CN109961130A - A kind of method and apparatus that target object position is determined based on particle swarm algorithm - Google Patents
A kind of method and apparatus that target object position is determined based on particle swarm algorithm Download PDFInfo
- Publication number
- CN109961130A CN109961130A CN201910181971.3A CN201910181971A CN109961130A CN 109961130 A CN109961130 A CN 109961130A CN 201910181971 A CN201910181971 A CN 201910181971A CN 109961130 A CN109961130 A CN 109961130A
- Authority
- CN
- China
- Prior art keywords
- particle
- fitness value
- population
- speed
- target object
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Feedback Control In General (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
This specification embodiment provides a kind of method and apparatus for determining target object position based on particle swarm algorithm, wherein the described method includes: determining the speed of each particle and position in population；According to speed and position, the fitness value of each particle is determined；It according to the size of fitness value, sorts to the fitness value of particle each in population, and by selecting the biggish some particles of fitness value to enter cistern of chiasma in ranking results, some particles are determined according to population scale and crossover probability；It makes a variation to the particle in population；When meeting the termination condition of particle swarm algorithm, using the corresponding position of the optimal solution of particle swarm algorithm as the position of the target object.
Description
Technical field
This disclosure relates to which machine learning techniques field, in particular to one kind determine target object position based on particle swarm algorithm
Method and apparatus.
Background technique
Particle swarm algorithm be it is a kind of by simulate flock of birds look for food during migrate with clustering behavior, and propose a kind of base
In the global random searching algorithm of swarm intelligence.The application scenarios of particle swarm algorithm are more, and one of application scenarios can be
For searching for the position of some object, which is properly termed as target object.Traditional particle swarm algorithm, can be from solving at random
Hair finds position of the optimal solution as target object by iteration.But this algorithm is often easily trapped into locally optimal solution, and
And convergence rate is unsatisfactory, affects efficiency and accuracy to target object position search.
Summary of the invention
In view of this, this specification one or more embodiment, which provides one kind, determines target object position based on particle swarm algorithm
The method and apparatus set, to improve the efficiency and accuracy of location finding.
Specifically, this specification one or more embodiment is achieved by the following technical solution:
In a first aspect, providing a kind of method of position for determining target object based on particle swarm algorithm, which comprises
Determine the speed of each particle and position in population；
According to the speed and position, the fitness value of each particle, the fitness value and the particle are determined
The distance between target object is related；
It according to the size of the fitness value, sorts to the fitness value of each particle in the population, and by sorting
As a result the selection biggish some particles of fitness value enter cistern of chiasma in, and some particles are according to population scale and intersect general
Rate determines, by selecting particle to be intersected in the cistern of chiasma；
It makes a variation to the particle in the population；
When meeting the termination condition of particle swarm algorithm, using the corresponding position of the optimal solution of particle swarm algorithm as the mesh
Mark the position of object.
Second aspect, provides a kind of device of position that target object is determined based on particle swarm algorithm, and described device includes:
Data determining module, for determining the speed of each particle and position in population；
Cross processing module, it is described suitable for determining the fitness value of each particle according to the speed and position
Answer the distance between angle value and the particle and target object related；And according to the size of the fitness value, to the particle
The fitness value sequence of each particle in group, and by selecting the biggish some particles of fitness value to enter intersection in ranking results
Pond, some particles are determined according to population scale and crossover probability, by selecting particle to be intersected in the cistern of chiasma；
Make a variation processing module, for making a variation to the particle in the population；
Position determination module, for when meeting the termination condition of particle swarm algorithm, by the optimal solution pair of particle swarm algorithm
Position of the position answered as the target object.
The third aspect, provides a kind of equipment of position that target object is determined based on particle swarm algorithm, and the equipment includes
Memory, processor and storage on a memory and the computer program that can run on a processor, processor execution institute
The method and step for determining target object position described in this specification any embodiment based on particle swarm algorithm is realized when stating program.
The method and apparatus of this specification one or more embodiment introduce particle swarm algorithm by that will intersect and make a variation,
Convergence rate is improved, and the optimal solution effect found is more preferable.
Detailed description of the invention
In order to illustrate more clearly of this specification one or more embodiment or technical solution in the prior art, below will
A brief introduction will be made to the drawings that need to be used in the embodiment or the description of the prior art, it should be apparent that, it is described below
Attached drawing is only some embodiments recorded in this specification one or more embodiment, and those of ordinary skill in the art are come
It says, without any creative labor, is also possible to obtain other drawings based on these drawings.
Fig. 1 is a kind of position that target object is determined based on particle swarm algorithm that this specification one or more embodiment provides
The method set；
Flying robot group when Fig. 2 is initialization；
Fig. 3 to Fig. 5 is the process schematic that flying robot searches for life signal；
Fig. 6 is that one kind that this specification one or more embodiment provides determines target object position based on particle swarm algorithm
Device structural schematic diagram.
Specific embodiment
In order to make those skilled in the art more fully understand the technical solution in this specification one or more embodiment,
Below in conjunction with the attached drawing in this specification one or more embodiment, to the technology in this specification one or more embodiment
Scheme is clearly and completely described, it is clear that described embodiment is only a part of the embodiment, rather than whole realities
Apply example.Based on this specification one or more embodiment, those of ordinary skill in the art are not making creative work premise
Under every other embodiment obtained, shall fall within the protection scope of the present application.
Particle swarm algorithm is one, and for solving the algorithm of optimal solution, the application scenarios of the particle swarm algorithm are relatively broad,
For example, can be used for solving the shortest path in route planning；Alternatively, can be also used for solving the minimum in commercial profit project
Expense, etc..The different problems that can according to need solution, construct corresponding different functions, which is properly termed as fitness letter
It counts, a fitness value can be calculated in each particle in population according to the fitness function, to pass through fitness
Value measures superiority and inferiority of the particle relative to some problem target.
At least one embodiment of this specification provides a kind of combination genetic algorithm modified particle swarm optiziation, and heredity is calculated
Intersection and variation in method are introduced into population.Wherein, genetic algorithm be simulate Darwinian evolutionism natural selection and
The computation model of the biological evolution process of genetic mechanisms is a kind of side by simulating natural evolution process searches optimal solution
Method.
Fig. 1 be at least one embodiment of this specification provide a kind of target object position is determined based on particle swarm algorithm
Method, as shown in Figure 1, this method may include handling as follows:
In step 100, the speed of each particle and position in population are determined.
For example, this step may include following processing:
Firstly, population, including population size are initialized, the position and speed of each particle in population.
In this step, the population size of population is initialized, for example, sharing M particle in the population.
Also, the position and speed of each particle can also be initialized, wherein the position of particle can be a vector Xi
=(xi1,xi2,…,xiD), i=1,2 ... .M；The speed of particle can be expressed as vector: Vi=(vi1,vi2..., viD), i=1,
2….M.Wherein, D indicates that the search space of particle swarm algorithm search optimal solution is the search space of D dimension.
Initial particle position and speed can be random setting.
Secondly, calculating the fitness value of each particle.
In this step, the numerical value that each particle of fitness value is calculated according to fitness function, different application scenarios
Different fitness functions can be constructed.In population, each particle when initialization is equivalent to a RANDOM SOLUTION,
In the iteration renewal process of population, each particle is a potential optimal solution.If the fitness value of a particle compared with
Height indicates that the particle is closer apart from optimal solution, and the particle of better performances is belonged to during searching for optimal solution.
F can be usedit[i] indicates the fitness value of i-th of particle in population.
Then, each particle updates institute by the way that the fitness value of oneself to be compared with individual extreme value and global extremum
State speed and the position of particle.
Wherein, the optimal location that i-th of particle searches so far is known as individual extreme value, is denoted as pbest(i)。
The optimal location that entire population searches so far is global extremum, is denoted as gbest。
Particle swarm algorithm is to find optimal solution by successive ignition, wherein in iteration each time, each particle can be with
Update oneself position and speed by tracking two extreme values: an extreme value is exactly the optimal solution that particle itself is found, i.e., on
The individual extreme value stated；Another extreme value is the optimal solution that entire population is found, i.e., above-mentioned global extremum.
Wherein, the position of particle is one of parameter in fitness function, which can consider to be exactly fitness
The solution of function, if two particles compare, the fitness value of a particle is more excellent than the fitness value of another particle, then this is suitable
Answering angle value, more preferably the corresponding position of particle is also preferably to solve.
The fitness value of oneself being calculated can be compared by each particle with individual extreme value and global extremum, with
Speed and the position of the particle are updated, as follows:
By the fitness value of particle compared with individual extreme value, if the fitness value is greater than individual extreme value, with the grain
The position of son replaces the individual extreme value.By the fitness value of the particle compared with global extremum, if the fitness value
Greater than global extremum, then the global extremum is replaced with the position of the particle.
For by taking global extremum as an example: assuming that fitness function f (x, y)=x+y, wherein the position that (x, y) is particle is sat
Mark.Entire population when searching for global extremum, be using closest to the corresponding particle position of the fitness value of expected result as
Optimal location, that is, global extremum.For example, desired f (x, y)=1, and the position of some particle is x=0.5, y=
0.49, f (x, y)=0.99, then the error of the fitness value and desired fitness value is exactly 0.01.And it can preset
Error range is within 2%, then 0.01 error meets the condition of error range, then the corresponding position f (x, y)=0.99
(0.5,0.49) can replace original global extremum.That is, global extremum is the position of a particle, if particle
Fitness value is closest to model answer, closest to optimal solution, then using the position of the particle as global extremum.
Further according to the updated individual extreme value and global extremum, speed and the position of the particle are updated.It can be by
The update of particle rapidity and position is carried out according to following formula (1) and formula (2):
vid=vid+c1r1(pid-xid)+c2r2(pgd-xid) (1)
xid=xid+vid (2)
Wherein, c1And c2For Studying factors, also referred to as aceleration pulse (acceleration constant), r1And r2For [0,
1] uniform random number in range.It is consisted of three parts on the right of formula (1), first part is " inertia (inertia) " or " momentum
(momentum) " part reflects the movement " habit (habit) " of particle, and representing particle has becoming for oneself previous velocity of maintenance
Gesture；Second part is " cognition (cognition) " part, reflects particle to the memory (memory) of itself historical experience or returns
Recall (remembrance), represents the trend that itself oriented history optimum position of particle is approached；Part III is " society
(social) " part reflects group's historical experience of cooperative cooperating and knowledge sharing between particle, represent the oriented group of particle or
The trend that neighborhood history optimum position is approached.vidIt is the speed of particle, vid∈[-vmax,vmax], vmaxIt is constant, is set by user
The fixed speed for being used to limit particle.r1And r2It is the random number between [0,1].
By above-mentioned processing, the speed of available updated particle and position.
In a step 102, according to the speed of updated particle and position, the fitness value of each particle, and root are calculated
It is ranked up according to the size of fitness value；Cistern of chiasma is entered by selected section particle in ranking results, which is basis
Population scale and crossover probability obtain.
In this step, it is the particle by selecting to be intersected in population, forms cistern of chiasma.For example, can be according to such as
Lower processing obtains cistern of chiasma, it should be understood that, it is not limited in actual implementation and executes sequence.
It is possible, firstly, to calculate the fitness value of each particle according to the speed of updated particle and position.The fitness
Value can be related to the distance between particle and target object, for example, fitness value is bigger, indicates between particle and target object
Distance it is closer.
Then, it can be ranked up according to the size of fitness value.
Then, some particles of cistern of chiasma are entered according to population scale and crossover probability determination.For example, can be by grain
Subgroup scale obtains the number of particles N in cistern of chiasma multiplied by crossover probability.
Furthermore it can be according to number of particles N, by selecting the biggish top n particle of fitness value to enter friendship in ranking results
Pitch pond.
Wherein, crossover probability can be the data withdrawal ratio being artificially arranged, between 0 to 1.Cistern of chiasma size N
=population scale M* crossover probability, for example, current population shares 100 particles, crossover probability=0.1, then intersecting
Pond size=10.Then just the selection best particle of preceding 10 fitness values is put into cistern of chiasma.Particle in cistern of chiasma is all suitable
Answer angle value higher, it is believed that be all preferably to solve.It is this generate cistern of chiasma mode, can to avoid choose fitness compared with
The particle of difference does intersection particle.
At step 104, by selecting particle to be intersected in the cistern of chiasma.
This step is specific crossover operation.
Can be random by selecting two different particles in cistern of chiasma, carry out real number intersection as parent particle.
Fitness value is calculated to the filial generation particle obtained after intersection.
If the fitness value of the filial generation particle is higher than the fitness value of parent particle, by the position of the filial generation particle
Set the position and speed for replacing parent particle with speed.Otherwise, then the position and speed of parent particle does not change.By to particle
Particle in group carries out crossover operation, can increase the diversity of particle, jumps out local optimum, accelerates convergence rate.
For example: a filial generation particle can be generated in two parent particle intersections, if than two parents of filial generation particle
The fitness value of particle is all high, then the filial generation particle is just better than parent particle.Such as, it is assumed that two parent particles are parent grains
Sub- A and parent particle B, f (x, y)=x+y above, model answer are f (x, y)=1, if that parent particle A f (x,
Y)=0.8, the f (x, y)=0.7 of parent particle B, if the f (x, y)=0.9 of filial generation particle, this filial generation particle just compares
Two parent particles are all good.The position of the filial generation particle can be retained, and parent particle can be by removing in population.Population
Size remains unchanged, for example, population scale when initial is 200, remains 200.
By to particle carry out crossover operation, and with generate filial generation particle substitution parent particle so that filial generation particle after
The advantages of having held parent particle theoretically strengthens the search capability to interparticle region.For example, two parent particles are located
It in different local optimum regions, tends to get rid of local optimum then the two intersects the filial generation particle generated, and is changed
Into search result.
In step 106, it makes a variation to the particle in the population.
In this step, Gaussian mutation can be carried out to the particle in the population.The local search ability of Gaussian mutation
Preferably.Gaussian mutation is to be replaced when being made a variation with a random number of the normal distribution that a mean value is μ, variance is σ 2
Original genic value.
Wherein it is possible to which random selects particle to carry out Gaussian mutation from entire population according to mutation probability.For example,
Mutation probability can be 0.5, before a particle will make a variation, and the random value between one 0~1 be obtained, if the random value is greater than
0.5, then it makes a variation；Otherwise, it does not make a variation.
Calculate variation after particle fitness value, if variation after particle fitness value than make a variation before particle adaptation
Angle value is more excellent, then can replace the position and speed of particle before making a variation with the position and speed of particle after variation.Otherwise, may be used not
Change.
In step 108, judge whether the termination condition for meeting particle swarm algorithm.
This step can check whether the termination condition for meeting particle swarm algorithm.
The termination condition can be fitness value that final global extremum calculates within the precision of satisfaction, than
Such as, in example above-mentioned, it is assumed that fitness function f (x, y)=x+y, desired f (x, y)=1, particle
Position is x=0.5, y=0.49, f (x, y)=0.99, and the error of the fitness value and desired fitness value meets precision and wants
It asks, which can be used as global extremum, i.e. the optimal location that finds of population.The optimal solution of particle swarm algorithm is exactly complete
Office's extreme value.
If the judging result of this step is to meet, optimal solution, i.e. step 110 are obtained.
Otherwise, the update continued to execute to the position and speed of particle in population, i.e. return step 100 are returned.Also,
After the speed of each more new particle and position, the fitness value of particle will be recalculated.
In step 110, optimal solution is obtained.
By above-mentioned example, it will intersect and variation introduce particle swarm algorithm, improve convergence rate, and find optimal
It is more preferable to solve effect.
The particle swarm algorithm of this specification at least one embodiment is applied to the position of search target object, the mesh as follows
Mark object for example can be the life in disaster relief, such as in the application scenarios applied to flying robot's search life.
After the disasters such as earthquake, fire, mine disaster occur, positioning survivor is searched in ruins, gives necessary medical rescue,
And rescuing captive as early as possible is three urgent tasks that rescue personnel faces.Among these, the search positioning of survivor seems outstanding
It is important.Signals filtering sniffing robot can well solve the above problem.At earthquake initial stage, in order to evade the danger of earth's surface
Danger can complete the reconnaissance probe of life using flying robot.And consider the size of ruins area and the efficiency of search,
The search positioning of life can be completed using a group flying robot.Predation side of such way of search relatively similar to flock of birds
Formula, bevy only have one piece of food in random search food this region, and all birds do not know food at that
In.But they know that current position is also how far from food.Particle swarm algorithm can be utilized at this time, by using flight
Robot reconnaissance probe life, quick detection to life.
In the present example, each flying robot can have store function and communication function.
Wherein, each flying robot itself can carry out the iterative processing of particle swarm algorithm, and flying robot can be
The optimal solution oneself found is updated in every single-step iteration of algorithm, and is saved.I.e. each flying robot itself can look for
To and store individual extreme value.Therefore, robot can have the processing function of simple store function and algorithm.
In addition, according to particle swarm algorithm, each flying robot can also be according to oneself optimal solution and global optimal
Solution updates speed and the position of oneself, so robot can have simple communication function, for mutual between robot
The interaction for carrying out optimal solution, to know global extremum.Robot each in this way can be according to individual extreme value and global extremum
Renewal speed and position.
Therefore, each flying robot may include the power plant modules such as battery, motor, provide the movement function of robot
Energy.It can also include communication module, be responsible for relatively recording the position of global optimum.Memory module can be used for storing simple journey
Sequence instruction and individual extreme value, global extremum.Calculated speed can also be acted on power mould by control module by robot
Block, the flight of driving robot.
Wherein, in above-mentioned scene, the corresponding fitness function of fitness value is according to target object (for example, above-mentioned
Survivor's life) location finding application construction function, fitness value indicates the signal that the life that receives of robot issues
The size of (for example, the signal can be sound) intensity, signal is stronger, and the fitness value is bigger.When fitness value is bigger, table
The bright robot is closer apart from life, and the position of the robot is more excellent.Individual extreme value be robot for scanning for
Nearest position between life, global extremum are positions nearest between life in all robots.
Illustratively, the fitness function in this example may is that
In above formula (1), MaxZ can be fitness value, and s can be the signal strength that the robot receives,
N is the step number of the robot iteration, and i and j are the two-dimensional coordinates of the robot, and x is the position of the robot, and d is distance, xij's
Mean node i to node j, dijMean node i to node j distance.Wherein, node i and node j are robot fortune
Two dynamic positions.
For example, flying robot may be considered executes search mission in two-dimensional space, if robot group scale is
M, each robot i (1≤i≤m) is just like properties in group: xi (t)=(xi1, xi2) the location of when t walks iteration,
Flying speed vi (t)=(vi1, vi2), the position that flying robot remembers and real-time update itself is nearest with the life searched for
Pi (individual extreme value), the nearest position of the life of (entire group regards neighborhood as) of neighborhood and search where i-th of robot
Pl (global extremum).
It then, can be according to algorithm flow shown in FIG. 1, search of the Lai Zhihang robot to life.Method and step can be with
In conjunction with referring to Fig. 1, no longer it is described in detail.
Wherein, Fig. 2 is referred to, Fig. 2 is the group of flying robot when initializing, the region 21 near life in figure
Color is most deep, and the pore in figure is flying robot 22, and the depth of color indicates the life received when robot is at this
The power of signal.For example, the fitness value that the robot calculates is lower if flying robot is at color shallower region,
Indicate that the life signal received is weaker.
With the successive ignition of algorithm, each time in iteration, robot is all that the individual extreme value of tracking and global extremum come more
The newly position and speed of oneself, is equivalent to and flies towards better position.That is flying robot is in search process, by certainly
Body and group are pulled in shift position.It is shown in Figure 3, after iteration, it is distributed loose flying robot originally gradually
It is drawn close to saturate region.
It refers to shown in Fig. 4 and Fig. 5, with continuing for iteration, flying robot ground more strong to life signal
Side is drawn close, and is eventually collected in global life signal most strength, here it is the positions of global extremum.Particle swarm algorithm is received
It holds back, flying robot also has found the life of survival.
In this example embodiment, each flying robot can update oneself position and speed according to individual extreme value and global extremum
Degree intersects to executing and when the processing of variation, a central processor equipment can be arranged except these flying robots, in this
Centre processing equipment can receive the fitness value that each flying robot sends, and is ranked up according to the size of fitness value；
By selecting the biggish top n particle of fitness value to enter cistern of chiasma in ranking results.Central processor equipment can be by the intersection
It selects particle to be intersected in pond, and sends corresponding parent flying machine for the speed for intersecting obtained filial generation particle and position
Device people, to realize that filial generation particle replaces parent particle.Central processor equipment can also become the particle in the population
It is different, and particle before corresponding variation is sent by the position and speed of the particle obtained after variation, to realize that variation updates.When
So, be a kind of illustrative implement scene above, other modes can also be used, for example, by one of flying robot Lai
Execute the processing of above-mentioned central processor equipment.
Fig. 6 is a kind of position that target object is determined based on particle swarm algorithm that at least one embodiment of this specification provides
Device structural schematic diagram, as shown in fig. 6, the apparatus may include: data determining module 61, cross processing module 62, becoming
Different processing module 63 and position determination module 64.
Data determining module 61, for determining the speed of each particle and position in population；
Cross processing module 62, it is described for determining the fitness value of each particle according to the speed and position
Fitness value is related to the distance between the particle and target object；And according to the size of the fitness value, to the grain
The fitness value sequence of each particle in subgroup, and by selecting the biggish some particles of fitness value to enter intersection in ranking results
Pond, some particles are determined according to population scale and crossover probability, by selecting particle to be intersected in the cistern of chiasma；
Make a variation processing module 63, for making a variation to the particle in the population；
Position determination module 64, for when meeting the termination condition of particle swarm algorithm, by the optimal solution of particle swarm algorithm
Position of the corresponding position as the target object.
In one example, data update module 63 are specifically used for: by the fitness value of the particle and individual ratio of extreme values
Compared with if the fitness value replaces the individual extreme value greater than individual extreme value, with the position of the particle；By the particle
Fitness value compared with global extremum, if the fitness value be greater than global extremum, replaced with the position of the particle
The global extremum；According to the updated individual extreme value and global extremum, speed and the position of the particle are updated.
In one example, make a variation processing module 65, specifically for carrying out Gaussian mutation to the particle in the population.
In one example, the particle in the population is the robot for scanning for；It is described individual extreme value be
The proximal most position of the robot and target object；
The global extremum is the proximal most position of all robots and target object in population；
The fitness value indicates the size for the signal strength that the target object that the robot receives issues, signal
Stronger, the fitness value is bigger.
Each step in process shown in above method embodiment, execution sequence are not limited to suitable in flow chart
Sequence.In addition, the description of each step, can be implemented as software, hardware or its form combined, for example, those skilled in the art
Member can implement these as the form of software code, can be can be realized the computer of the corresponding logic function of the step can
It executes instruction.When it is realized in the form of software, the executable instruction be can store in memory, and by equipment
Processor execute.
The device or module that above-described embodiment illustrates can specifically realize by computer chip or entity, or by having
The product of certain function is realized.A kind of typically to realize that equipment is computer, the concrete form of computer can be personal meter
Calculation machine, laptop computer, cellular phone, camera phone, smart phone, personal digital assistant, media player, navigation are set
It is any several in standby, E-mail receiver/send equipment, game console, tablet computer, wearable device or these equipment
The combination of equipment.
For convenience of description, it is divided into various modules when description apparatus above with function to describe respectively.Certainly, implementing this
The function of each module can be realized in the same or multiple software and or hardware when specification one or more embodiment.
It should be understood by those skilled in the art that, this specification one or more embodiment can provide for method, system or
Computer program product.Therefore, complete hardware embodiment can be used in this specification one or more embodiment, complete software is implemented
The form of example or embodiment combining software and hardware aspects.Moreover, this specification one or more embodiment can be used one
It is a or it is multiple wherein include computer usable program code computer-usable storage medium (including but not limited to disk storage
Device, CD-ROM, optical memory etc.) on the form of computer program product implemented.
These computer program instructions, which may also be stored in, is able to guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works, so that it includes referring to that instruction stored in the computer readable memory, which generates,
Enable the manufacture of device, the command device realize in one box of one or more flows of the flowchart and/or block diagram or
The function of being specified in multiple boxes.
These computer program instructions also can be loaded onto a computer or other programmable data processing device, so that counting
Series of operation steps are executed on calculation machine or other programmable devices to generate computer implemented processing, thus in computer or
The instruction executed on other programmable devices is provided for realizing in one or more flows of the flowchart and/or block diagram one
The step of function of being specified in a box or multiple boxes.
It should also be noted that, the terms "include", "comprise" or its any other variant are intended to nonexcludability
It include so that the process, method, commodity or the equipment that include a series of elements not only include those elements, but also to wrap
Include other elements that are not explicitly listed, or further include for this process, method, commodity or equipment intrinsic want
Element.In the absence of more restrictions, the element limited by sentence "including a ...", it is not excluded that including described want
There is also other identical elements in the process, method of element, commodity or equipment.
This specification one or more embodiment can computer executable instructions it is general on
It hereinafter describes, such as program module.Generally, program module includes executing particular task or realization particular abstract data type
Routine, programs, objects, component, data structure etc..Can also practice in a distributed computing environment this specification one or
Multiple embodiments, in these distributed computing environments, by being executed by the connected remote processing devices of communication network
Task.In a distributed computing environment, the local and remote computer that program module can be located at including storage equipment is deposited
In storage media.
At least one embodiment of this specification additionally provides a kind of position that target object is determined based on particle swarm algorithm
Equipment, the equipment include memory, processor and storage on a memory and the computer program that can run on a processor,
The processor performs the steps of when executing described program
Initialize the position and speed of each particle in population；
Calculate the fitness value of each particle；
For each particle, by the way that the fitness value of the particle and individual extreme value and global extremum are compared
Compared with updating speed and the position of the particle；
According to the speed of updated particle and position, the fitness value of each particle is calculated, and according to the fitness
The size of value is ranked up；By selecting the biggish top n particle of fitness value to enter cistern of chiasma in ranking results, the N is by grain
Subgroup scale is obtained multiplied by crossover probability；The N is natural number；The fitness value is higher, between particle and target object
Apart from closer；
By selecting particle to be intersected in the cistern of chiasma；
It makes a variation to the particle in the population；
Judge whether the termination condition for meeting particle swarm algorithm, if satisfied, then that the optimal solution of particle swarm algorithm is corresponding
Position of the position as target object；Otherwise, the update continued to execute to the position and speed of particle in population is returned.
All the embodiments in this specification are described in a progressive manner, same and similar portion between each embodiment
Dividing may refer to each other, and each embodiment focuses on the differences from other embodiments.It is adopted especially for data
For collecting equipment or data processing equipment embodiment, since it is substantially similar to the method embodiment, so the comparison of description is simple
Single, the relevent part can refer to the partial explaination of embodiments of method.
It is above-mentioned that this specification specific embodiment is described.Other embodiments are in the scope of the appended claims
It is interior.In some cases, the movement recorded in detail in the claims or step can be come according to the sequence being different from embodiment
It executes and desired result still may be implemented.In addition, process depicted in the drawing not necessarily require show it is specific suitable
Sequence or consecutive order are just able to achieve desired result.In some embodiments, multitasking and parallel processing be also can
With or may be advantageous.
The foregoing is merely the preferred embodiments of this specification one or more embodiment, not to limit this public affairs
It opens, all within the spirit and principle of the disclosure, any modification, equivalent substitution, improvement and etc. done should be included in the disclosure
Within the scope of protection.
Claims (12)
1. a kind of method for the position for determining target object based on particle swarm algorithm, which comprises
Determine the speed of each particle and position in population；
According to the speed and position, the fitness value of each particle, the fitness value and the particle and mesh are determined
Mark the distance between object correlation；
It according to the size of the fitness value, sorts to the fitness value of each particle in the population, and by ranking results
The biggish some particles of middle selection fitness value enter cistern of chiasma, and some particles are true according to population scale and crossover probability
It is fixed, by selecting particle to be intersected in the cistern of chiasma；
It makes a variation to the particle in the population；
When meeting the termination condition of particle swarm algorithm, using the corresponding position of the optimal solution of particle swarm algorithm as the target pair
The position of elephant.
2. according to the method described in claim 1, described by selecting the biggish some particles of fitness value to enter in ranking results
Cistern of chiasma, some particles are determined according to population scale and crossover probability, comprising:
By selecting the biggish top n particle of fitness value to enter cistern of chiasma in ranking results；
The N is to be obtained by population scale multiplied by crossover probability.
3. according to the method described in claim 1,
The speed of each particle and position in the determining population, comprising:
Initialize the position and speed of each particle in population；
Calculate the fitness value of each particle；
For each particle, by the way that the fitness value of the particle to be compared with individual extreme value and global extremum, more
The speed of the new particle and position；
The fitness value that each particle is determined according to the speed and position, specifically: according to updated particle
Speed and position, calculate the fitness value.
4. according to the method described in claim 3, each particle passes through by the fitness value of oneself and individual extreme value and entirely
Office's extreme value is compared, and updates speed and the position of the particle, comprising:
By the fitness value of the particle compared with individual extreme value, if the fitness value is greater than individual extreme value, with the grain
The position of son replaces the individual extreme value；
By the fitness value of the particle compared with global extremum, if the fitness value is greater than global extremum, with the grain
The position of son replaces the global extremum；
According to the updated individual extreme value and global extremum, speed and the position of the particle are updated.
5. according to the method described in claim 3, the particle in the population is the robot for searching for target object；
The individual extreme value is the proximal most position of the robot and target object；
The global extremum is the proximal most position of all robots and target object in population；
The fitness value indicates the size for the signal strength that the target object that the robot receives issues, and signal is stronger,
The fitness value is bigger.
6. according to the method described in claim 1, the particle in the population makes a variation, specifically: to described
Particle in population carries out Gaussian mutation.
7. according to the method described in claim 1, described by selecting particle to be intersected in cistern of chiasma, comprising:
It is random by selecting two different particles in the cistern of chiasma, intersected as parent particle；
Fitness value is calculated to the filial generation particle obtained after intersection；
If the fitness value of the filial generation particle is higher than the fitness value of parent particle, parent grain is replaced by the filial generation particle
Son.
8. a kind of device for the position for determining target object based on particle swarm algorithm, described device include:
Data determining module, for determining the speed of each particle and position in population；
Cross processing module, for determining the fitness value of each particle, the fitness according to the speed and position
It is worth related to the distance between the particle and target object；And according to the size of the fitness value, in the population
The fitness value of each particle sorts, and by selecting the biggish some particles of fitness value to enter cistern of chiasma in ranking results, institute
It states some particles to be determined according to population scale and crossover probability, by selecting particle to be intersected in the cistern of chiasma；
Make a variation processing module, for making a variation to the particle in the population；
Position determination module, it is for when meeting the termination condition of particle swarm algorithm, the optimal solution of particle swarm algorithm is corresponding
Position of the position as the target object.
9. device according to claim 8,
The data determining module, when for determining the speed of each particle and position in population, comprising: initialization particle
The position and speed of each particle in group；Calculate the fitness value of each particle；For each particle, by by institute
The fitness value for stating particle is compared with individual extreme value and global extremum, updates speed and the position of the particle；Described
The fitness value of each particle is determined according to the speed and position, specifically: according to the speed of updated particle and position
It sets, calculates the fitness value.
10. device according to claim 9, the particle in the population is the robot for scanning for；It is described
Individual extreme value is the proximal most position of the robot and target object；
The global extremum is the proximal most position of all robots and target object in population；
The fitness value indicates the size for the signal strength that the target object that the robot receives issues, and signal is stronger,
The fitness value is bigger.
11. device according to claim 8,
The variation processing module, specifically for carrying out Gaussian mutation to the particle in the population.
12. a kind of equipment for the position for determining target object based on particle swarm algorithm, the equipment include memory, processor and
The computer program that can be run on a memory and on a processor is stored, the processor realizes right when executing described program
It is required that 1 to 7 any method and step.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910181971.3A CN109961130A (en) | 2019-03-11 | 2019-03-11 | A kind of method and apparatus that target object position is determined based on particle swarm algorithm |
| PCT/CN2020/073825 WO2020181934A1 (en) | 2019-03-11 | 2020-01-22 | Method and device for determining position of target object on the basis of particle swarm algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910181971.3A CN109961130A (en) | 2019-03-11 | 2019-03-11 | A kind of method and apparatus that target object position is determined based on particle swarm algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN109961130A true CN109961130A (en) | 2019-07-02 |
Family
ID=67024207
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910181971.3A Pending CN109961130A (en) | 2019-03-11 | 2019-03-11 | A kind of method and apparatus that target object position is determined based on particle swarm algorithm |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN109961130A (en) |
| WO (1) | WO2020181934A1 (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020181934A1 (en) * | 2019-03-11 | 2020-09-17 | 阿里巴巴集团控股有限公司 | Method and device for determining position of target object on the basis of particle swarm algorithm |
| CN112862057A (en) * | 2021-04-07 | 2021-05-28 | 京东数字科技控股股份有限公司 | Modeling method, modeling device, electronic equipment and readable medium |
| CN112965530A (en) * | 2021-02-09 | 2021-06-15 | 辽宁警察学院 | Multi-unmanned aerial vehicle self-adaptive variable-scale dynamic target searching method |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120084742A1 (en) * | 2010-09-30 | 2012-04-05 | Ispir Mustafa | Method and apparatus for using entropy in ant colony optimization circuit design from high level synthesis |
| CN107506821A (en) * | 2017-10-13 | 2017-12-22 | 集美大学 | A kind of improved particle group optimizing method |
| CN108399450A (en) * | 2018-02-02 | 2018-08-14 | 武汉理工大学 | Improvement particle cluster algorithm based on biological evolution principle |
| CN108846472A (en) * | 2018-06-05 | 2018-11-20 | 北京航空航天大学 | A kind of optimization method of Adaptive Genetic Particle Swarm Mixed Algorithm |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8620631B2 (en) * | 2011-04-11 | 2013-12-31 | King Fahd University Of Petroleum And Minerals | Method of identifying Hammerstein models with known nonlinearity structures using particle swarm optimization |
| CN104638637A (en) * | 2014-12-08 | 2015-05-20 | 国家电网公司 | Coordinative optimization control method based on AGC and AVC |
| CN106126479B (en) * | 2016-07-07 | 2019-04-12 | 重庆邮电大学 | Order Oscillating population blind source separation method based on hereditary variation optimization |
| CN106502092B (en) * | 2016-10-21 | 2019-05-31 | 东南大学 | A kind of thermal process model parameter identification method using improvement Hybrid Particle Swarm |
| CN107253442B (en) * | 2017-06-21 | 2019-10-25 | 太原科技大学 | A method of braking force distribution in optimization Electro-hydraulic brake system |
| CN109039173A (en) * | 2018-08-09 | 2018-12-18 | 沈阳工业大学 | A kind of PMLSM iterative learning control method and system based on hybridization particle group optimizing |
| CN109961130A (en) * | 2019-03-11 | 2019-07-02 | 阿里巴巴集团控股有限公司 | A kind of method and apparatus that target object position is determined based on particle swarm algorithm |
-
2019
- 2019-03-11 CN CN201910181971.3A patent/CN109961130A/en active Pending
-
2020
- 2020-01-22 WO PCT/CN2020/073825 patent/WO2020181934A1/en active Application Filing
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20120084742A1 (en) * | 2010-09-30 | 2012-04-05 | Ispir Mustafa | Method and apparatus for using entropy in ant colony optimization circuit design from high level synthesis |
| CN107506821A (en) * | 2017-10-13 | 2017-12-22 | 集美大学 | A kind of improved particle group optimizing method |
| CN108399450A (en) * | 2018-02-02 | 2018-08-14 | 武汉理工大学 | Improvement particle cluster algorithm based on biological evolution principle |
| CN108846472A (en) * | 2018-06-05 | 2018-11-20 | 北京航空航天大学 | A kind of optimization method of Adaptive Genetic Particle Swarm Mixed Algorithm |
Non-Patent Citations (3)
| Title |
|---|
| 专祥涛: "《最优化方法基础》", 31 March 2018 * |
| 雷斌, 李文锋: ""基于粒子群优化的多机器人合作目标搜索算法"", 《武汉理工大学学报》 * |
| 靳志宏，计明军: "《现代优化技术》", 28 February 2017 * |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2020181934A1 (en) * | 2019-03-11 | 2020-09-17 | 阿里巴巴集团控股有限公司 | Method and device for determining position of target object on the basis of particle swarm algorithm |
| CN112965530A (en) * | 2021-02-09 | 2021-06-15 | 辽宁警察学院 | Multi-unmanned aerial vehicle self-adaptive variable-scale dynamic target searching method |
| CN112965530B (en) * | 2021-02-09 | 2024-03-19 | 辽宁警察学院 | Multi-unmanned aerial vehicle self-adaptive variable-scale dynamic target searching method |
| CN112862057A (en) * | 2021-04-07 | 2021-05-28 | 京东数字科技控股股份有限公司 | Modeling method, modeling device, electronic equipment and readable medium |
| CN112862057B (en) * | 2021-04-07 | 2023-11-03 | 京东科技控股股份有限公司 | Modeling method, modeling device, electronic equipment and readable medium |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2020181934A1 (en) | 2020-09-17 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN107169608B (en) | Distribution method and device for multiple unmanned aerial vehicles to execute multiple tasks | |
| CN107103164B (en) | Distribution method and device for unmanned aerial vehicle to execute multiple tasks | |
| CN112307622B (en) | Autonomous planning system and planning method for generating force by computer | |
| Yu et al. | A knee-guided differential evolution algorithm for unmanned aerial vehicle path planning in disaster management | |
| Debnath et al. | A review on graph search algorithms for optimal energy efficient path planning for an unmanned air vehicle | |
| Wang et al. | Efficient object search with belief road map using mobile robot | |
| CN109961130A (en) | A kind of method and apparatus that target object position is determined based on particle swarm algorithm | |
| Cagnoni et al. | Genetic and evolutionary computation for image processing and analysis | |
| Wang et al. | Rank-driven salp swarm algorithm with orthogonal opposition-based learning for global optimization | |
| CN109269502A (en) | A kind of no-manned plane three-dimensional Route planner based on more stragetic innovation particle swarm algorithms | |
| CN109242290B (en) | Automatic generation method for unmanned aerial vehicle group action scheme | |
| Xu et al. | Explore-bench: Data sets, metrics and evaluations for frontier-based and deep-reinforcement-learning-based autonomous exploration | |
| CN106295793A (en) | Group robot mixed search algorithm based on biological foraging behavior | |
| Hamami et al. | A systematic review on particle swarm optimization towards target search in the swarm robotics domain | |
| Yu et al. | Balanced multi-region coverage path planning for unmanned aerial vehicles | |
| Parvez et al. | Path planning optimization using genetic algorithm | |
| Kareem et al. | Planning the Optimal 3D Quadcopter Trajectory Using a Delivery System-Based Hybrid Algorithm. | |
| Hu et al. | Enhanced multi-strategy bottlenose dolphin optimizer for UAVs path planning | |
| Makarov et al. | Voronoi-based Path Planning based on Visibility and Kill/Death RatioTactical Component. | |
| Kim et al. | Reducing the search space for pathfinding in navigation meshes by using visibility tests | |
| Mishra et al. | Weapon target assignment problem: multi-objective formulation, optimisation using MOPSO and TOPSIS | |
| Von Mammen et al. | An organic computing approach to self-organizing robot ensembles | |
| Nategh et al. | University-timetabling problem and its solution using GELS algorithm: a case study | |
| CN110501905A (en) | Multi-agent system self-adaptive method and system based on packing model | |
| Hornby et al. | Learning comparative user models for accelerating human-computer collaborative search |
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 | ||
| REG | Reference to a national code |
Ref country code: HK Ref legal event code: DE Ref document number: 40010418 Country of ref document: HK |
|
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190702 |
|
| RJ01 | Rejection of invention patent application after publication |