CN113609761B - Calculation method, device, equipment and storage medium of model parameters - Google Patents
Calculation method, device, equipment and storage medium of model parameters Download PDFInfo
- Publication number
- CN113609761B CN113609761B CN202110822225.5A CN202110822225A CN113609761B CN 113609761 B CN113609761 B CN 113609761B CN 202110822225 A CN202110822225 A CN 202110822225A CN 113609761 B CN113609761 B CN 113609761B
- Authority
- CN
- China
- Prior art keywords
- particle
- model
- strategy
- population
- iteration number
- 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.)
- Active
Links
- 238000004364 calculation method Methods 0.000 title claims abstract description 51
- 238000000034 method Methods 0.000 claims abstract description 32
- 238000011161 development Methods 0.000 claims abstract description 31
- 238000013461 design Methods 0.000 claims abstract description 27
- 230000000739 chaotic effect Effects 0.000 claims abstract description 26
- 238000013507 mapping Methods 0.000 claims abstract description 22
- 238000005457 optimization Methods 0.000 claims abstract description 15
- 230000002028 premature Effects 0.000 claims abstract description 14
- 239000002245 particle Substances 0.000 claims description 177
- 230000006870 function Effects 0.000 claims description 45
- 238000004590 computer program Methods 0.000 claims description 13
- 238000005452 bending Methods 0.000 claims description 7
- 230000006978 adaptation Effects 0.000 claims description 6
- 230000003247 decreasing effect Effects 0.000 claims 2
- 238000004422 calculation algorithm Methods 0.000 abstract description 14
- 230000000694 effects Effects 0.000 abstract description 4
- 238000010586 diagram Methods 0.000 description 7
- 230000007423 decrease Effects 0.000 description 5
- 230000008569 process Effects 0.000 description 4
- 230000004044 response Effects 0.000 description 4
- 238000005094 computer simulation Methods 0.000 description 3
- 241000272478 Aquila Species 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000001514 detection method Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000013528 artificial neural network Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000006872 improvement Effects 0.000 description 1
- 239000007788 liquid Substances 0.000 description 1
- 230000035800 maturation Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000012549 training Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
-
- 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]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/10—Numerical modelling
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
- Y02T10/00—Road transport of goods or passengers
- Y02T10/10—Internal combustion engine [ICE] based vehicles
- Y02T10/40—Engine management systems
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Evolutionary Computation (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computer Hardware Design (AREA)
- Databases & Information Systems (AREA)
- Algebra (AREA)
- Computational Mathematics (AREA)
- Medical Informatics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Pure & Applied Mathematics (AREA)
- Geometry (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The embodiment of the invention provides a method, a device, equipment and a storage medium for calculating model parameters, and relates to the technical field of model parameter solving. Wherein, this calculation method at least comprises steps S1 and S2. S1, acquiring a calculation model of a target object to be solved. S2, solving a calculation model through an improved Alquacrak model to obtain design parameters of the target object. The improved Alquacrak optimization model calculates an initial population through logistic chaotic mapping, and adopts elite hunting strategy to carry out local development in the development stage so as to avoid premature convergence. By introducing Logistic map chaotic mapping in a population initialization stage, the diversity of the initialized population distribution is improved, elite hunting strategy is introduced in a local development stage, a population library in a small range is established, elite individuals and common individuals are used as guidance for local development, algorithm premature convergence is avoided, and a better calculation effect can be obtained.
Description
Technical Field
The invention relates to the technical field of model parameter solving, in particular to a method, a device, equipment and a storage medium for calculating model parameters.
Background
The nyquagmill model is designed with an exploration stage and a development stage by simulating the hunting mode of the nyquagmill. Aloft flight search and vertical bend-over attacks of the nyquagmus and fly around the prey and short-range glide attacks are simulated during the exploration phase. The low-altitude flight of the Alquacrak is simulated in the development stage, and the descent attack and the ground short-distance attack are simulated.
The nyquagmus model can be used to solve solutions of various computational models. However, it tends to fall into a local optimum, resulting in a decrease in the resolution accuracy. When solving the computational model, an optimal solution may not be obtained.
The pressure vessel is a closed device for containing gas or liquid and bearing a certain pressure. In the development of various industries, various devices must be relied on, wherein pressure vessels are very widely used and very important devices in many areas of our country.
The design of pressure vessels is a typical nonlinear mixed integer programming problem. The design of the pressure vessel has 4 parameters to be optimized, namely the length, thickness, inner radius and hemispherical thickness of the middle part of the vessel. In the prior art, the calculation method of the parameters of the pressure vessel generally has a large number of local optimal points, and is easy to sink into the local optimal points, so that the problems of low convergence accuracy, low convergence speed and the like are caused. In view of this, the applicant has studied the prior art and has made the present application.
Disclosure of Invention
The invention provides a calculation method, a device, equipment and a storage medium of model parameters, which are used for solving the problems of low convergence accuracy and low convergence speed of a design method in the related technology.
A first aspect,
The embodiment of the invention provides a calculation method of model parameters, which at least comprises steps S1 and S2.
S1, acquiring a calculation model of a target object to be solved.
And S2, solving the calculation model through an improved Alquacrak model to obtain the design parameters of the target object. The improved Alquail optimization model calculates an initial population through logistic chaotic mapping, and adopts elite hunting strategy to carry out local development in the development stage so as to avoid premature convergence.
In an alternative embodiment, step S2 includes at least steps S21 to S26.
S21, randomly generating an initial particle population according to the calculation model, and initializing the iteration times and the maximum iteration times.
S22, calculating to obtain a diversified particle population through a logistic chaotic mapping model according to the initial particle population.
S23, calculating the fitness of each particle in the diversified particle population, and judging whether the iteration number is not more than a preset iteration number.
And S24, when the iteration number is not greater than the preset iteration number, updating the position of the particle population by adopting a large-range search attack strategy, adding one to the iteration number, and judging whether the maximum iteration number is reached.
And S25, when the iteration times are greater than the preset iteration times, updating the positions of the particle population by adopting a local attack strategy, calculating to obtain elite positions by adopting an elite hunting strategy, obtaining the optimal positions by adopting a greedy strategy, adding one to the iteration times, and judging whether the maximum iteration times are reached.
And S26, when judging that the maximum iteration number is not reached, calculating the fitness of each particle in the updated particle population, and judging whether the iteration number is not more than the preset iteration number. Otherwise, outputting the optimal position and the corresponding fitness of the optimal position when the maximum iteration number is output so as to obtain the design parameters.
In an alternative embodiment, step S21 includes at least step S211 and step S212.
S211, acquiring the upper bound and the lower bound of the search space according to the calculation model.
S212, generating the initial particle population in the search space according to the random function. Specifically, the initial population generation model is:
X=rand×(UB-LB)+LB
where X is the particle position, rand is a random number uniformly distributed between 0 and 1, UB is the upper boundary of the search space, and LB is the lower boundary of the search space.
In an alternative embodiment, the logistc chaotic map model is:
X(t+1)=a×X(t)×(1-X(t))
where X (t+1) is the updated particle position, a=4, and X (t) is the current particle position.
In an alternative embodiment, the fitness calculation model of the fitness of the particles is:
wherein F (x) is a penalty function value, F (x) is an original function value, M is a penalty factor, g i (x) As a constraint, n is the number of particles.
In an alternative embodiment, step S24 includes at least steps S241 to S243.
S241, acquiring a first random number, and judging whether the random number is smaller than a first preset value.
And S242, when the first random number is judged to be smaller than a first preset value, the position of the particle population is updated by adopting a strategy of high-altitude flight searching and vertical bending attack. Specifically, the location update model for high altitude flight search and vertical bend-over attack is:
wherein X is 1 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, X M (T) is the average position of all the current particles, T is the current iteration number, T is the maximum iteration number, N is the population number, r 1 A random number between 0 and 1, X i (t) is the current ith particle position.
And S243, when the first random number is judged not to be smaller than a first preset value, updating the position of the particle population by adopting a strategy of flying around the prey and attacking in a short-distance gliding mode. Specifically, the location update model for flying around a prey and short-range glide attack is:
X 2 (t+1)=X best (t)×LF(D)+X R (t)+(y-x)×r 2
Wherein X is 2 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, LF (D) represents the Lewy flight function, D represents the population dimension, X R (t) represents the position of any particle, r 2 For random numbers uniformly distributed between 0 and 1, s=0.01, u and v are random numbers between 0 and 1, and β=1.5.
In an alternative embodiment, step S25 includes at least steps S251 to S253.
S251, a second random number is acquired, and whether the random number is smaller than a second preset value is judged.
And S252, when the second random number is judged to be smaller than the second preset value, the particle position is updated by adopting a strategy of low-altitude flight and slow-descent attack. Specifically, the location update model of the low-altitude flight and slow-descent attack is as follows:
X 3 (t+1)=(X best (t)-X M (t))×α-r 4 +((UB-LB)×r 5 +LB)×δ
wherein X is 3 (t+1) represents the next updated position of the particle, X best (t) Represents the best position, X, which is currently obtained M (t) is the average position of all particles present, alpha and delta are adaptation parameters, their values are 0.1, UB is the upper boundary, LB is the lower boundary, r 4 And r 5 A random number between 0 and 1. X is X i (t) the number of N particles for the current ith particle position.
And S253, when the second random number is judged to be not smaller than the second preset value, updating the particle position by adopting a ground short-distance attack hunting strategy. Specifically, the location update model of the ground proximity attack prey is:
Wherein X is 4 (t+1) denotes the next updated position of the particle, QF (t) is a mass function for balancing its search strategy, X best (t) represents the best position currently obtained, G 1 The movement parameter of the Alquacral tracking hunting object is [ -1,1]Random number in between, X (t) is the current particle position, r 6 、r 7 And r 8 A random number between 0 and 1, G 2 The flight slope of the chase hunting in the air decreases linearly from 2 to 0, LF (D) represents the Lewy flight function, T is the current iteration number, and T is the maximum iteration number.
In an alternative embodiment, the elite position calculation model for elite position calculated by elite hunting strategy is:
X E (t+1)=X E-best +cd×(X(t)-X E-normal (t))+σ×(X rand1 -X rand2 )
cd=z 0 +0.1×tan(π×(r-0.5))
wherein X is E (t+1) is a new location, X, generated based on elite hunting strategy E-best Cd is the coincidence of individuals randomly selected from 5 elite individualsThe random number of the Coxil distribution, X (t) is the current particle position, X E-normal (t) 15 positions selected randomly from the warehouse, X rand1 And X is rand2 For random selection from the remaining 15 individuals, exponent=2, σ initial =1σ final =0,z 0 =2, T is the current iteration number, T is the maximum iteration number, r is the distribution in [0,1]Random numbers in between.
In an alternative embodiment, the preset number of iterations is two-thirds of the maximum number of iterations.
In an alternative embodiment, the target object is a pressure vessel.
A second aspect,
The embodiment of the invention provides a calculation device of model parameters, which comprises the following components:
and the model acquisition module is used for acquiring a calculation model of the target object to be solved.
And the model solving module is used for solving the calculation model through an improved Alquacrak model so as to obtain the design parameters of the target object. The improved Alquail optimization model calculates an initial population through logistic chaotic mapping, and adopts elite hunting strategy to carry out local development in the development stage so as to avoid premature convergence.
In an alternative embodiment, the model solving module specifically includes:
and the initialization unit is used for randomly generating an initial particle population according to the calculation model and initializing the iteration times and the maximum iteration times.
And the chaotic mapping unit is used for calculating and obtaining the diversified particle population through a logistic chaotic mapping model according to the initial particle population.
And the fitness unit is used for calculating the fitness of each particle in the diversified particle population and judging whether the iteration number is not more than the preset iteration number.
And the large-range unit is used for updating the position of the particle population by adopting a large-range search attack strategy when the iteration number is not more than the preset iteration number, adding one to the iteration number, and judging whether the maximum iteration number is reached.
And the local unit is used for updating the position of the particle population by adopting a local attack strategy when the iteration number is greater than the preset iteration number, calculating to obtain the elite position by adopting an elite hunting strategy, acquiring the optimal position by adopting a greedy strategy, adding one to the iteration number, and judging whether the maximum iteration number is reached.
And the output unit is used for calculating the fitness of each particle in the updated particle population when the maximum iteration number is not reached, and judging whether the iteration number is not more than the preset iteration number. Otherwise, outputting the optimal position and the corresponding fitness of the optimal position when the maximum iteration number is output so as to obtain the design parameters.
A third aspect,
An embodiment of the invention provides a computing device for model parameters, which comprises a processor, a memory and a computer program stored in the memory. The computer program is executable by the processor to implement the method of calculating model parameters as described in the first aspect.
A fourth aspect,
An embodiment of the present invention provides a computer readable storage medium, where the computer readable storage medium includes a stored computer program, where the computer program when run controls a device in which the computer readable storage medium is located to execute a method for calculating model parameters according to the first aspect.
By adopting the technical scheme, the invention can obtain the following technical effects:
according to the embodiment of the invention, the Logistic map chaotic map is introduced in the population initialization stage, so that the diversity of the initialized population distribution is improved, the elite hunting strategy is introduced in the local development stage, a small-range population library is established, the elite individuals and the common individuals are used as the guide for local development, the premature convergence of the algorithm is avoided, and a better calculation effect can be obtained.
In order to make the above objects, features and advantages of the present invention more comprehensible, preferred embodiments accompanied with figures are described in detail below.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings that are needed in the embodiments will be briefly described below, it being understood that the following drawings only illustrate some embodiments of the present invention and therefore should not be considered as limiting the scope, and other related drawings may be obtained according to these drawings without inventive effort for a person skilled in the art.
Fig. 1 is a flowchart of a method for calculating model parameters according to a first embodiment of the present invention.
Fig. 2 is a logical block diagram of a solution to a computational model by a modified nyquagmus model.
FIG. 3 is a plot of a set of points of Logistic chaos.
Fig. 4 is a pressure vessel structural model.
Fig. 5 is a schematic structural diagram of a device for calculating model parameters according to a second embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
For a better understanding of the technical solution of the present invention, the following detailed description of the embodiments of the present invention refers to the accompanying drawings.
It should be understood that the described embodiments are merely some, but not all, embodiments of the invention. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
The terminology used in the embodiments of the invention is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in this application and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
It should be understood that the term "and/or" as used herein is merely one relationship describing the association of the associated objects, meaning that there may be three relationships, e.g., a and/or B, may represent: a exists alone, A and B exist together, and B exists alone. In addition, the character "/" herein generally indicates that the front and rear associated objects are an "or" relationship.
Depending on the context, the word "if" as used herein may be interpreted as "at … …" or "at … …" or "in response to a determination" or "in response to detection". Similarly, the phrase "if determined" or "if detected (stated condition or event)" may be interpreted as "when determined" or "in response to determination" or "when detected (stated condition or event)" or "in response to detection (stated condition or event), depending on the context.
References to "first\second" in the embodiments are merely to distinguish similar objects and do not represent a particular ordering for the objects, it being understood that "first\second" may interchange a particular order or precedence where allowed. It is to be understood that the "first\second" distinguishing aspects may be interchanged where appropriate, such that the embodiments described herein may be implemented in sequences other than those illustrated or described herein.
The invention is described in further detail below with reference to the attached drawings and detailed description:
embodiment one:
referring to fig. 1 to 4, a method for calculating model parameters according to a first embodiment of the present invention may be performed by a computing device for model parameters. In particular, performed by one or more processors in the computing device to implement at least steps S1 and S2.
S1, acquiring a calculation model of a target object to be solved. In this embodiment, the target object is a pressure vessel. It should be noted that the design of the pressure vessel is a typical nonlinear mixed integer programming problem. There are a number of local optimum points for pressure vessel design issues. Therefore, at the time of design, an optimal solution is often not obtained.
As shown in fig. 4, the pressure vessel design problem has 4 parameters to be optimized, namely the length L, thickness Ts, inner radius R and hemispherical thickness Th of the vessel middle portion. Specifically, the calculation model of the pressure vessel is:
wherein Ts is x 1 Th is x 2 R is x 3 L is x 4 。
It will be appreciated that the computing device may be a local computer, a laptop portable computer, a local server, or a cloud server, as the invention is not limited in this regard. In other embodiments, the target object may be another research object, which is not limited in the present invention.
S2, solving a calculation model through an improved Alquacrak model to obtain design parameters of the target object. The improved Alquacrak optimization model calculates an initial population through logistic chaotic mapping, and adopts elite hunting strategy to carry out local development in the development stage so as to avoid premature convergence.
The Aquila Optimizer (AO) has paid a lot of attention to the algorithm novelty and high optimization performance, and has been used for classification problems, neural network training, feature selection and other problems at present, and good effects are obtained. Although AO has certain advantage in solving certain problems, the algorithm is only guided according to the current optimal position in the later optimizing process, so that the algorithm is easy to sink into local optimal, and the problems of low convergence accuracy, low convergence speed and the like are caused.
In order to solve the problems, the embodiment provides an improved nyquagmill optimization model (Improved Aquila Optimizer, IAO), which introduces Logistic map chaotic mapping in a population initialization stage, improves diversity of initialized population distribution, introduces elite hunting strategy in a local development stage, establishes a small-range population library, and utilizes elite individuals and common individuals as guidance to carry out local development so as to avoid premature convergence of algorithms. The improved Alquacrak optimizing model is obviously better than the original Alquacrak optimizing model, and particularly, the improved Alquacrak optimizing model achieves a better level on the design problem of the pressure vessel.
Specifically, as shown in fig. 2, in an alternative embodiment of the present invention, step S2 includes at least steps S21 to S26.
S21, randomly generating an initial particle population according to the calculation model, and initializing the iteration times and the maximum iteration times. Specifically, step S21 includes at least step S211 and step S212.
S211, acquiring the upper bound and the lower bound of the search space according to the calculation model.
S212, generating an initial particle population in the search space according to the random function. Specifically, the initial population generation model is:
X=rand×(UB-LB)+LB
Where X is the particle position, rand is a random number uniformly distributed between 0 and 1, UB is the upper boundary of the search space, and LB is the lower boundary of the search space.
S22, calculating to obtain a diversified particle population through a logistic chaotic mapping model according to the initial particle population. Specifically, the logistic chaotic mapping model is:
x (t+1) =a×x (t) × (1-X (t)) … … … … … … (formula 1)
Where X (t+1) is the updated particle position, a=4, and X (t) is the current particle position.
It should be noted that, as shown in fig. 3, a feature of the Logistic chaotic nonlinear system is mathematically defined as a random number generated by a simple deterministic function. In the embodiment, the chaos idea is integrated into an improved Alquacrak optimization model, and the core of the chaos idea is to replace random variables obeying standard probability distribution by utilizing randomness, ergodic property and irregularity of chaos variables to perform optimization search, so that the whole solution space is searched, individuals with even distribution are generated, and the diversity of the initialized population distribution and the initial population quality are improved.
S23, calculating the fitness of each particle in the diversified particle population, and judging whether the iteration number is not more than the preset iteration number. In this embodiment, the preset number of iterations is two-thirds of the maximum number of iterations. In other embodiments, the preset number of iterations may be other ratios, or other predetermined numbers, which are not particularly limited by the present invention. Specifically, the fitness calculation model of the fitness of the particles is:
Wherein F (x) is a penalty function value, F (x) is an original function value, M is a penalty factor, gi (x) is a constraint condition, and n is the number of particles.
The pressure vessel design problem is a constraint optimization problem, which is converted into an unconstrained optimization problem by a punishment function method. For the individual x in the population, which does not violate the constraint condition, since all the constraint conditions are satisfied at this time, that is, g1 is less than or equal to 0, g2 is less than or equal to 0, g3 is less than or equal to 0, and g4 is less than or equal to 0, the maximum of 0 and g1 (x) is 0, so that max (0, g1 (x))=0 can be obtained, wherein max is a maximum function. Similarly, max (0, g2 (x))=0, …, max (0, gn (x))=0, thereby satisfying the fitness calculation model.
For the individual x without violating constraint condition, adding punishment function value F (x) after punishment and original function value F #x) are equal, penalty termDoes not play any role; for individual x that violates a constraint, its penalty term is due to the violation of the constraint>0, would result in the penalty function F (x) for that individual being greater than the original function F (x), such that individual is phased out during evolution due to the worse penalty function.
The penalty function method adds a penalty term in the unqualified solutions, so that some unqualified solutions which only violate the constraint in a certain dimension cannot be abandoned and iteration can be continued, the problem that fewer and fewer qualified solutions in the population are avoided, meanwhile, solutions with poor quality are obtained, solutions which do not violate the constraint are made into good solutions, and the solutions with poor quality are eliminated gradually.
And S24, when the iteration number is not greater than the preset iteration number, updating the position of the particle population by adopting a large-range search attack strategy, adding one to the iteration number, and judging whether the maximum iteration number is reached. It should be noted that the nyquagmill model has an exploration phase and a development phase, and is in the exploration phase when the iteration number is not greater than the preset iteration number, and is in the development phase when the iteration number is greater than the preset number.
The search stage, namely the nyquagmill, realizes the wide-range search of hunting objects and attack in a search space in a random range through two different hunting modes, and the two hunting modes are introduced as follows:
1. high-altitude flight search and vertical bend-over attack:
the alopecie flies at high altitude, a prey flying below the alopecie is observed through eyes, and after the range of the prey is determined, the vertical bending attack is carried out, and the calculation formula is as follows:
wherein X is 1 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, X M (T) is the average position of all the current particles, T is the current iteration number, T is the maximum iteration number, N is the population number, r 1 A random number between 0 and 1, X i (t) is the current ith particle position.
2. Flying around the prey and short-range glide attack:
the nyquagmill typically employs this hunting method, gradually lowering the altitude and encircling the prey, and employing short range glide to attack the prey, the location update formula is as follows:
X 2 (t+1)=X best (t)×LF(D)+X R (t)+(y-x)×r 2
wherein X is 2 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, LF (D) represents the Lewy flight function, D represents the population dimension, X R (t) represents the position of any particle, r 2 For random numbers uniformly distributed between 0 and 1, s=0.01, u and v are random numbers between 0 and 1, and β=1.5.
Specifically, the "update the location of the particle population with the broad search attack strategy" in step S24 includes at least steps S241 to S243.
S241, acquiring a first random number, and judging whether the random number is smaller than a first preset value.
And S242, when the first random number is judged to be smaller than a first preset value, the position of the particle population is updated by adopting a strategy of high-altitude flight searching and vertical bending attack. Specifically, the location update model for high altitude flight search and vertical bend-over attack is:
wherein X is 1 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, X M (T) is the average position of all the current particles, T is the current iteration number, T is the maximum iteration number, N is the population number, r 1 A random number between 0 and 1, X i (t) is the current ith particle position.
And S243, when the first random number is judged not to be smaller than a first preset value, the position of the particle population is updated by adopting a strategy of flying around the prey and attacking in a short-distance gliding mode. Specifically, the location update model for flying around a prey and short-range glide attack is:
X 2 (t+1)=X best (t)×LF(D)+X R (t)+(y-x)×r 2 … … … … … (4)
Wherein X is 2 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, LF (D) represents the Lewy flight function, D represents the population dimension, X R (t) represents the position of any particle, r 2 For random numbers uniformly distributed between 0 and 1, s=0.01, u and v are random numbers between 0 and 1, and β=1.5.
And S25, when the iteration times are greater than the preset iteration times, updating the positions of the particle population by adopting a local attack strategy, calculating to obtain elite positions by adopting an elite hunting strategy, obtaining the optimal positions by adopting a greedy strategy, adding one to the iteration times, and judging whether the maximum iteration times are reached. It should be noted that, when the iteration number is greater than the preset iteration number, the nyquist hawk model enters the development stage, and the nyquist hawk executes the local attack strategy in the development stage. Local search for hunting and attack in the search space is achieved by two different hunting methods, which are described as follows:
3. Low-altitude flight and slow-descent attack
The position of the hunting object is determined by the nyquagmill in the exploring stage, so the nyquagmill adopts low-altitude flight and slowly descends to attack the hunting object, and the hunting object with lower escape capability is predated by the method, and the position updating formula is as follows:
X 3 (t+1)=(X best (t)-X M (t))×α-r 4 +((UB-LB)×r 5 +LB)×δ
wherein X is 3 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, X M (t) is the average position of all particles present, alpha and delta are adaptation parameters, their values are 0.1, UB is the upper boundary, LB is the lower boundary, r 4 And r 5 A random number between 0 and 1. X is X i (t) the number of N particles for the current ith particle position.
4. Ground short-distance attack hunting object
The alopecie chases the hunting object according to the hunting track at low altitude and launches attack to the hunting object on the ground, and the position updating formula is as follows:
wherein X is 4 (t+1) denotes the next updated position of the particle, QF (t) is a mass function for balancing its search strategy, X best (t) represents the best position currently obtained, G 1 The movement parameter of the Alquacral tracking hunting object is [ -1,1]Random number in between, X (t) is the current particle position, r 6 、r 7 And r 8 A random number between 0 and 1, G 2 The flight slope of the chase hunting in the air decreases linearly from 2 to 0, LF (D) represents the Lewy flight function, T is the current iteration number, and T is the maximum iteration number.
Specifically, the "update the position of the particle population with the local attack strategy" in step S25 includes at least steps S251 to S253.
S251, a second random number is acquired, and whether the random number is smaller than a second preset value is judged.
And S252, when the second random number is judged to be smaller than a second preset value, the particle position is updated by adopting a strategy of low-altitude flight and slow-descent attack. Specifically, the location update model of the low-altitude flight and slow-descent attack is as follows:
X 3 (t+1)=(X best (t)-X M (t))×α-r 4 +((UB-LB)×r 5 +LB). Times.delta. … … … (5)
Wherein X is 3 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, X M (t) is the average position of all particles present, alpha and delta are adaptation parameters, their values are 0.1, UB is the upper boundary, LB is the lower boundary, r 4 And r 5 A random number between 0 and 1. X is X i (t) the number of N particles for the current ith particle position.
And S253, when the second random number is judged to be not smaller than a second preset value, the particle position is updated by adopting a ground short-distance attack prey strategy. Specifically, the location update model of the ground proximity attack prey is:
wherein X is 4 (t+1) denotes the next updated position of the particle, QF (t) is a mass function for balancing its search strategy, X best (t) represents the best position currently obtained, G 1 The movement parameter of the Alquacral tracking hunting object is [ -1,1]Random number in between, X (t) is the current particle position, r 6 、r 7 And r 8 A random number between 0 and 1, G 2 The flight slope of the chase hunting in the air decreases linearly from 2 to 0, LF (D) represents the Lewy flight function, T is the current iteration number, and T is the maximum iteration number.
To prevent the improved nyquagmill model, a locally optimal solution is trapped during the development phase, resulting in premature convergence. In the embodiment, through elite hunting strategy, elite individuals and common individuals are utilized for local development, and early maturation convergence of the Alquacrak model is avoided. Specifically, an elite position calculation model for calculating an elite position by an elite hunting strategy is:
X E (t+1)=X E-best +cd×(X(t)-X E-normal (t))+σ×(X rand1 -X rand2 )
/>
cd=z 0 +0.1×tan(π×(r-0.5))
wherein X is E (t+1) is a new location, X, generated based on elite hunting strategy E-best For randomly selected individuals from the 5 elite individuals, cd is a random number conforming to the cauchy distributionX (t) is the current particle position, X E-normal (t) 15 positions selected randomly from the warehouse, X rand1 And X is rand2 For random selection from the remaining 15 individuals, exponent=2, σ initial =1σ final =0,z 0 =2, T is the current iteration number, T is the maximum iteration number, r is the distribution in [0,1 ]Random numbers in between.
It should be noted that, the elite hunting strategy is to build a population library, dispatch individual elite to go out and explore, and make the individuals fully interact information, avoid the algorithm to be in local optimum, and make the algorithm not easy to converge in premature.
Elite hunting strategy principle: firstly, a warehouse is established, 15 populations are taken out and put into the warehouse in each iteration, elite individuals (5) and the rest are ordinary individuals (10) are determined according to the fitness value sequence, and at the moment, if the total population number is 30, the total number of individuals outside the warehouse is 15. And calculating according to the elite position calculation model.
Specifically, the obtaining the optimal position through the greedy strategy is a conventional technical means for those skilled in the art, and the present invention is not described herein.
And S26, when judging that the maximum iteration number is not reached, calculating the fitness of each particle in the updated particle population, and judging whether the iteration number is not more than the preset iteration number. Otherwise, outputting the optimal position and the corresponding adaptability of the optimal position when the maximum iteration number is output so as to obtain the design parameters.
According to the calculation method of the model parameters, an improved Alquacrak model is adopted, and the population distribution diversity is improved by introducing Logistic chaotic mapping in the population initialization stage. And introducing elite hunting strategies in a local development stage, establishing a population library to fully exert potential capability of elite individuals and common individuals, and selecting the optimal position to enter the next iteration through a greedy algorithm to avoid easy premature convergence of the algorithm. Finally, the constraint design problem is converted into an unconstrained optimization problem through a punishment function, so that an IAO algorithm can be used for obtaining an optimal solution. More accurate design parameters of the model parameters can be obtained.
A second aspect,
The embodiment of the invention provides a calculation device of model parameters, which comprises the following components:
the model acquisition module 1 is used for acquiring a calculation model of the target object to be solved. Preferably, the target object is a pressure vessel.
And the model solving module 2 is used for solving the calculation model through the improved Alquacrak model so as to obtain the design parameters of the target object. The improved Alquacrak optimization model calculates an initial population through logistic chaotic mapping, and adopts elite hunting strategy to carry out local development in the development stage so as to avoid premature convergence.
The calculation device of the model parameters provided by the embodiment adopts an improved Alquacrak model, and improves population distribution diversity by introducing Logistic chaotic mapping in a population initialization stage. And introducing elite hunting strategies in a local development stage, establishing a population library to fully exert potential capability of elite individuals and common individuals, and selecting the optimal position to enter the next iteration through a greedy algorithm to avoid easy premature convergence of the algorithm. Finally, the constraint design problem is converted into an unconstrained optimization problem through a punishment function, so that an IAO algorithm can be used for obtaining an optimal solution. More accurate design parameters of the model parameters can be obtained.
In an alternative embodiment, the model solving module 2 specifically includes:
and the initialization unit is used for randomly generating an initial particle population according to the calculation model, and initializing the iteration times and the maximum iteration times.
And the chaotic mapping unit is used for calculating and obtaining a diversified particle population through a logistic chaotic mapping model according to the initial particle population.
And the fitness unit is used for calculating the fitness of each particle in the diversified particle population and judging whether the iteration times are not more than the preset iteration times. Preferably, the preset number of iterations is two-thirds of the maximum number of iterations.
And the large-range unit is used for updating the position of the particle population by adopting a large-range search attack strategy when the iteration number is not more than the preset iteration number, adding one to the iteration number, and judging whether the maximum iteration number is reached.
And the local unit is used for updating the position of the particle population by adopting a local attack strategy when the iteration number is greater than the preset iteration number, calculating to obtain the elite position by adopting an elite hunting strategy, acquiring the optimal position by adopting a greedy strategy, adding one to the iteration number, and judging whether the maximum iteration number is reached.
And the output unit is used for calculating the fitness of each particle in the updated particle population when the maximum iteration number is not reached, and judging whether the iteration number is not more than the preset iteration number. Otherwise, outputting the optimal position and the corresponding adaptability of the optimal position when the maximum iteration number is output so as to obtain the design parameters.
In an alternative embodiment, the initialization unit is specifically configured to:
and acquiring the upper bound and the lower bound of the search space according to the calculation model.
The initial population of particles is generated within a search space according to a random function. Specifically, the initial population generation model is x=rand× (UB-LB) +lb, where X is the particle position, rand is a random number uniformly distributed between 0 and 1, UB is the upper boundary of the search space, and LB is the lower boundary of the search space.
In an alternative embodiment, the logistic chaotic map model is:
X(t+1)=a×X(t)×(1-X(t))
where X (t+1) is the updated particle position, a=4, and X (t) is the current particle position.
In an alternative embodiment, the fitness calculation model of the fitness of the particles is:
wherein F (x) is a penalty function value, F (x) is an original function value, M is a penalty factor, g i (x) As a constraint, n is the number of particles.
In an alternative embodiment, when updating the location of a particle population using a broad search attack strategy, a broad range unit is specifically used to:
and acquiring a first random number, and judging whether the random number is smaller than a first preset value.
And when the first random number is judged to be smaller than a first preset value, updating the position of the particle population by adopting a strategy of high-altitude flight searching and vertical bending attack. Specifically, the location update model for high altitude flight search and vertical bend-over attack is:
wherein X is 1 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, X M (T) is the average position of all the current particles, T is the current iteration number, T is the maximum iteration number, N is the population number, r 1 A random number between 0 and 1, X i (t) is the current ith particle position.
And when the first random number is not smaller than a first preset value, updating the position of the particle population by adopting a strategy of flying around the prey and attacking in a short-distance gliding mode. Specifically, the location update model for flying around a prey and short-range glide attack is:
X 2 (t+1)=X best (t)×LF(D)+X R (t)+(y-x)×r 2
wherein X is 2 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, LF (D) represents the Lewy flight function, D represents the population dimension, X R (t) represents the position of any particle, r 2 For random numbers uniformly distributed between 0 and 1, s=0.01, u and v are random numbers between 0 and 1, and β=1.5.
In an alternative embodiment, the local unit is specifically configured to, when updating the location of the particle population using a local attack strategy:
and acquiring a second random number, and judging whether the random number is smaller than a second preset value.
And when the second random number is judged to be smaller than the second preset value, the particle position is updated by adopting a strategy of low-altitude flight and slow-descent attack. Specifically, the location update model of the low-altitude flight and slow-descent attack is as follows:
X 3 (t+1)=(X best (t)-X M (t))×α-r 4 +((UB-LB)×r 5 +LB)×δ
wherein X is 3 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, X M (t) is the average position of all particles present, alpha and delta are adaptation parameters, their values are 0.1, UB is the upper boundary, LB is the lower boundary, r 4 And r 5 A random number between 0 and 1. X is X i (t) the number of N particles for the current ith particle position.
And when the second random number is not smaller than the second preset value, updating the particle position by adopting a strategy of short-distance ground attack hunting. Specifically, the location update model of the ground proximity attack prey is:
Wherein X is 4 (t+1) denotes the next updated position of the particle, QF (t) is a mass function for balancing its search strategy, X best (t) represents the best position currently obtained, G 1 The movement parameter of the Alquacral tracking hunting object is [ -1,1]Random number in between, X (t) is the current particle position, r 6 、r 7 And r 8 A random number between 0 and 1, G 2 The flight slope of the chase hunting in the air decreases linearly from 2 to 0, LF (D) represents the Lewy flight function, T is the current iteration number, and T is the maximum iteration number.
In an alternative embodiment, the elite position calculation model for elite position calculated by elite hunting strategy is:
X E (t+1)=X E-best +cd×(X(t)-X E-normal (t))+σ×(X rand1 -X rand2 )
cd=z 0 +0.1×tan(π×(r-0.5))
wherein X is E (t+1) is a new location, X, generated based on elite hunting strategy E-best For randomly selected individuals from the 5 elite individuals, cd is a random number conforming to the cauchy distribution, X (t) is the current particle position, X E-normal (t) 15 positions selected randomly from the warehouse, X rand1 And X is rand2 For random selection from the remaining 15 individuals, exponent=2, σ initial =1σ final =0,z 0 =2, T is the current iteration number, T is the maximum iteration number, r is the distribution in [0,1]Random numbers in between.
A third aspect,
The embodiment of the invention provides a computing device for model parameters, which comprises a processor, a memory and a computer program stored in the memory. The computer program is executable by a processor to implement the method of calculating model parameters as described in the first aspect.
A fourth aspect,
An embodiment of the invention provides a computer readable storage medium comprising a stored computer program, wherein the computer readable storage medium is controlled to execute a method for calculating model parameters according to the first aspect in a device in which the computer readable storage medium is located when the computer program is run.
In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other manners. The apparatus and method embodiments described above are merely illustrative, for example, flow diagrams and block diagrams in the figures illustrate the architecture, functionality, and operation of possible implementations of apparatus, methods and computer program products according to various embodiments of the present invention. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of code, which comprises one or more executable instructions for implementing the specified logical function(s). It should also be noted that in some alternative implementations, the functions noted in the block may occur out of the order noted in the figures. For example, two blocks shown in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It will also be noted that each block of the block diagrams and/or flowchart illustration, and combinations of blocks in the block diagrams and/or flowchart illustration, can be implemented by special purpose hardware-based systems which perform the specified functions or acts, or combinations of special purpose hardware and computer instructions.
In addition, functional modules in the embodiments of the present invention may be integrated together to form a single part, or each module may exist alone, or two or more modules may be integrated to form a single part.
The functions, if implemented in the form of software functional modules and sold or used as a stand-alone product, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention may be embodied essentially or in a part contributing to the prior art or in a part of the technical solution, in the form of a software product stored in a storage medium, comprising several instructions for causing a computer device (which may be a personal computer, an electronic device, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a removable hard disk, a Read-Only Memory (ROM), a random access Memory (RAM, random Access Memory), a magnetic disk, or an optical disk, or other various media capable of storing program codes. It should be noted that, in this document, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising one … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (5)
1. A method for calculating model parameters, comprising:
acquiring a calculation model of a target object to be solved;
solving the calculation model through an improved Alquacrak model to obtain design parameters of the target object; the improved Alquail optimization model calculates an initial population through logistic chaotic mapping, and adopts elite hunting strategy to carry out local development in the development stage so as to avoid premature convergence;
solving the calculation model through an improved Alquacrak model to obtain design parameters of the target object, wherein the method specifically comprises the following steps:
according to the calculation model, randomly generating an initial particle population, and initializing iteration times and maximum iteration times;
according to the initial particle population, calculating to obtain a diversified particle population through a logistic chaotic mapping model;
Calculating the fitness of each particle in the diversified particle population, and judging whether the iteration number is not more than a preset iteration number;
when the iteration number is not greater than the preset iteration number, updating the position of the particle population by adopting a large-range search attack strategy, adding one to the iteration number, and judging whether the maximum iteration number is reached;
when the iteration number is greater than the preset iteration number, updating the position of the particle population by adopting a local attack strategy, calculating to obtain an elite position by adopting an elite hunting strategy, then obtaining the optimal position by adopting a greedy strategy, adding one to the iteration number, and judging whether the maximum iteration number is reached;
when the maximum iteration times are not reached, calculating the fitness of each particle in the updated particle population, and judging whether the iteration times are not more than preset iteration times or not; otherwise, outputting the optimal position and the corresponding fitness of the optimal position when the maximum iteration number is output so as to obtain the design parameters;
the method for updating the position of the particle population by adopting a large-scale search attack strategy specifically comprises the following steps:
acquiring a first random number, and judging whether the random number is smaller than a first preset value;
when the first random number is judged to be smaller than a first preset value, the position of the particle population is updated by adopting a strategy of high-altitude flight searching and vertical bending attack; the position updating model for high-altitude flight searching and vertical bending attack is as follows:
Wherein X is 1 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, X M (T) is the average position of all the current particles, T is the current iteration number, T is the maximum iteration number, N is the population number, r 1 A random number between 0 and 1, X i (t) is the current ith particle position;
when the first random number is judged to be not smaller than a first preset value, a strategy of flying around the prey and attacking in a short-distance gliding mode is adopted to update the position of the particle population; the position updating model of the short-distance gliding attack flying around the prey is as follows:
X 2 (t+1)=X best (t)×LF(D)+X R (t)+(y-x)×r 2
wherein X is 2 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, LF (D) represents the Lewy flight function, D represents the population dimension, X R (t) represents any particleThe position of the son, r 2 For random numbers uniformly distributed between 0 and 1, s=0.01, u and v are random numbers between 0 and 1, β=1.5;
the method for updating the position of the particle population by adopting the local attack strategy specifically comprises the following steps:
acquiring a second random number, and judging whether the random number is smaller than a second preset value;
when the second random number is judged to be smaller than the second preset value, a strategy of low-altitude flight and slow-descent attack is adopted to update the particle position; the position updating model of the low-altitude flight and slow-descent attack is as follows:
X 3 (t+1)=(X best (t)-X M (t))×α-r 4 +((UB-LB)×r 5 +LB)×δ
Wherein X is 3 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, X M (t) is the average position of all particles present, alpha and delta are adaptation parameters, their values are 0.1, UB is the upper boundary, LB is the lower boundary, r 4 And r 5 A random number between 0 and 1; x is X i (t) the number of N particles for the current ith particle position;
when the second random number is judged to be not smaller than the second preset value, the particle position is updated by adopting a strategy of short-distance ground attack of the hunting object; the position updating model of the ground short-distance attack hunting object is as follows:
wherein X is 4 (t+1) denotes the next updated position of the particle, QF (t) is a mass function for balancing its search strategy, X best (t) represents the best position currently obtained, G 1 The movement parameter of the Alquacral tracking hunting object is [ -1,1]Between which are locatedX (t) is the current particle position, r 6 、r 7 And r 8 A random number between 0 and 1, G 2 The flight slope of the hunting object is linearly decreased from 2 to 0 in the air, LF (D) represents the Lewy flight function, T is the current iteration number, and T is the maximum iteration number;
an elite position calculation model for calculating an elite position through an elite hunting strategy is as follows:
X E (t+1)=X E-best +cd×(X(t)-X E-normal (t))+σ×(X rand1 -X rand2 )
cd=z 0 +0.1×tan(π×(r-0.5))
Wherein X is E (t+1) is a new location, X, generated based on elite hunting strategy E-best For randomly selected individuals from the 5 elite individuals, cd is a random number conforming to the cauchy distribution, X (t) is the current particle position, X E-normal (t) 15 positions selected randomly from the warehouse, X rand1 And X is rand2 For random selection from the remaining 15 individuals, exponent=2, σ initial =1σ final =0,z 0 =2, T is the current iteration number, T is the maximum iteration number, r is the distribution in [0,1]Random numbers in between;
the preset iteration times are two thirds of the maximum iteration times;
the target object is a pressure vessel.
2. The method for calculating model parameters according to claim 1, wherein the generating the initial particle population randomly according to the calculation model comprises:
acquiring an upper bound and a lower bound of a search space according to the calculation model;
generating the initial particle population in a search space according to a random function; the initial population generation model is as follows: x=rand× (UB-LB) +lb, X being the particle position, rand being a random number uniformly distributed between 0 and 1, UB being the upper boundary of the search space, LB being the lower boundary of the search space;
the logistic chaotic mapping model is as follows: x (t+1) =a×x (t) × (1-X (t)), where X (t+1) is the updated particle position, a=4, and X (t) is the current particle position;
The fitness calculation model of the fitness of the particles is as follows:
wherein F (x) is a penalty function value, F (x) is an original function value, M is a penalty factor, g i (x) As a constraint, n is the number of particles.
3. A computing device for model parameters, comprising:
the model acquisition module is used for acquiring a calculation model of the target object to be solved;
the model solving module is used for solving the calculation model through an improved Alquacrak model so as to obtain design parameters of the target object; the improved Alquail optimization model calculates an initial population through logistic chaotic mapping, and adopts elite hunting strategy to carry out local development in the development stage so as to avoid premature convergence;
the model solving module specifically comprises:
the initialization unit is used for randomly generating an initial particle population according to the calculation model, and initializing the iteration times and the maximum iteration times;
the chaotic mapping unit is used for calculating and obtaining a diversified particle population through a logistic chaotic mapping model according to the initial particle population;
the fitness unit is used for calculating the fitness of each particle in the diversified particle population and judging whether the iteration number is not more than a preset iteration number;
The large-range unit is used for updating the position of the particle population by adopting a large-range search attack strategy when the iteration number is not more than the preset iteration number, adding one to the iteration number, and judging whether the maximum iteration number is reached;
the local unit is used for updating the position of the particle population by adopting a local attack strategy when the iteration number is greater than the preset iteration number, calculating to obtain an elite position by adopting an elite hunting strategy, obtaining the optimal position by adopting a greedy strategy, adding one to the iteration number, and judging whether the maximum iteration number is reached;
the output unit is used for calculating the fitness of each particle in the updated particle population when judging that the maximum iteration number is not reached, and judging whether the iteration number is not more than the preset iteration number or not; otherwise, outputting the optimal position and the corresponding fitness of the optimal position when the maximum iteration number is output so as to obtain the design parameters;
when the location of the particle population is updated using a broad-range search attack strategy, the broad-range unit is specifically used to:
acquiring a first random number, and judging whether the random number is smaller than a first preset value;
when the first random number is judged to be smaller than a first preset value, the position of the particle population is updated by adopting a strategy of high-altitude flight searching and vertical bending attack; specifically, the location update model for high altitude flight search and vertical bend-over attack is:
Wherein X is 1 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, X M (T) is the average position of all the current particles, T is the current iteration number, T is the maximum iteration number, N is the population number, r 1 A random number between 0 and 1, X i (t) is the current ith particle position;
when the first random number is judged to be not smaller than a first preset value, a strategy of flying around the prey and attacking in a short-distance gliding mode is adopted to update the position of the particle population; the position updating model of the short-distance gliding attack flying around the prey is as follows:
X 2 (t+1)=X best (t)×LF(D)+X R (t)+(y-x)×r 2
wherein X is 2 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, LF (D) represents the Lewy flight function, D represents the population dimension, X R (t) represents the position of any particle, r 2 For random numbers uniformly distributed between 0 and 1, s=0.01, u and v are random numbers between 0 and 1, β=1.5;
when the local attack strategy is adopted to update the position of the particle population, the local unit is specifically used for:
acquiring a second random number, and judging whether the random number is smaller than a second preset value;
when the second random number is judged to be smaller than the second preset value, a strategy of low-altitude flight and slow-descent attack is adopted to update the particle position; specifically, the location update model of the low-altitude flight and slow-descent attack is as follows:
X 3 (t+1)=(X best (t)-X M (t))×α-r 4 +((UB-LB)×r 5 +LB)×δ
Wherein X is 3 (t+1) represents the next updated position of the particle, X best (t) represents the best position currently obtained, X M (t) is the average position of all particles present, alpha and delta are adaptation parameters, their values are 0.1, UB is the upper boundary, LB is the lower boundary, r 4 And r 5 A random number between 0 and 1; x is X i (t) the number of N particles for the current ith particle position;
when the second random number is judged to be not smaller than the second preset value, the particle position is updated by adopting a strategy of short-distance ground attack of the hunting object; specifically, the location update model of the ground proximity attack prey is:
wherein X is 4 (t+1) denotes the next updated position of the particle, QF (t) is a mass function for balancing its search strategy, X best (t) represents the best position currently obtained, G 1 The movement parameter of the Alquacral tracking hunting object is [ -1,1]Random number in between, X (t) is the current particle position, r 6 、r 7 And r 8 A random number between 0 and 1, G 2 The flight slope of the hunting object is linearly decreased from 2 to 0 in the air, LF (D) represents the Lewy flight function, T is the current iteration number, and T is the maximum iteration number;
an elite position calculation model for calculating an elite position through an elite hunting strategy is as follows:
X E (t+1)=X E-best +cd×(X(t)-X E-normal (t))+σ×(X rand1 -X rand2 )
cd=z 0 +0.1×tan(π×(r-0.5))
Wherein X is E (t+1) is a new location, X, generated based on elite hunting strategy E-best For randomly selected individuals from the 5 elite individuals, cd is a random number conforming to the cauchy distribution, X (t) is the current particle position, X E-normal (t) 15 positions selected randomly from the warehouse, X rand1 And X is rand2 For random selection from the remaining 15 individuals, exponent=2, σ initial =1σ final =0,z 0 =2, T is the current iteration number, T is the maximum iteration number, r is the distribution in [0,1]Random numbers in between.
4. A computing device for model parameters, comprising a processor, a memory, and a computer program stored in the memory; the computer program being executable by the processor to implement a method of calculating model parameters according to any one of claims 1 to 2.
5. A computer readable storage medium, characterized in that the computer readable storage medium comprises a stored computer program, wherein the computer program, when run, controls a device in which the computer readable storage medium is located to perform the method of calculating model parameters according to any one of claims 1 to 2.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110822225.5A CN113609761B (en) | 2021-07-21 | 2021-07-21 | Calculation method, device, equipment and storage medium of model parameters |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202110822225.5A CN113609761B (en) | 2021-07-21 | 2021-07-21 | Calculation method, device, equipment and storage medium of model parameters |
Publications (2)
Publication Number | Publication Date |
---|---|
CN113609761A CN113609761A (en) | 2021-11-05 |
CN113609761B true CN113609761B (en) | 2024-02-20 |
Family
ID=78304881
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202110822225.5A Active CN113609761B (en) | 2021-07-21 | 2021-07-21 | Calculation method, device, equipment and storage medium of model parameters |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN113609761B (en) |
Families Citing this family (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115313763A (en) * | 2022-08-09 | 2022-11-08 | 重庆大学 | Heat pipe control method for motor of new energy automobile |
CN117692914A (en) * | 2024-02-02 | 2024-03-12 | Ut斯达康通讯有限公司 | Service function chain mapping method, system, electronic equipment and storage medium |
Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107317338A (en) * | 2017-08-30 | 2017-11-03 | 广东工业大学 | The optimal load flow computational methods and device of a kind of power system |
CN108510074A (en) * | 2018-05-30 | 2018-09-07 | 江苏理工学院 | A kind of implementation method for improving GWO algorithms |
CN109753680A (en) * | 2018-11-20 | 2019-05-14 | 南京南瑞集团公司 | A kind of swarm of particles intelligent method based on chaos masking mechanism |
CN110516831A (en) * | 2019-06-18 | 2019-11-29 | 国网(北京)节能设计研究院有限公司 | A kind of short-term load forecasting method based on MWOA algorithm optimization SVM |
CN110728001A (en) * | 2019-09-29 | 2020-01-24 | 温州大学 | Engineering optimization method of Harris eagle algorithm based on multi-strategy enhancement |
CN111292808A (en) * | 2020-02-14 | 2020-06-16 | 大连大学 | DNA storage coding optimization method based on improved Harris eagle algorithm |
CN111310885A (en) * | 2020-02-27 | 2020-06-19 | 辽宁工程技术大学 | Chaotic space cattle herd search algorithm introducing variation strategy |
CN111506856A (en) * | 2020-03-10 | 2020-08-07 | 燕山大学 | Photovoltaic cell parameter identification method based on improved Harris eagle optimization algorithm |
CN111625816A (en) * | 2020-04-21 | 2020-09-04 | 江西理工大学 | Intrusion detection method and device |
CN111709524A (en) * | 2020-07-03 | 2020-09-25 | 江苏科技大学 | RBF neural network optimization method based on improved GWO algorithm |
CN111832135A (en) * | 2020-07-28 | 2020-10-27 | 郑州轻工业大学 | Pressure container structure optimization method based on improved Harris eagle optimization algorithm |
CN112861427A (en) * | 2021-01-15 | 2021-05-28 | 湖南科技大学 | Permanent magnet synchronous motor parameter identification method based on improved wolf optimization algorithm |
CN112926139A (en) * | 2021-03-23 | 2021-06-08 | 中国人民解放军火箭军工程大学 | Improved intelligent sparrow optimization method based on chaotic mapping and golden sine strategy |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US8504491B2 (en) * | 2010-05-25 | 2013-08-06 | Microsoft Corporation | Variational EM algorithm for mixture modeling with component-dependent partitions |
US10227145B2 (en) * | 2015-02-27 | 2019-03-12 | Space Systems/Loral, Llc | Truss structure |
-
2021
- 2021-07-21 CN CN202110822225.5A patent/CN113609761B/en active Active
Patent Citations (13)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107317338A (en) * | 2017-08-30 | 2017-11-03 | 广东工业大学 | The optimal load flow computational methods and device of a kind of power system |
CN108510074A (en) * | 2018-05-30 | 2018-09-07 | 江苏理工学院 | A kind of implementation method for improving GWO algorithms |
CN109753680A (en) * | 2018-11-20 | 2019-05-14 | 南京南瑞集团公司 | A kind of swarm of particles intelligent method based on chaos masking mechanism |
CN110516831A (en) * | 2019-06-18 | 2019-11-29 | 国网(北京)节能设计研究院有限公司 | A kind of short-term load forecasting method based on MWOA algorithm optimization SVM |
CN110728001A (en) * | 2019-09-29 | 2020-01-24 | 温州大学 | Engineering optimization method of Harris eagle algorithm based on multi-strategy enhancement |
CN111292808A (en) * | 2020-02-14 | 2020-06-16 | 大连大学 | DNA storage coding optimization method based on improved Harris eagle algorithm |
CN111310885A (en) * | 2020-02-27 | 2020-06-19 | 辽宁工程技术大学 | Chaotic space cattle herd search algorithm introducing variation strategy |
CN111506856A (en) * | 2020-03-10 | 2020-08-07 | 燕山大学 | Photovoltaic cell parameter identification method based on improved Harris eagle optimization algorithm |
CN111625816A (en) * | 2020-04-21 | 2020-09-04 | 江西理工大学 | Intrusion detection method and device |
CN111709524A (en) * | 2020-07-03 | 2020-09-25 | 江苏科技大学 | RBF neural network optimization method based on improved GWO algorithm |
CN111832135A (en) * | 2020-07-28 | 2020-10-27 | 郑州轻工业大学 | Pressure container structure optimization method based on improved Harris eagle optimization algorithm |
CN112861427A (en) * | 2021-01-15 | 2021-05-28 | 湖南科技大学 | Permanent magnet synchronous motor parameter identification method based on improved wolf optimization algorithm |
CN112926139A (en) * | 2021-03-23 | 2021-06-08 | 中国人民解放军火箭军工程大学 | Improved intelligent sparrow optimization method based on chaotic mapping and golden sine strategy |
Non-Patent Citations (1)
Title |
---|
一种改进的多目标粒子群优化算法;何骞 等;北京石油化工学院学报;第24卷(第3期);37-43 * |
Also Published As
Publication number | Publication date |
---|---|
CN113609761A (en) | 2021-11-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN113609761B (en) | Calculation method, device, equipment and storage medium of model parameters | |
US20210382445A1 (en) | Control sequence generation system and methods | |
CN106953862B (en) | Sensing method and device for network security situation and sensing model training method and device | |
CN111353582B (en) | Particle swarm algorithm-based distributed deep learning parameter updating method | |
Battiti et al. | Reactive search and intelligent optimization | |
CN105488528B (en) | Neural network image classification method based on improving expert inquiry method | |
EP3732628A1 (en) | Learning data augmentation policies | |
EP3583553A1 (en) | Neural architecture search for convolutional neural networks | |
CN111262858A (en) | Network security situation prediction method based on SA _ SOA _ BP neural network | |
EP3602419B1 (en) | Neural network optimizer search | |
CN104899431B (en) | Based on ant colony and swarm of particles into LSSVM fluctuating wind speed Forecasting Methodologies | |
CN112329948B (en) | Multi-agent strategy prediction method and device | |
CN110134165B (en) | Reinforced learning method and system for environmental monitoring and control | |
CN109508042B (en) | Control method, system and storage medium for intelligent decision-making of unmanned aerial vehicle formation | |
CN105556401B (en) | For area of computer aided control and/or regulate the method for technological system | |
CN108563119A (en) | A kind of unmanned boat motion control method based on fuzzy support vector machine algorithm | |
CN113190037A (en) | Unmanned aerial vehicle optimal path searching method based on improved fluid disturbance and sparrow algorithm | |
CN112272074B (en) | Information transmission rate control method and system based on neural network | |
CN113625569B (en) | Small unmanned aerial vehicle prevention and control decision method and system based on hybrid decision model | |
CN111832135A (en) | Pressure container structure optimization method based on improved Harris eagle optimization algorithm | |
Alhafedh et al. | Two-stage gene selection in microarray dataset using fuzzy mutual information and binary particle swarm optimization | |
US11941867B2 (en) | Neural network training using the soft nearest neighbor loss | |
CN114358197A (en) | Method and device for training classification model, electronic equipment and storage medium | |
CN111667028B (en) | Reliable negative sample determination method and related device | |
Dolan et al. | Learning-based parameter optimization to support dynamic aircraft hazard area generation for space operations |
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 | ||
GR01 | Patent grant | ||
GR01 | Patent grant |