CN108062585A - A kind of method that Function Extreme value is calculated based on a flying moth darts into the fire algorithm - Google Patents

A kind of method that Function Extreme value is calculated based on a flying moth darts into the fire algorithm Download PDF

Info

Publication number
CN108062585A
CN108062585A CN201711480246.3A CN201711480246A CN108062585A CN 108062585 A CN108062585 A CN 108062585A CN 201711480246 A CN201711480246 A CN 201711480246A CN 108062585 A CN108062585 A CN 108062585A
Authority
CN
China
Prior art keywords
mrow
moth
flameno
algorithm
flame
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
Application number
CN201711480246.3A
Other languages
Chinese (zh)
Inventor
王平
石晓飞
白蛟
肖楠
张彩娜
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Casic Wisdom Industrial Development Co Ltd
Original Assignee
Casic Wisdom Industrial Development Co Ltd
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by Casic Wisdom Industrial Development Co Ltd filed Critical Casic Wisdom Industrial Development Co Ltd
Priority to CN201711480246.3A priority Critical patent/CN108062585A/en
Publication of CN108062585A publication Critical patent/CN108062585A/en
Pending legal-status Critical Current

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)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Health & Medical Sciences (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The invention discloses a kind of methods that Function Extreme value is calculated based on a flying moth darts into the fire algorithm, are related to function optimization solution technique field.This method, by using backward learning, to the prior art, a flying moth darts into the fire that algorithm is improved, add the number of Candidate Set, remain elite, so that result of calculation is closer to accurately solve, moreover, using method provided in an embodiment of the present invention, convergence precision and convergence rate are in the utilization of some functions also apparently higher than a flying moth darts into the fire algorithm of the prior art.

Description

A kind of method that Function Extreme value is calculated based on a flying moth darts into the fire algorithm
Technical field
The present invention relates to function optimization solution technique field more particularly to one kind, based on a flying moth darts into the fire, algorithm calculates function most The method of value.
Background technology
At present, as technique becomes increasingly complex, calculation amount is increasing, and the complexity being related to is higher and higher.If using Traditional auxiliary tool is calculated, and not only can greatly be lost time, and increases the complexity of work, can also expend substantial amounts of essence Power, material resources, financial resources.And with the development of big data technology, it has been obtained very in engineering optimization based on data mining scheduling algorithm It is widely applied.
Engineering optimization is engineering mathematics problem, and optimization process is exactly that the process of optimal solution is found to particular problem. Optimization method can substantially be divided into two major class of deterministic optimization and random optimization.Though deterministic optimization studies relative maturity, There is more exacting terms to engineer application, so, randomized optimization process is rapidly developed.
Swarm intelligence algorithm is the current more extensive randomized optimization process of research, including more classical ant group algorithm, is lost Propagation algorithm, particle cluster algorithm etc., each colony intelligence optimization algorithm have its unique advantage.
The biologies such as the insect, the animal that live in nature, they complete huge task by cooperative cooperating.People are therefrom It is inspired, the insect in group, social behavior in animal, group life habit in mimic biology system has invented group Body intelligent optimization algorithm.
Colony intelligence optimization algorithm is in simulation biocenose life, and individual and mutually exchanging and cooperating between individual are used Simply, limited individual behavior and intelligence by cooperating with each other, form the powerful ability of group.Swarm intelligence algorithm is simple, robust Property is strong, scalability is high, has many advantages, such as wide applicability.Algorithm can be solved quickly in the case where lacking global information Certainly various challenges become the common method for solving challenge and optimization problem, receive every field, domestic and foreign scholars Extensive concern and utilization.Such as patent《A kind of new colony intelligence optimization algorithm-dove group's algorithm》,《Based on the automatic of genetic algorithm Floor truck dispatch control method》Deng, wherein,《A kind of new colony intelligence optimization algorithm-dove group's algorithm》It is slow there are convergence rate, Convergence precision is low, the problem of being easily trapped into local optimum,《Automatic transporting vehicle dispatching control method based on genetic algorithm》It deposits In higher algorithm complexity, the defects of convergence rate is slow.
A flying moth darts into the fire, and algorithm (MFO) is that the algorithm generated is inspired by moth, as one kind of swarm intelligence algorithm, is being solved There is preferable application effect in Function Extreme value problem and some industrial problems.But a flying moth darts into the fire algorithm and most of groups Intelligent optimization algorithm is equally of problems identical, is easily trapped into Local Extremum, later stage of evolution convergence rate is slow, precision is poor Deng.
The content of the invention
It is existing so as to solve it is an object of the invention to provide a kind of method that Function Extreme value is calculated based on a flying moth darts into the fire algorithm There are foregoing problems present in technology.
To achieve these goals, the technical solution adopted by the present invention is as follows：
A kind of method that Function Extreme value is calculated based on a flying moth darts into the fire algorithm, is included the following steps：
S1, using a flying moth darts into the fire, algorithm generates n RANDOM SOLUTION；
S2, obtain n RANDOM SOLUTION using backward learning method n reversely solve：
S3, n RANDOM SOLUTIONs and n reversed solutions form 2n Candidate Set, and preferably n is used as initialization in 2n Candidate Set Moth；
S4 using the moth population of initialization, is iterated function processing, if iterations be less than T, calculate obtain with The flame flameno of machine selection simultaneously performs S5, otherwise performs S7, T is maximum iteration；
S5 compares the relation between the moth number i flame numbers flameno of current iteration processing, if i≤ Flameno obtains locally optimal solution in a manner that spiral is circled in the air；If i ＞ flameno, randomly selected is utilized Flameno flame, the method approached using spiral are carried out global exploration, the operation are repeated, until i=n；
S6 repeats S4-S5, until iterations is equal to T；
S7 returns to adaptive value.
Preferably, the mathematical model of the backward learning is：
x*=ub+lb-x,
In formula, x is RANDOM SOLUTION, x*It is the reversed solution of x, ub, lb are the corresponding maximums of x and minimum value respectively.
Preferably, in S3, the preferably n moth population as initialization in 2n Candidate Set, using such as lower section Method is implemented：2n Candidate Set is substituted into function respectively, solves corresponding adaptive value, chooses n minimum adaptive value pair The n Candidate Set answered, the moth population as initialization.
Preferably, it is described to calculate the flame flameno for obtaining and randomly selecting in S4, it is calculated according to equation below：
Wherein, l is current iteration number, and N is the maximum of flame quantity, and T is maximum iteration.
Preferably, in S5, if the i≤flameno, locally optimal solution is obtained in a manner that spiral is circled in the air, specifically To be calculated using equation below：
Mi=S (Mi,Fj),
S(Mi,Fj)=Di·ebt·cos(2πt)+Fj,
Di=| Fj-Mi|,
Wherein, MiRepresent the position of i-th moth, FjRepresent the position of j-th of flame, S is Spirallike Functions；DiRepresent i-th The distance between a moth and j-th flame, b are constant, a random number between t ∈ [- 1,1].
Preferably, in a flying moth darts into the fire the algorithm, adaptive weighting is introduced, then according to following adaptive weighting mathematics Model：
Obtain following improved Spirallike Functions：
S(Mi,Fj)=wDi·ebt·cos(2πt)+Fj,
Wherein, wmaxFor constant, 0.9, w is traditionally arranged to bemin0.4, Iter is arranged to as current iteration number, w is adaptive Answer weights.
Preferably, in S5, if the i ＞ flameno, using the flameno flames randomly selected, are forced using spiral Near method carries out global exploration, specifically, being calculated using equation below：
Mi=S (Mi,Fj),
S(Mi,Fj)=Di·ebt·cos(2πt)+Fflameno,
Di=| Fj-Mi|,
Wherein, MiRepresent the position of i-th moth, FflamenoRepresent the flameno flame number randomly selected, S is spiral shell Revolve function；DiRepresent the distance between i-th moth and j-th flame, b is constant, and one between t ∈ [- 1,1] is random Number.
Preferably, in a flying moth darts into the fire the algorithm, adaptive weighting is introduced, then according to following adaptive weighting mathematics Model：
Obtain following improved Spirallike Functions：
S(Mi,Fj)=wDi·ebt·cos(2πt)+Fflameno,
Wherein, wmaxFor constant, 0.9, w is traditionally arranged to bemin0.4, Iter is arranged to as current iteration number, w is adaptive Answer weights.
The beneficial effects of the invention are as follows：The side provided in an embodiment of the present invention that Function Extreme value is calculated based on a flying moth darts into the fire algorithm Method, by using backward learning, to the prior art, a flying moth darts into the fire that algorithm is improved, and adds the number of Candidate Set, retains Elite, so that result of calculation is closer to accurately solve, moreover, using method provided in an embodiment of the present invention, convergence Precision and convergence rate are in the utilization of some functions also apparently higher than a flying moth darts into the fire algorithm of the prior art.
Description of the drawings
Fig. 1 is the method flow schematic diagram provided in an embodiment of the present invention that Function Extreme value is calculated based on a flying moth darts into the fire algorithm；
Fig. 2 is the schematic diagram for the function adaptive value not obtained using backward learning method initialization population；
Fig. 3 is the schematic diagram of the function adaptive value obtained using backward learning method initialization population；
Fig. 4 is the schematic diagram for the function adaptive value not obtained using adaptive weighting；
Fig. 5 is the schematic diagram of the function adaptive value obtained using adaptive weighting；
Fig. 6 is function f in the present invention1Convergence curve；
Fig. 7 is function f in the present invention2Convergence curve；
Fig. 8 is function f in the present invention3Convergence curve；
Fig. 9 is function f in the present invention5Convergence curve.
Specific embodiment
In order to make the purpose , technical scheme and advantage of the present invention be clearer, below in conjunction with attached drawing, to the present invention into Row is further described.It should be appreciated that the specific embodiments described herein are only used to explain the present invention, it is not used to Limit the present invention.
As shown in Figure 1, an embodiment of the present invention provides a kind of method that Function Extreme value is calculated based on a flying moth darts into the fire algorithm, bags Include following steps：
S1, using a flying moth darts into the fire, algorithm generates n RANDOM SOLUTION；
S2, obtain n RANDOM SOLUTION using backward learning method n reversely solve；
S3, n RANDOM SOLUTIONs and n reversed solutions form 2n Candidate Set, and preferably n is used as initialization in 2n Candidate Set Moth；
S4 using the moth population of initialization, is iterated function processing, if iterations be less than T, calculate obtain with The flame flameno of machine selection simultaneously performs S5, otherwise performs S7, T is maximum iteration；
S5 compares the relation between the moth number i and flame number flameno of current iteration processing, if i≤ Flameno obtains locally optimal solution in a manner that spiral is circled in the air；If i ＞ flameno, randomly selected is utilized Flameno flames, the method approached using spiral are carried out global exploration, the operation are repeated, until i=n；
S6 repeats S4-S5, until iterations is equal to T；
S7 returns to adaptive value.
A flying moth darts into the fire, and algorithm can randomly generate initialization population, but the quantity of the initialization population randomly generated compares It is few.Therefore, relatively good candidate solution can be missed, in order to solve this problem, chooses preferably initialization moth population, the present invention is real In the above method that example offer is provided, the method initialization moth population of backward learning is employed, is obtained by the method for backward learning To initialization moth population, the quantity of the moth population formed together with original population increases 1 times, if, it is random to produce Raw initialization population quantity is n, then after initializing moth population using backward learning method, obtained population quantity is just 2n, then from 2n population can preferably n as final initialization moth population.
So compared with a flying moth darts into the fire algorithm of the prior art, in the present invention, by using backward learning method to Machine group is handled, and after obtaining double initialization population quantity, then therefrom preferably goes out the Candidate Set of identical quantity as most Whole population, improves the diversity of initialization moth population, while saves the elite of population.
In the embodiment of the present invention, by using functionThe effect of the above method is analyzed, Fig. 2 is not have The volume function adaptive value distribution map obtained using backward learning method, Fig. 3 are that the function obtained using backward learning method is adapted to Distribution value figure.
As can be seen that the function adaptive value fluctuation range obtained using backward learning method is [0,8e from Fig. 2 and Fig. 3-33] in, function adaptive value is more nearly globally optimal solution 0, the fluctuation for the function adaptive value not obtained using backward learning method Scope is in [0,6.5e4] in, with using compared with backward learning method, not using backward learning method when, function adaptive value is from most Excellent solution is farther.
As it can be seen that using backward learning method function adaptive value can be caused to be more nearly globally optimal solution.
Wherein, the mathematical model of the backward learning can be：
x*=ub+lb-x,
In formula, x is RANDOM SOLUTION, x*It is the reversed solution of x, ub, lb are the corresponding maximums of x and minimum value respectively.
In a preferred embodiment of the present invention, in S3, described preferred in 2n Candidate Set n flies as initialization Moth population may be employed following method and be implemented：2n Candidate Set is substituted into function respectively, solves corresponding adaptive value, Choose the n minimum corresponding n Candidate Set of adaptive value, the moth population as initialization.
In the embodiment of the present invention, in S4, described calculate obtains the flame flameno that randomly selects, according to equation below into Row calculates：
Wherein, l is current iteration number, and N is the maximum of flame quantity, and T is maximum iteration.
In a flying moth darts into the fire algorithm, if moth is only moved towards flame, can cause MFO algorithms be absorbed in quickly it is local most It is excellent, in order to avoid such case, one can be chosen for each moth and update their position with reference to flame.Iteration each time After updating flame, flame sorts according to their fitness value, rear moth towards corresponding update flame location updating they The position of oneself, first moth always compared with optimal flame location updating oneself position, and that last moth Compared with the position of the location updating oneself of worst flame, it can thus make the quantity of flame adaptation value in an iterative process Reduce.According to equation below, the position of the reference flame flameno that randomly selects can be calculated：
Wherein, l is current iteration number, and N is the maximum of flame quantity, and T is maximum iteration.
At the beginning of iteration, moth carries out global exploration, then selects local optimal searching, and in the last of iteration, fly Moth updates their position using only best flame, and global optimum is obtained, so as to which MFO algorithms be avoided to be absorbed in part quickly most It is excellent.
In the embodiment of the present invention, in S5, if the i≤flameno, local optimum is obtained in a manner that spiral is circled in the air Solution, specifically, being calculated using equation below：
Mi=S (Mi,Fj),
S(Mi,Fj)=Di·ebt·cos(2πt)+Fj,
Di=| Fj-Mi|,
Wherein, MiRepresent the position of i-th moth, FjRepresent the position of j-th of flame, S is Spirallike Functions；DiRepresent i-th The distance between a moth and j-th flame, b are constant, a random number between t ∈ [- 1,1].
MFO algorithms can randomly generate solution space, and solution space corresponds to moth population, therefore moth population is candidate solution, use Matrix M represents the set of candidate solution, and matrix is as follows：
Wherein, n represents the quantity of moth, and d is the dimension of variable.
Flame matrix is represented with using F, and dimension and M are equal, and matrix represents as follows：
Wherein, n represents the quantity of moth, and d is the dimension of variable.
OF is the adaptive value of moth, during a flying moth darts into the fire can more new position, so as to obtain adaptive value, adapt to value matrix such as Shown in lower：
MFO algorithms are to be similar to global optimal triple in optimization problem：
MFO=(I, P, T)
Wherein, I is a function for seeking moth adaptive value, and P is search function of the moth in search space, and T is function Return value, condition meets, and T function returns true；It is unsatisfactory for, T function returns to vacation, and mathematical model displaying is as follows respectively：
I:φ→{M,OM}
P:M→M
T:M→{true,false}
A flying moth darts into the fire, using Spirallike Functions, updates each moth compared with the position of flame, spiral using vrille The mathematical model of function is as follows：
Mi=S (Mi,Fj)
Wherein, MiRepresent the position of i-th moth, FjRepresenting the position of j-th of flame, S is Spirallike Functions, wherein, spiral Function is as follows：
S(Mi,Fj)=Di·ebt·cos(2πt)+Fj
Wherein, DiRepresent the distance between i-th moth and j-th flame, b is constant, one between t ∈ [- 1,1] Random number, wherein, DiMathematical expression show it is as follows：
Di=| Fj-Mi|
Wherein, MiRepresent the position of i-th of moth, FjRepresent the position of j-th of flame, DiRepresent i-th of moth and jth The distance between a flame.DiSolution formula simulate moth flight path, determined by being calculated with the distance of flame Next position.In this equation, wherein t represents that (t=-1 represents closest to the next degree for being closely located to flame of moth Flame, and t=1 explanations are farthest apart from flame).
In a flying moth darts into the fire the algorithm, adaptive weighting is introduced, then according to following adaptive weighting mathematical model：
Obtain following improved Spirallike Functions：
S(Mi,Fj)=wDi·ebt·cos(2πt)+Fj,
Wherein, wmaxFor constant, 0.9, w is traditionally arranged to bemin0.4, Iter is arranged to as current iteration number, w is adaptive Answer weights.
In particle cluster algorithm, in order to balance global exploration and local optimal searching ability, inertia weight is introduced, experiment shows There is good effect using the particle swarm optimization algorithm of inertia weight.
In the embodiment of the present invention, in a flying moth darts into the fire algorithm, adaptive weighting is introduced, when moth is remote apart from flame distance When, the speed of moth flight is fast, accelerates global exploration ability, when moth apart from flame apart from it is near when, moth flight Speed it is slow so that local optimal searching is more accurate.
The it is proposed of adaptive weighting is in order to enable moth is easier to search out locally optimal solution, can be seen by Fig. 4 Go out, not using adaptive weighting a flying moth darts into the fire algorithm, adaptive value is more scattered, illustrate if using adaptively weighing Weight, function local optimal searching ability are poor.As seen in Figure 5, using adaptive weighting a flying moth darts into the fire algorithm, adaptive value is more Add concentration, illustrate that, using adaptive weighting, moth local optimal searching ability is enhanced.
In S5, if the i ＞ flameno, using the flameno flames randomly selected, the side approached using spiral Method carries out global exploration, specifically, being calculated using equation below：
Mi=S (Mi,Fj),
S(Mi,Fj)=Diebt·cos(2πt)+Fflameno,
Di=| Fj-Mi|,
Wherein, MiRepresent the position of i-th moth, FflamenoRepresent the position of the flameno flame randomly selected, S For Spirallike Functions；DiRepresent the distance between i-th moth and j-th flame, b is constant, one between t ∈ [- 1,1] with Machine number.
In a flying moth darts into the fire the algorithm, adaptive weighting is introduced, then according to following adaptive weighting mathematical model：
Obtain following improved Spirallike Functions：
S(Mi,Fj)=wDi·ebt·cos(2πt)+Fflameno,
Wherein, wmaxFor constant, 0.9, w is traditionally arranged to bemin0.4, Iter is arranged to as current iteration number, w is adaptive Answer weights.
In order to test the validity of method provided in an embodiment of the present invention, 16 test functions, test function bag are had chosen Unimodal function, Solving Multimodal Function and fixed dimension function are included, test function is as shown in table 1-3：
1 unimodal function of table
2 Solving Multimodal Function of table
3 fixed dimension function of table
Using method provided in an embodiment of the present invention to all equal iteration of function 1000 times, population scale is arranged to 30, Using window7 64bit systems, experiment simulation is carried out using matlab2015b, the results are shown in Table 4：
4 experimental result of table
As can be seen from Table 4, method provided in an embodiment of the present invention (improved a flying moth darts into the fire algorithm, OWMFO) is with flying Moth algorithm (MFO) of putting out the fire compares, function f1~f3, f5, f7, f9, f10, f13, f15Effect of optimization obtained apparent improvement, And f6, f8, f11, f16The effect of several functions remains basically stable.From standard deviation as can be seen that except f7In addition, remaining several function Stability be improved.
So with a flying moth darts into the fire algorithm comparison of the prior art, using method provided in an embodiment of the present invention, in function There is better effect of optimization in the solution being most worth.
Fig. 6-Fig. 9 respectively shows function f1、f2、f3、f5The effect of optimization of OWMFO algorithms and MFO algorithms is respectively adopted Figure it can be seen from the figure that the adaptive value obtained by OWMFO algorithms will be less than MFO algorithms, will also realize that OWMFO algorithms exist as a result, The most value of function has better effect of optimization in solving.
By using above-mentioned technical proposal disclosed by the invention, following beneficial effect has been obtained：The embodiment of the present invention carries The method that Function Extreme value is calculated based on a flying moth darts into the fire algorithm supplied, by using backward learning, to the prior art, a flying moth darts into the fire Algorithm is improved, and is added the number of Candidate Set, is remained elite, so that result of calculation is closer to accurately solve, Moreover, using method provided in an embodiment of the present invention, convergence precision and the convergence rate also apparent height in the utilization of some functions In a flying moth darts into the fire algorithm of the prior art.
The above is only the preferred embodiment of the present invention, it is noted that for the ordinary skill people of the art For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications also should Depending on protection scope of the present invention.

Claims (8)

  1. A kind of 1. method that Function Extreme value is calculated based on a flying moth darts into the fire algorithm, which is characterized in that include the following steps：
    S1, using a flying moth darts into the fire, algorithm generates n RANDOM SOLUTION；
    S2, obtain n RANDOM SOLUTION using backward learning method n reversely solve：
    S3, n RANDOM SOLUTIONs and n reversed solutions form 2n Candidate Set, and preferably n flies as initialization in 2n Candidate Set Moth；
    S4 using the moth population of initialization, is iterated function processing, if iterations is less than T, calculates and obtains random choosing The flame flameno that takes simultaneously performs S5, otherwise performs S7, T is maximum iteration；
    S5 compares the relation between the moth number i flame numbers flameno of current iteration processing, if i≤flameno, adopts The mode circled in the air with spiral obtains locally optimal solution；If i ＞ flameno, using the flameno flame randomly selected, The method approached using spiral carries out global exploration, the operation is repeated, until i=n；
    S6 repeats S4-S5, until iterations is equal to T；
    S7 returns to adaptive value.
  2. 2. the method according to claim 1 that Function Extreme value is calculated based on a flying moth darts into the fire algorithm, which is characterized in that described anti- It is to mathematical model of learning：
    x*=ub+lb-x,
    In formula, x is RANDOM SOLUTION, x*It is the reversed solution of x, ub, lb are the corresponding maximums of x and minimum value respectively.
  3. 3. the method according to claim 1 that Function Extreme value is calculated based on a flying moth darts into the fire algorithm, which is characterized in that in S3, The preferably n moth population as initialization in 2n Candidate Set, is implemented with the following method：By 2n candidate Collection is substituted into function respectively, solves corresponding adaptive value, the n minimum corresponding n Candidate Set of adaptive value is chosen, as initial The moth population of change.
  4. 4. the method according to claim 1 that Function Extreme value is calculated based on a flying moth darts into the fire algorithm, which is characterized in that in S4, It is described to calculate the flame flameno for obtaining and randomly selecting, it is calculated according to equation below：
    <mrow> <mi>f</mi> <mi>l</mi> <mi>a</mi> <mi>m</mi> <mi>e</mi> <mi>n</mi> <mi>o</mi> <mo>=</mo> <mi>r</mi> <mi>o</mi> <mi>u</mi> <mi>n</mi> <mi>d</mi> <mrow> <mo>(</mo> <mi>N</mi> <mo>-</mo> <mi>l</mi> <mo>*</mo> <mfrac> <mrow> <mi>N</mi> <mo>-</mo> <mn>1</mn> </mrow> <mi>T</mi> </mfrac> <mo>)</mo> </mrow> <mo>,</mo> </mrow>
    Wherein, l is current iteration number, and N is the maximum of flame quantity, and T is maximum iteration.
  5. 5. the method according to claim 1 that Function Extreme value is calculated based on a flying moth darts into the fire algorithm, which is characterized in that in S5, If i≤the flameno obtains locally optimal solution in a manner that spiral is circled in the air, specifically, being counted using equation below It calculates：
    Mi=S (Mi,Fj),
    S(Mi,Fj)=Di·ebt·cos(2πt)+Fj,
    Di=| Fj-Mi|,
    Wherein, MiRepresent the position of i-th moth, FjRepresent the position of j-th of flame, S is Spirallike Functions；DiIt represents to fly for i-th The distance between moth and j-th flame, b are constant, a random number between t ∈ [- 1,1].
  6. 6. the method according to claim 5 that Function Extreme value is calculated based on a flying moth darts into the fire algorithm, which is characterized in that described In a flying moth darts into the fire algorithm, adaptive weighting is introduced, then according to following adaptive weighting mathematical model：
    <mrow> <mi>w</mi> <mo>=</mo> <mfrac> <mrow> <msub> <mi>w</mi> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> </mrow> </msub> <mo>-</mo> <mrow> <mo>(</mo> <msub> <mi>w</mi> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> </mrow> </msub> <mo>-</mo> <msub> <mi>w</mi> <mrow> <mi>m</mi> <mi>i</mi> <mi>n</mi> </mrow> </msub> <mo>)</mo> </mrow> </mrow> <mrow> <mi>I</mi> <mi>t</mi> <mi>e</mi> <mi>r</mi> </mrow> </mfrac> <mo>,</mo> </mrow>
    Obtain following improved Spirallike Functions：
    S(Mi,Fj)=wDi·ebt·cos(2πt)+Fj,
    Wherein, wmaxFor constant, 0.9, w is traditionally arranged to bemin0.4, Iter is arranged to as current iteration number, w is adaptively to weigh Value.
  7. 7. the method according to claim 1 that Function Extreme value is calculated based on a flying moth darts into the fire algorithm, which is characterized in that in S5, If the i ＞ flameno, using the flameno flames randomly selected, the overall situation is carried out using the method that spiral is approached and is surveyed It visits, specifically, being calculated using equation below：
    Mi=S (Mi,Fj),
    S(Mi,Fj)=Di·ebt·cos(2πt)+Fflameno,
    Di=| Fj-Mi|,
    Wherein, MiRepresent the position of i-th moth, FflamenoRepresent the flameno flame number randomly selected, S is spiral letter Number；DiRepresent the distance between i-th moth and j-th flame, b is constant, a random number between t ∈ [- 1,1].
  8. 8. the method according to claim 7 that Function Extreme value is calculated based on a flying moth darts into the fire algorithm, which is characterized in that described In a flying moth darts into the fire algorithm, adaptive weighting is introduced, then according to following adaptive weighting mathematical model：
    <mrow> <mi>w</mi> <mo>=</mo> <mfrac> <mrow> <msub> <mi>w</mi> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> </mrow> </msub> <mo>-</mo> <mrow> <mo>(</mo> <msub> <mi>w</mi> <mrow> <mi>m</mi> <mi>a</mi> <mi>x</mi> </mrow> </msub> <mo>-</mo> <msub> <mi>w</mi> <mi>min</mi> </msub> <mo>)</mo> </mrow> </mrow> <mrow> <mi>I</mi> <mi>t</mi> <mi>e</mi> <mi>r</mi> </mrow> </mfrac> <mo>,</mo> </mrow>
    Obtain following improved Spirallike Functions：
    S(Mi,Fj)=wDi·ebt·cos(2πt)+Fflameno,
    Wherein, wmaxFor constant, 0.9, w is traditionally arranged to bemin0.4, Iter is arranged to as current iteration number, w is adaptively to weigh Value.
CN201711480246.3A 2017-12-29 2017-12-29 A kind of method that Function Extreme value is calculated based on a flying moth darts into the fire algorithm Pending CN108062585A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN201711480246.3A CN108062585A (en) 2017-12-29 2017-12-29 A kind of method that Function Extreme value is calculated based on a flying moth darts into the fire algorithm

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN201711480246.3A CN108062585A (en) 2017-12-29 2017-12-29 A kind of method that Function Extreme value is calculated based on a flying moth darts into the fire algorithm

Publications (1)

Publication Number Publication Date
CN108062585A true CN108062585A (en) 2018-05-22

Family

ID=62140870

Family Applications (1)

Application Number Title Priority Date Filing Date
CN201711480246.3A Pending CN108062585A (en) 2017-12-29 2017-12-29 A kind of method that Function Extreme value is calculated based on a flying moth darts into the fire algorithm

Country Status (1)

Country Link
CN (1) CN108062585A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110109350A (en) * 2019-03-29 2019-08-09 广东工业大学 A kind of power capture optimization method of wave-power device that catching flame algorithm based on chaos moth
CN111125885A (en) * 2019-12-03 2020-05-08 杭州电子科技大学 ASF correction table construction method based on improved kriging interpolation algorithm
CN111880402A (en) * 2020-07-30 2020-11-03 广州大学 Method and device for controlling product parameters of fluorescent powder layer and storage medium
WO2023134403A1 (en) * 2022-01-11 2023-07-20 中国科学院深圳先进技术研究院 Internet of things resource allocation method and system, terminal and storage medium
CN117253555A (en) * 2023-11-17 2023-12-19 山东阜丰发酵有限公司 Method for improving xanthan gum fermentation process and operation system thereof

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN110109350A (en) * 2019-03-29 2019-08-09 广东工业大学 A kind of power capture optimization method of wave-power device that catching flame algorithm based on chaos moth
CN111125885A (en) * 2019-12-03 2020-05-08 杭州电子科技大学 ASF correction table construction method based on improved kriging interpolation algorithm
CN111880402A (en) * 2020-07-30 2020-11-03 广州大学 Method and device for controlling product parameters of fluorescent powder layer and storage medium
WO2023134403A1 (en) * 2022-01-11 2023-07-20 中国科学院深圳先进技术研究院 Internet of things resource allocation method and system, terminal and storage medium
CN117253555A (en) * 2023-11-17 2023-12-19 山东阜丰发酵有限公司 Method for improving xanthan gum fermentation process and operation system thereof

Similar Documents

Publication Publication Date Title
CN108062585A (en) A kind of method that Function Extreme value is calculated based on a flying moth darts into the fire algorithm
CN108133258B (en) Hybrid global optimization method
CN103971160B (en) particle swarm optimization method based on complex network
Panda et al. Evolutionary techniques for model order reduction of large scale linear systems
Wang et al. An improved hybrid Aquila Optimizer and Harris Hawks Optimization for global optimization
CN111553469A (en) Wireless sensor network data fusion method, device and storage medium
Meng et al. A novel hybrid bat algorithm with differential evolution strategy for constrained optimization
CN111105045A (en) Method for constructing prediction model based on improved locust optimization algorithm
CN106991442A (en) The self-adaptive kernel k means method and systems of shuffled frog leaping algorithm
Özsoy et al. Estimating the parameters of nonlinear regression models through particle swarm optimization
Gorjestani et al. A hybrid COA-DEA method for solving multi-objective problems
CN108229536A (en) Optimization method, device and the terminal device of classification prediction model
CN114758180B (en) Knowledge distillation-based lightweight flower identification method
Zhou et al. CCEO: cultural cognitive evolution optimization algorithm
Liu et al. Financial data forecasting using optimized echo state network
CN103593504B (en) A kind of based on the netting Reliability of Microprocessor emulation mode improving quality amplifying technique
Bekker Applying the cross-entropy method in multi-objective optimisation of dynamic stochastic systems
CN107590538A (en) A kind of dangerous source discrimination based on online Sequence Learning machine
Zhang et al. Self-adaptive hybrid mutation slime mould algorithm: case studies on uav path planning, engineering problems, photovoltaic models and infinite impulse response
Wei et al. A modified fireworks algorithm with dynamic search interval based on closed-loop control
CN108803330A (en) A kind of learning aid optimization sliding-mode control based on personality coefficient adjustment
CN109447231B (en) Method for solving multi-attribute bilateral matching problem under shared economic background by ant colony algorithm
Zhao et al. Application of LDHA-BP in prediction of atmospheric pm2. 5 concentration
Cui et al. Application of RBF neural network improved by PSO algorithm in fault diagnosis
Liu Shortest path selection algorithm for cold chain logistics transportation based on improved artificial bee colony

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication

Application publication date: 20180522

RJ01 Rejection of invention patent application after publication