unmanned plane to energy at this time
The influence of source consumption.
2014, Qiang Wang et al. proposed particle swarm optimization algorithm in no-manned plane three-dimensional environment path planning problem
Three improvement strategies.First, the rand function in canonical algorithm is generated into a random number, is changed to generate a random number square
Battle array, can guarantee that the learning probability of the particle in each dimension is reliable in this way.Second, occur crossing the border when particle individual
When situation, its range is limited using two random numbers.Third introduces asynchronous training operator to promote ability of searching optimum.
2016, Yao Peng et al. proposed a kind of mixing air route based on improved interference fluid dynamics system and grey wolf optimum theory
Planning algorithm, for the three-dimensional path planning problem of unmanned plane under complicated landform environment.Establish a kind of improved disturbance fluid
Dynamic system mathematical model obtains disturbance flow field by correcting initial point.The streamline in flow field can be regarded as planning road
Line, it is possible to prevente effectively from stationary point and local trap.By simulating the hierarchical structure and Forging strategy of wolf pack, individual memory function is introduced
It can optimize barrier response coefficient with survival of the fittest rule, cook up the three dimension pathline that can smoothly fly.2016, Gai-
Ge Wang et al. publishes thesis " Three-dimensional path planning for UACV using an impoved
Bat algorithm " handles routeing in three-dimensional war environment using improving bat optimization algorithm for the first time in text, and
And the air route cooked up is smoothed using B-spline algorithm, by improve bat optimization algorithm and evolution algorithm into
Row comparative analysis show that the secure path for avoiding obstacle can effectively be cooked up by improving bat algorithm.
The presence all more or less when carrying out routeing of original Path Planning for Unmanned Aircraft Vehicle algorithm falls into local optimum
The problem of solution, leading to unmanned plane is not to navigate by water according to optimal route in entire flight course.Herein in three-dimensional environment
Path Planning for Unmanned Aircraft Vehicle propose more stragetic innovation particle swarm algorithms, promote path planning efficiency in three-dimensional environment.
Summary of the invention
It is different according to the complexity of terrain environment, it should to select during solving practical solution Path Planning for Unmanned Aircraft Vehicle
The algorithm for meeting environmental requirement mission requirements is selected, existing algorithm is to carry out a mathematics to landform when carrying out emulation experiment
The foundation of model, does not under various circumstances test algorithm.Swarm intelligence algorithm is solving the path rule in three-dimensional space
When the problem of drawing, time complexity and space complexity are all different, and generally existing to fall into local optimum, convergence rate is low, algorithm
Stability it is not high enough the problems such as.Therefore, simulation modeling is carried out herein for two kinds of true environment data, and is directed to landform
Actual conditions more stragetic innovations are carried out to particle swarm algorithm.
It is an object of the invention to overcome problem above of the existing technology, provide a kind of based on more stragetic innovation particles
The no-manned plane three-dimensional Route planner of group's algorithm.
To realize above-mentioned technical purpose and the technique effect, the invention is realized by the following technical scheme: step 1:
Necessary parameter during setting particle swarm optimization algorithm, specifically includes that inertia weight range, self-teaching factor range, society
Meeting Studying factors range, the number of iterations etc..
(1) selection of population scale num
In optimization algorithm, the number of particles of selection is excessive, is able to ascend the ability of searching optimum of algorithm, but also can simultaneously
Convergence speed of the algorithm is substantially reduced, time loss also will increase dramatically;The number of particles of selection is less, is capable of the time of reduction
Consumption promotes convergence rate, also improves a possibility that falling into local optimum.Herein for the practical problem of routeing, in advance
A large amount of test has first been carried out, has used suitable population scale for improved particle swarm optimization algorithm.It is tested by repetition test
Card chooses number of particles n=150, can not only cook up reasonable path under artificially generated terrain at this time, moreover it is possible to guarantee algorithm
Time loss and ability of searching optimum.
(2) inertia weight range is arranged
During algorithm initialization, need to make the range of inertia weight setting, the selection of inertia weight range
It will affect the efficiency and performance of algorithm.When ω value range is larger, when population particle solution space range is larger, the algorithm overall situation is searched
Rope capability improving, but will lead to search precision reduction;When ω value range is smaller, when population particle solution space range is smaller,
Algorithm local search ability is promoted, but algorithm is made to have fallen into the puzzlement of local extremum, cannot cook up global consumption cost most
Low optimal path.For the global and local search capability of balanced algorithm, predecessor experience is combined to carry out before being configured parameter
A large amount of tests propose that using the value range of ω be [0.4,0.7], and algorithm can cook up unmanned plane in simulated environment at this time
Can flight road.
(3) Studying factors c1And c2Selection
Studying factors c1And c2The ability for having self-teaching and learning to optimized individual, and then enable individual particles in population
It is enough close to global optimum's particle and local optimal particle.c1Adjust the maximum step-length that particle flies to individual optimal direction, c2It adjusts
The maximum step-length that section particle flies to optimal direction, the two factors directly determine population at individual experience and group's experience to particle
The influence of self-operating track, the information exchange between reaction particle group.c1And c2Collocation select it is different, to the shadow of algorithm
The degree of sound is also different.It in practical applications, is largely to choose by rule of thumb, often there is no fixed parameter selection fixing means
Parameter selection is c1=c2=1.49618 and ω=0.7298 combination, repeatedly test hair is carried out to this group of data herein
Now there is certain feasibility, carried out under circumstance of initialization herein according to this group of data.But in different environment and difference
Artificially generated terrain in the case of according to practical problem change.
Step 2: introducing Chaos Variable and population particle is initialized, by initialization informations such as the speed of particle, positions
It is mapped in reasonable chaos system using formula (1).
Xn+1=4Xn(1-Xn) (1)
Chaos phenomenon refers to the irregular movement for seeming random in deterministic system.The system table described by certainty theory
It is now non-repeatability and unpredictability.Chaos Variable is not meant to all at random, but one kind has certain inherent law
The phenomenon that.Although chaos system has certain randomness, chaology is with regularity;Chaotic motion can be one
Determine to carry out in range, " regularity " of its own is stateful without repeating through institute.Therefore, more had using Chaos Variable
The search of effect is advantageously than random search.
The basic thought for introducing Chaotic Optimization Strategy is the solution space that Chaos Variable is mapped to optimized variable, and is utilized mixed
Ignorant variable scans for.One typical chaos system of Logistics model, Logistics mapping mathematical expression formula are formula
(2)。
Xn+1=Xn·μ·(1-Xn) (2)
In formula (2), control parameter μ ∈ [0,4], X ∈ (0,1).Studies have shown that system is in completely mixed as μ=4
It is ignorant, on (0,1) section with distribution;μ=4 are taken to make the Chaos Variable X generated hereinn+1With good ergodic, use
Chaos Variable deinitialization population particle.
Fig. 2 is chaos and random particles contrast schematic diagram, and the random number range of selection is x ∈ (0,100), y ∈ (0,
100), two kinds of distributions generate 40 particles, by result figure it can be seen that the particle that completely random generates concentrates on x ∈ (0,40)
In range, equally distributed particle can be distributed in x ∈ (0,100), in the entire space y ∈ (0,100), as can be seen from the results into
When row routeing, population uniform particle is distributed in the feasibility for being conducive to boosting algorithm in space.
In conclusion do not change the randomness of population particle using chaotic model initialization particle position and particle rapidity,
And improve intragroup particle diversity and search spread, it proposes to adapt to a kind of based on chaos optimization of practical problem
Particle swarm optimization algorithm has certain feasibility.
Step 3: calculating the fitness value of individual particles in population, assess the fitness value size of particle, and in population
Global optimum and local optimal particle initialize.
Step 4: for each particle in population, being selected using survival of the fittest strategy, consumption cost will be threatened higher
Particle eliminate, and update global optimum's particle and local optimal particle accordingly.
Not only to consider that can algorithm cook up the path that flies for meeting mission requirements, will also examine in Path Planning for Unmanned Aircraft Vehicle
Consider the path cooked up and threaten cost consumption (whether distance consumption is optimal, and whether energy consumption is most excellent).Therefore, algorithm carries out
When emulation experiment, not only to consider to threaten obstacle avoidance problem, it is also contemplated that threatening the minimum problem of cost evaluation functional value.Tradition
Particle swarm optimization algorithm when carrying out routeing, particle from starting point to target point is normally to give up in an iteration circulation
It abandons, terminates search process.But this mode is easily trapped into local extremum trap in the later period of routeing, algorithm will guarantee
The equilibrium of global optimum and local optimal performance could be that unmanned plane cooks up optimal air line.Therefore, in order to avoid being advised in air route
Draw the later period causes the threat cost of algorithm to consume problems of too because algorithm local search ability enhances, and this chapter is that population is excellent
Change algorithm and introduces survival of the fittest strategy.
Survival of the fittest strategy mainly allows those to give up the particle for falling into local extremum trap and threat in search process
Cost consumes excessive particle, it is made to be no longer participate in the route searching of next round, terminates in advance the life cycle of poor particle,
In entire algorithm implementation procedure, by comparing the threat cost value of each particle, eliminates and cost is threatened to consume high particle,
It thus can preferably guarantee that whole threat consumption cost is optimal.The pseudocode of survival of the fittest strategy is expressed as follows:
Step 5: updating position, the velocity information of individual particles according to the particle swarm algorithm model of linear decrease, and update
Inertia weight value in next round iteration.
In D dimension space, population is made of n particle, by vector X=(X1,X2,X3,…,Xn) indicate, i-th particle
Position vector Xi=(xi1,xi2,xi3,…,xiD)TIt indicates, the speed of i-th of particle is expressed as vi=(vi1,vi2,vi3,…,
viD)T, the local extremum of the t times iteration is pt, the global extremum of the t times iteration is gt。
PSO algorithm is the positional value of the random initializtion population at individual within the scope of solution space and the velocity space when initial
And velocity amplitude, the position and speed of individual particles carries out group's particle according to formula (3) and formula (4) respectively in the t times iteration
Update operation.
xij(t+1)=xij(t)+vij(t+1) (3)
vij(t+1)=ω vij(t)+c1r1(pij(t)-xij(t))+c2r2(gij(t)-xij(t)) (4)
Wherein, c1And c2It for Studying factors, and is to obey the equally distributed random number of 0-1.ω is inertia weight coefficient.Speed
Degree more new formula is divided into three parts.First part is the speed of the last iteration of particle, it indicates the current kinetic of particle
State.Second part is cognitive learning part, i.e., current point is directed toward the vector between optimum point, and Part III is portion of social learning
Point, it refers to the vector between current point and global optimum's particle.
Increasing different inertia weight ω is the method for solving to fall into a kind of relative efficiency of local optimum, and inertia weight increases
Greatly, enhance ability of searching optimum, inertia weight is smaller, and ability of searching optimum reduces, therefore introduces in different practical problems
Different inertia weights is very crucial.Adaptive inertia weight is introduced into dove group's algorithm by this patent, is needed in conjunction with practical problem
The specific performance of summation algorithm, introduces the inertia weight coefficient of linear decrease in particle swarm optimization algorithm, and boosting algorithm is searched
Suo Nengli.In three-dimensional environment path planning, according to the difference of environment complexity and mission requirements in routeing initial stage, enhancing
The ability of searching optimum of population particle guarantees the optimality in particle two-dimensional surface, uses biggish inertia weight coefficient;It is navigating
The circuit planning later period has been substantially achieved the approximate location of globally optimal solution, carries out route searching optimizing in a small range, then needs
Enhance local search ability, uses lesser inertia weight coefficient.Linear decrease model is shown in formula (5).
ω=ωmax-t*(ωmax-ωmin)/tmax (5)
Wherein, ωmaxIt is the maximum value of inertia weight, ωminIt is the minimum value of inertia weight, t indicates current iteration number,
tmaxIndicate the maximum number of iterations of algorithm.
The speed more new formula for increasing Linear recurring series particle swarm algorithm is then expressed as formula (6).
vij(t+1)=(ωmax-t*(ωmax-ωmin)/tmax)vij(t)+c1r1(pij(t)-xij(t))+c2r2(gij(t)-
xij(t)) (6)
Step 6: comparing the relationship of current iteration number and maximum number of iterations, if current iteration number is greater than greatest iteration
Number then terminates;Otherwise, return step 3.
The beneficial effects of the present invention are: the present invention is different according to the complexity of terrain environment, selector closes environmental requirement
The algorithm of mission requirements carries out simulation modeling for two kinds of true environment data, and for the actual conditions of landform to population
Algorithm carries out more stragetic innovations.The experimental results showed that under undulating topography, the planning path length of the IPSO algorithm proposed,
Shorten 5.2% compared with AWPIO, shortens 5.9% compared with PSO;The time loss of IPSO algorithm shortens compared with AWPIO
5.25%, shorten 4.15% compared with PSO;Under alpine terrain, the planning path length of the IPSO algorithm proposed, with
AWPIO shortens 2.69% compared to shortening 1.3% compared with PSO;The time loss of IPSO algorithm shortens compared with AWPIO
24.3%, shorten 26.3% compared with PSO.Therefore, IPSO is a kind of intelligent algorithm for being more suitable for routeing.
Detailed description of the invention
The drawings described herein are used to provide a further understanding of the present invention, constitutes part of this application, this hair
Bright illustrative embodiments and their description are used to explain the present invention, and are not constituted improper limitations of the present invention.In the accompanying drawings:
Fig. 1 is the flow chart for inventing the algorithm;
Fig. 2 is chaos and random particles contrast schematic diagram;
Fig. 3 is hills environment map described in the present embodiment;
Fig. 4 is mountain area environment map described in the present embodiment.
Specific embodiment
Following will be combined with the drawings in the embodiments of the present invention, and technical solution in the embodiment of the present invention carries out clear, complete
Site preparation description, it is clear that described embodiments are only a part of the embodiments of the present invention, instead of all the embodiments.It is based on
Embodiment in the present invention, it is obtained by those of ordinary skill in the art without making creative efforts all other
Embodiment shall fall within the protection scope of the present invention.
Embodiment 1
Three-dimensional environment design
(1) hills environment
Hills environment is as shown in figure 3, S and D respectively indicate the starting point and target point of task.Wherein, it is risen under original state
Initial point coordinate is (2,2,96), and coordinate of ground point is (170,170,76), and three cylindrical bodies in map are three kinds of different prestige
The three-dimensional threat modeling of range, Different Effects rank is coerced, the range (radius) of threat is respectively 12km, 17km, 20km, wherein
The more specific location information in threat source is expressed as (50,130,80,30) in three-dimensional system of coordinate, (80,42,80,25),
(140,100,80,25).Wherein (x, y, z) is geographical coordinate of the threat source in coordinate system, and h is threat source effective height;Fig. 3
For undulating topography environment map;
(2) mountain area environment
Mountain area environment is as shown in figure 4, S and D respectively indicate starting point and target point, wherein starting point coordinate under original state
For (15,2,60), coordinate of ground point is (160,160,55), three cylindrical bodies in map be three kinds of different threat ranges,
The range (radius) of the three-dimensional threat modeling of Different Effects rank, threat is respectively 12km, 17km, 20km.Threat source it is specific
Location information is expressed as (40,30,50,15) in three-dimensional system of coordinate, (60,100,50,16), (140,135,50,16),
(140,135,50,14).Wherein (x, y, z) is geographical coordinate of the threat source in coordinate system, and h is threat source effective height;Fig. 4
For alpine terrain environment map;The analysis of three-dimensional environment route programming result.
The present invention pass through many experiments after find, two kinds of algorithms of path length and elapsed time that many experiments are cooked up it
Between gap be not obvious, and the performance between algorithm also has certain fluctuation, and algorithm comparative analysis cannot be according to one
Group experimental data concludes the performance of algorithm.Therefore, it is increased in the present invention to the mean analysis of experiment sample and variance analysis
Method assesses the performance of algorithm.Mean value is for assessing overall value average level, and variance is to data dispersion degree
Statistic, therefore statistically also there is certain meaning using the method for mean analysis and variance analysis.
During Path Planning for Unmanned Aircraft Vehicle, population mean is smaller, illustrates the fewer away from discrete time of consumption, algorithm performance is got over
It is excellent.Conversely, algorithm performance is poorer.Population variance is smaller, illustrates that the fluctuation of algorithm is smaller, algorithm performance is more excellent.Conversely, algorithm
Performance is poorer.Therefore, for Path Planning for Unmanned Aircraft Vehicle problem, it should be algorithm mean value it is the smaller the better, variance be also it is smaller more
It is good.
(1) route programming result is analyzed under undulating topography environment
In the environment of hills, AWPIO algorithm, PSO algorithm is respectively adopted and improves PSO algorithm progress simulation comparison experiment,
For the comparativity for guaranteeing two algorithms, three algorithms use identical the number of iterations when emulation experiment, and 300 times.It carries out respectively
After 20 experiments, it is shown in Table 1 using the performance data that AWPIO algorithm obtains in the environment of hills, the performance obtained using PSO algorithm
Data are shown in Table 2, are shown in Table 3 using the performance data that PSO algorithm obtains is improved.
Table 1AWPIO algorithm performance tables of data
Table 2PSO algorithm performance tables of data
Table 3 improves PSO algorithm performance tables of data
By the method for mean analysis in table 1, table 2 and table 3, from optimal path length angle analysis, AWPIO algorithm is on mound
Mound landform many experiments mean value is 285.9km, and mean value of the PSO algorithm in identical environment is 287.65km, improves PSO algorithm and exists
The mean value 270.9km of identical environment, the simulation experiment result show to improve PSO algorithm path planning length most in the environment of hills
Short, AWPIO algorithm path planning length is better than PSO algorithm.From time loss angle analysis, AWPIO algorithm is in undulating topography ring
The time loss average value that Multi simulation running experiment is carried out under border is 401.18ms, time loss of the PSO algorithm in identical environment
Average value is 405.41ms, and improving time loss average value of the PSO algorithm in identical environment is 422.25ms, emulation experiment knot
Fruit shows to analyze from time loss angle, and PSO algorithm time loss is improved in undulating topography environment and is higher than other two kinds calculations
Method, AWPIO algorithm performance are optimal.
The method that variance analysis is used in table 1, table 2 and table 3, is tested from optimal path length angle A WPIO algorithm 20 times
Population variance is that 4.7, PSO algorithm population variance is 4.22, and improving PSO algorithm population variance is 4.72.By further using
Variance analysis, which obtains, improves the variance that PSO algorithm variance is higher than AWPIO algorithm, also above PSO algorithm variance, although shortest path
Length is small, but stability is slightly lower.
(2) route programming result is analyzed under alpine terrain environment
In the environment of mountain area, AWPIO algorithm, PSO algorithm is respectively adopted and improves PSO algorithm progress Path Planning for Unmanned Aircraft Vehicle and imitates
True comparative experiments, three algorithms are 300 times using the number of iterations when emulation experiment, carry out 20 experiments respectively.In mountain area ring
4 are shown in Table using the performance data that AWPIO algorithm obtains in border, 5 is shown in Table using the performance data that PSO algorithm obtains, uses improvement
The performance data that PSO algorithm obtains is shown in Table 6.
Table 4AWPIO algorithm performance tables of data
Table 5PSO algorithm performance tables of data
Table 6 improves PSO algorithm performance tables of data
The method that mean analysis is used in table 4, table 5 and table 6, from optimal path length angle analysis, AWPIO algorithm is on mountain
Optimal path length mean value is 223.15km under area's landform, and PSO algorithm optimal path length mean value under identical environment is
229.25km, improving PSO algorithm optimal path length mean value under identical environment is 220.35km, and the simulation experiment result shows
The optimal path length mean value that PSO algorithm is improved in the environment of mountain area is most short, followed by AWPIO algorithm, PSO algorithm optimal path
Length mean value longest.From time loss angle analysis, AWPIO algorithm carried out under undulating topography environment Multi simulation running experiment when
Between consumption average value be 247.32ms, time loss average value of the PSO algorithm in identical environment be 342.43ms, improve PSO
Time loss average value of the algorithm in identical environment is 335.62ms, and the simulation experiment result shows to come from time loss angle
Analysis, AWPIO algorithm calculating speed is most fast in undulating topography environment, improves PSO algorithm calculating speed and is slightly faster than PSO algorithm.
It is overall to carry out 20 experiments from optimal path length angle A WPIO algorithm for the method that variance analysis is used in table 4, table 5 and table 6
Variance is that 5.21, PSO algorithm population variance is 3.19, and improving PSO algorithm population variance is 6.7.The simulation experiment result shows
PSO algorithm has higher stability when solving optimal path in the environment of mountain area, and the stability of AWPIO algorithm is better than improvement
The stability of PSO algorithm.
In three-dimensional environment, 20 typical datas are chosen herein and are analyzed, by using mean analysis and variance analysis
The performance of method parser obtain: it is optimal that particle swarm algorithm routing cost is improved in the environment of hills, and stability is high;Mountain
Improvement particle swarm algorithm time loss is optimal in area's environment, and routing cost is optimal, and stability is also more preferable.
In conclusion the particle swarm algorithm of more stragetic innovations reduces algorithmic theory of randomness, algorithm is improved in three-dimensional space
Efficiency and stability.There is higher algorithm stability and applicability compared with other two kinds of algorithms, still in emulation experiment
Many experiments are it has also been found that algorithms of different adaptation environment is not quite similar, it is therefore desirable to make a concrete analysis of as the case may be, different
The superiority and inferiority of comprehensive consideration algorithm is wanted to select suitable algorithm in the case where environment, different task.
Claims (6)
1. a kind of no-manned plane three-dimensional Route planner based on more stragetic innovation particle swarm algorithms, which is characterized in that including such as
Lower step:
Step 1: the parameter during setting particle swarm optimization algorithm;
Step 2: introducing Chaos Variable and population particle is initialized, the initialization informations such as the speed of particle, position are made
It is mapped in reasonable chaos system with formula (1);
Xn+1=4Xn(1-Xn) (1)
Step 3: calculating the fitness value of individual particles in population, assess the fitness value size of particle, and to complete in population
The optimal and local optimal particle of office initializes;
Step 4: for each particle in population, being selected using survival of the fittest strategy, consume the higher grain of cost for threatening
Son eliminates, and updates global optimum's particle and local optimal particle accordingly;
Step 5: updating position, the velocity information of individual particles according to the particle swarm algorithm model of linear decrease, and update next
Take turns the inertia weight value in iteration;
In D dimension space, population is made of n particle, by vector X=(X1,X2,X3,…,Xn) indicate, the position of i-th of particle is used
Vector Xi=(xi1,xi2,xi3,…,xiD)TIt indicates, the speed of i-th of particle is expressed as vi=(vi1,vi2,vi3,…,viD)T, t
The local extremum of secondary iteration is pt, the global extremum of the t times iteration is gt;
PSO algorithm is the positional value and speed of the random initializtion population at individual within the scope of solution space and the velocity space when initial
Angle value, the position and speed of individual particles carries out the update of group's particle according to formula (3) and formula (4) respectively in the t times iteration
Operation;
xij(t+1)=xij(t)+vij(t+1) (3)
vij(t+1)=ω vij(t)+c1r1(pij(t)-xij(t))+c2r2(gij(t)-xij(t)) (4)
Wherein, c1And c2It for Studying factors, and is to obey the equally distributed random number of 0-1;ω is inertia weight coefficient;Speed is more
New formula is divided into three parts;First part is the speed of the last iteration of particle, it indicates the current motion state of particle;
Second part is cognitive learning part, i.e., current point is directed toward the vector between optimum point, and Part III is social learning part, it is
Refer to the vector between current point and global optimum's particle;R1 and r2 is that value range is [0,1], and is uniformly distributed in the section
Pseudo random number.
Step 6: comparing the relationship of current iteration number and maximum number of iterations, if current iteration number is greater than greatest iteration time
Number, then terminate;Otherwise, return step 3.
2. a kind of no-manned plane three-dimensional routeing side based on more stragetic innovation particle swarm algorithms according to claim 1
Method, which is characterized in that the parameter in the step 1 specifically includes that the number of iterations, self-teaching factor range, social learning because
Subrange, inertia weight range;
Parameter described in step 1 further include: (1) selection of population scale num: the population scale of particle swarm optimization algorithm chooses grain
Subnumber amount n=150;(2) inertia weight range ω is arranged: the value range of ω is [0.4,0.7], and algorithm can emulate at this time
That cooks up unmanned plane in environment can flight road;(3) Studying factors c1And c2Select: common parameter selection is c1=c2=
1.49618 and ω=0.7298 combination, repeatedly test is carried out to this group of data and being found to have certain feasibility, initializes feelings
It is carried out under condition according to this group of data;But change in different environment and different artificially generated terrains according to practical problem.
3. a kind of no-manned plane three-dimensional routeing side based on more stragetic innovation particle swarm algorithms according to claim 1
Method, which is characterized in that Logistics model is a typical chaos system, Logistics in chaos system described in step 2
Mapping mathematical expression formula is formula (5)
Xn+1=Xn·μ·(1-Xn) (5)
In formula (5), control parameter μ ∈ [0,4], X ∈ (0,1);Studies have shown that system is in Complete Chaos as μ=4,
(0,1) on section with distribution;μ=4 are taken to make the Chaos Variable X generated hereinn+1With good ergodic, chaos is used
Variable deinitialization population particle.
4. a kind of no-manned plane three-dimensional routeing side based on more stragetic innovation particle swarm algorithms according to claim 1
Method, which is characterized in that the pseudocode of survival of the fittest strategy described in step 4 is expressed as follows:
If newthreat≤oldthreat
Fitness (i)=newthreat;
Path (i :)=g_best;
else
Fitness (i)=oldthreat;
Path (i :)=Path (i :);
End。
5. a kind of no-manned plane three-dimensional routeing side based on more stragetic innovation particle swarm algorithms according to claim 1
Method, which is characterized in that the particle swarm algorithm model described in step 5 according to linear decrease is shown in formula (6):
ω=ωmax-t*(ωmax-ωmin)/tmax (6)
Wherein, ωmaxIt is the maximum value of inertia weight, ωminIt is the minimum value of inertia weight, t indicates current iteration number, tmax
Indicate the maximum number of iterations of algorithm;
The speed more new formula for increasing Linear recurring series particle swarm algorithm is then expressed as formula (7):
vij(t+1)=(ωmax-t*(ωmax-ωmin)/tmax)vij(t)+c1r1(pij(t)-xij(t))+c2r2(gij(t)-xij
(t)) (7) 。
6. a kind of no-manned plane three-dimensional routeing side based on more stragetic innovation particle swarm algorithms according to claim 1
Method, which is characterized in that the pseudocode for improving particle swarm algorithm is as follows:
。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811091779.7A CN109269502A (en) | 2018-09-19 | 2018-09-19 | A kind of no-manned plane three-dimensional Route planner based on more stragetic innovation particle swarm algorithms |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811091779.7A CN109269502A (en) | 2018-09-19 | 2018-09-19 | A kind of no-manned plane three-dimensional Route planner based on more stragetic innovation particle swarm algorithms |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109269502A true CN109269502A (en) | 2019-01-25 |
Family
ID=65197014
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811091779.7A Withdrawn CN109269502A (en) | 2018-09-19 | 2018-09-19 | A kind of no-manned plane three-dimensional Route planner based on more stragetic innovation particle swarm algorithms |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109269502A (en) |
Cited By (10)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110597280A (en) * | 2019-09-04 | 2019-12-20 | 中国航空工业集团公司沈阳飞机设计研究所 | Method for re-planning aircraft route |
CN110986954A (en) * | 2020-02-18 | 2020-04-10 | 中国人民解放军空军工程大学 | Military transport plane route planning method based on wolf optimization algorithm |
CN111307158A (en) * | 2020-03-19 | 2020-06-19 | 哈尔滨工程大学 | AUV three-dimensional route planning method |
CN111428849A (en) * | 2020-03-24 | 2020-07-17 | 国电南瑞科技股份有限公司 | Improved particle swarm algorithm-based transfer function model parameter identification method and device |
CN111709214A (en) * | 2020-06-18 | 2020-09-25 | 福州大学 | Orbit planning method based on discrete particle swarm optimization |
CN112230678A (en) * | 2020-10-29 | 2021-01-15 | 皖江工学院 | Three-dimensional unmanned aerial vehicle path planning method and planning system based on particle swarm optimization |
CN112967763A (en) * | 2021-03-08 | 2021-06-15 | 北京北排水环境发展有限公司 | Effluent total phosphorus prediction method based on fuzzy neural network, electronic equipment and medium |
CN114022538A (en) * | 2021-09-15 | 2022-02-08 | 苏州中科华影健康科技有限公司 | Route planning method and device for endoscope, terminal and storage medium |
CN116774725A (en) * | 2023-05-19 | 2023-09-19 | 北华航天工业学院 | Low-altitude unmanned aerial vehicle path planning and obstacle avoidance method capable of improving SAC algorithm |
CN117130263A (en) * | 2023-10-26 | 2023-11-28 | 博创联动科技股份有限公司 | Intelligent control method and system for whole vehicle based on big data of Internet of vehicles |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI410103B (en) * | 2008-06-25 | 2013-09-21 | Univ Shu Te | Chaos Communication Confidentiality System and Its Data Transmission Method and Proportional Integral Control Module Design Method |
CN103559396A (en) * | 2013-10-31 | 2014-02-05 | 华北水利水电大学 | Automatic pharmacy storage location optimizing method based on improved chaos particle swarm algorithm |
CN105988366A (en) * | 2015-02-13 | 2016-10-05 | 大连大学 | Space robot minimal base disturbance trajectory planning method |
CN106441308A (en) * | 2016-11-10 | 2017-02-22 | 沈阳航空航天大学 | Unmanned aerial vehicle path planning method based on adaptive weight pigeon flock algorithm |
-
2018
- 2018-09-19 CN CN201811091779.7A patent/CN109269502A/en not_active Withdrawn
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
TWI410103B (en) * | 2008-06-25 | 2013-09-21 | Univ Shu Te | Chaos Communication Confidentiality System and Its Data Transmission Method and Proportional Integral Control Module Design Method |
CN103559396A (en) * | 2013-10-31 | 2014-02-05 | 华北水利水电大学 | Automatic pharmacy storage location optimizing method based on improved chaos particle swarm algorithm |
CN105988366A (en) * | 2015-02-13 | 2016-10-05 | 大连大学 | Space robot minimal base disturbance trajectory planning method |
CN106441308A (en) * | 2016-11-10 | 2017-02-22 | 沈阳航空航天大学 | Unmanned aerial vehicle path planning method based on adaptive weight pigeon flock algorithm |
Non-Patent Citations (2)
Title |
---|
王尔申等: "基于混沌的改进粒子群滤波算法", 《北京航空航天大学学报》 * |
黄思铭: "基于改进鸽群算法的无人机航路规划研究", 《中国优秀硕士学位论文全文数据库》 * |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110597280A (en) * | 2019-09-04 | 2019-12-20 | 中国航空工业集团公司沈阳飞机设计研究所 | Method for re-planning aircraft route |
CN110986954B (en) * | 2020-02-18 | 2023-08-22 | 中国人民解放军空军工程大学 | Military transport plane route planning method based on gray wolf optimization algorithm |
CN110986954A (en) * | 2020-02-18 | 2020-04-10 | 中国人民解放军空军工程大学 | Military transport plane route planning method based on wolf optimization algorithm |
CN111307158A (en) * | 2020-03-19 | 2020-06-19 | 哈尔滨工程大学 | AUV three-dimensional route planning method |
CN111428849A (en) * | 2020-03-24 | 2020-07-17 | 国电南瑞科技股份有限公司 | Improved particle swarm algorithm-based transfer function model parameter identification method and device |
CN111709214B (en) * | 2020-06-18 | 2022-06-14 | 福州大学 | Orbit planning method based on discrete particle swarm optimization |
CN111709214A (en) * | 2020-06-18 | 2020-09-25 | 福州大学 | Orbit planning method based on discrete particle swarm optimization |
CN112230678A (en) * | 2020-10-29 | 2021-01-15 | 皖江工学院 | Three-dimensional unmanned aerial vehicle path planning method and planning system based on particle swarm optimization |
CN112967763A (en) * | 2021-03-08 | 2021-06-15 | 北京北排水环境发展有限公司 | Effluent total phosphorus prediction method based on fuzzy neural network, electronic equipment and medium |
CN114022538A (en) * | 2021-09-15 | 2022-02-08 | 苏州中科华影健康科技有限公司 | Route planning method and device for endoscope, terminal and storage medium |
CN116774725A (en) * | 2023-05-19 | 2023-09-19 | 北华航天工业学院 | Low-altitude unmanned aerial vehicle path planning and obstacle avoidance method capable of improving SAC algorithm |
CN116774725B (en) * | 2023-05-19 | 2024-01-16 | 北华航天工业学院 | Low-altitude unmanned aerial vehicle path planning and obstacle avoidance method capable of improving SAC algorithm |
CN117130263A (en) * | 2023-10-26 | 2023-11-28 | 博创联动科技股份有限公司 | Intelligent control method and system for whole vehicle based on big data of Internet of vehicles |
CN117130263B (en) * | 2023-10-26 | 2024-01-16 | 博创联动科技股份有限公司 | Intelligent control method and system for whole vehicle based on big data of Internet of vehicles |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109269502A (en) | A kind of no-manned plane three-dimensional Route planner based on more stragetic innovation particle swarm algorithms | |
CN106979784B (en) | Non-linear track planning based on hybrid pigeon swarm algorithm | |
CN109059931B (en) | A kind of paths planning method based on multiple agent intensified learning | |
CN109254588B (en) | Unmanned aerial vehicle cluster cooperative reconnaissance method based on cross variation pigeon swarm optimization | |
CN111176334B (en) | Multi-unmanned aerial vehicle cooperative target searching method | |
CN111722643B (en) | Unmanned aerial vehicle cluster dynamic task allocation method imitating wolf colony cooperative hunting mechanism | |
CN104573812B (en) | A kind of unmanned plane air route determining method of path based on particle firefly colony optimization algorithm | |
CN106705970A (en) | Multi-UAV(Unmanned Aerial Vehicle) cooperation path planning method based on ant colony algorithm | |
CN108388250B (en) | Water surface unmanned ship path planning method based on self-adaptive cuckoo search algorithm | |
CN112082552A (en) | Unmanned aerial vehicle flight path planning method based on improved hybrid particle swarm optimization algorithm | |
CN106441308A (en) | Unmanned aerial vehicle path planning method based on adaptive weight pigeon flock algorithm | |
CN112000131A (en) | Unmanned aerial vehicle cluster path planning method and system based on artificial potential field method | |
Asghari et al. | Multi‐swarm and chaotic whale‐particle swarm optimization algorithm with a selection method based on roulette wheel | |
Wang et al. | Rank-driven salp swarm algorithm with orthogonal opposition-based learning for global optimization | |
CN114840020A (en) | Unmanned aerial vehicle flight path planning method based on improved whale algorithm | |
Hu et al. | HG-SMA: hierarchical guided slime mould algorithm for smooth path planning | |
CN113836803A (en) | Improved wolf colony algorithm-based unmanned aerial vehicle cluster resource scheduling method | |
CN110162077A (en) | A kind of unmanned aerial vehicle flight path planing method based on flying fish algorithm | |
Long et al. | A multi-subpopulation bacterial foraging optimisation algorithm with deletion and immigration strategies for unmanned surface vehicle path planning | |
Wang et al. | On the mathematical models and applications of swarm intelligent optimization algorithms | |
CN113805609A (en) | Unmanned aerial vehicle group target searching method based on chaos lost pigeon group optimization mechanism | |
Qingwen et al. | Cooperative jamming resource allocation of UAV swarm based on multi-objective DPSO | |
CN112000126A (en) | Whale algorithm-based multi-unmanned-aerial-vehicle collaborative searching multi-dynamic-target method | |
CN111811532B (en) | Path planning method and device based on impulse neural network | |
Heidari et al. | Improved black hole algorithm for efficient low observable UCAV path planning in constrained aerospace |
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 |
Application publication date: 20190125 |
|
WW01 | Invention patent application withdrawn after publication |