CN106682682A - Method for optimizing support vector machine based on Particle Swarm Optimization - Google Patents
Method for optimizing support vector machine based on Particle Swarm Optimization Download PDFInfo
- Publication number
- CN106682682A CN106682682A CN201610916399.7A CN201610916399A CN106682682A CN 106682682 A CN106682682 A CN 106682682A CN 201610916399 A CN201610916399 A CN 201610916399A CN 106682682 A CN106682682 A CN 106682682A
- Authority
- CN
- China
- Prior art keywords
- particle
- population
- control degree
- optimal control
- fitness
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/241—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches
- G06F18/2411—Classification techniques relating to the classification model, e.g. parametric or non-parametric approaches based on the proximity to a decision surface, e.g. support vector machines
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Computational Linguistics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a method for optimizing a support vector machine (SVM) based on Particle Swarm Optimization (PSO). The embodiments of the invention disclose a method for optimizing a SVM based on PSO, and belongs to the technical field of computer artificial intelligence. According to the embodiments of the invention, on one hand, the method by adjusting inertia weight according to particle fitness, achieves adaptive adjustment of the inertia weight, increases diversity of the inertia weight and better balances global exploration capability and local searching capability of PSO; on the other hand, by taking the threshold values calculated by searching the position of successful particles as variation conditions, the method can better control the timing of particle variation, and after the variation of the particles, the method increases the ability of particles to jump out of local optimal solution, is conducive to optimization of the optimal value of parameters of the SVM, and increases classification accuracy of the SVM algorithm. According to the invention, the method, by optimizing the parameters of a SVM classification model, increase the classification accuracy of the SVM classification model, and can promote wide applications of the SVM classification model in the field of mode identification, system control, production scheduling, computer engineering and electronic communication.
Description
Technical field
The present invention relates to Artificial technical field of intelligence, more particularly to a kind of particle swarm optimization algorithm that is based on is to supporting
The optimization method of vector machine algorithm.
Background technology
Particle swarm optimization algorithm (Particle Swarm Optimization, abbreviation PSO) is a kind of by simulating certainly
The swarm intelligence algorithm of the intelligent behavior of biocenose (such as ant, bird and honeybee) in right boundary.It is each in particle swarm algorithm model
The feasible solution of individual optimization problem regards a particle as, and the oneself state of each particle is by one group of position vector and velocity vector
Description, the feasible solution of difference problem of representation and its direction of motion in D dimensions search space.Particle is by experience and constantly
Study finds its neighbours' optimal solution and group optimal solution, realizes that position changes.
Original particle cluster algorithm is that do not have inertia weight w, Shi and Eberhart to first proposed containing inertia weight w
Particle cluster algorithm, and point out that a larger inertia weight w makes the speed of particle have larger increase, be conducive to particle to go to visit
The unknown space of Suo Xin；One less inertia weight w makes the speed of particle have less change, is conducive to particle local to search
Rope.
During the present invention is realized, at least there are the following problems for the particle cluster algorithm of inventor's discovery prior art：
Because larger inertia weight w can increase global exploring ability, less inertia weight w can increase local search ability, such as
Fruit goes for global exploring ability and the balance both local search ability, it is necessary to which inertia weight w can be adaptive
Change.However, inertia weight w of the prior art is fixed value during PSO algorithm performs or is changed according to iterations,
But these inertia weight w of the prior art can not carry out Automatic adjusument according to the information of population, so just can not cause complete
Office's exploring ability and local search ability are preferably balanced.In addition, inertia weight w is the mechanism of fixed value so that PSO is calculated
Method is easily trapped into locally optimal solution, easy Premature Convergence.
SVMs (Support Vector Machine, abbreviation SVM), is grown up on the basis of statistical learning
Learning algorithm of new generation, the algorithm has stronger advantage on the basis of theory, and in recent years SVMs is in text classification, figure
As the aspects such as classification, bioinformatics, pattern-recognition, system control, production scheduling, computer engineering and data mining are obtained
Extensively application.SVMs improves the generalization ability of learning machine, i.e., as far as possible according to the structural risk minimization of Vapnik
The decision rule obtained by limited training sample, remains to access little error to independent test set.Additionally, SVMs
Algorithm is a convex double optimization problem, ensure that the minimax solution for finding is exactly globally optimal solution.These features make support to
Amount machine becomes a kind of outstanding learning algorithm.
During the present invention is realized, at least there are the following problems for the SVM algorithm of inventor's discovery prior art：
In svm classifier model, C is a parameter in svm classifier model, is represented to the tolerance of classification error or to classification error
Punishment dynamics, C is bigger to represent that punishment is bigger, more can't stand mistake, easily causes over-fitting, C more it is little in contrast, easily
Cause poor fitting.G is Radial basis kernel function (Radial Basis Function) radius, and it affects data to be mapped to new spy
The distribution behind space is levied, g is bigger, and supporting vector is fewer, g values are less, and supporting vector is more, and the number of supporting vector affects instruction
Practice the speed with prediction.So parameter C and g have an impact to the performance of algorithm, reasonable arrange parameter C and g can improve grader
Classification accuracy and grader training and predetermined speed, and existing method is to the limited in one's ability of the two parameter optimizations,
The unreasonable of parameter setting can be caused, so as to cause the classification accuracy of svm classifier model not high.
The content of the invention
The purpose of the present invention is to optimize two parameters of C and g in svm classifier model using modified particle swarm optiziation, is made
Obtain the two parameters and obtain optimal values, realize improving the classification accuracy of SVM algorithm, and then promote algorithm of support vector machine in mould
Formula identification, system control, production scheduling, computer engineering and electronic communication field are more widely applied.
One side according to embodiments of the present invention, there is provided a kind of optimized algorithm based on population is to SVMs
Optimization method, including：
Step S1, initializes to each parameter of population, and the parameter includes the population scale of population, iteration time
Number, search space dimension, the maximum of hunting zone, the minimum of a value of hunting zone, the speed of each particle, position in population
Put, the self-teaching factor and social learning's factor；
Step S2, brings the position initial value of each particle after initialization into fitness function, obtains each particle
Fitness；
Step S3, according to the fitness of each particle, calculates the personal best particle of each particle, individual adaptive optimal control degree
And population optimal location, the population adaptive optimal control degree of population；
Step S4, based on population adaptive optimal control degree and individual adaptive optimal control degree inertia weight is calculated；
Step S5, based on inertia weight, the self-teaching factor, social learning's factor, each particle personal best particle
With the population optimal location of population, speed and the position of each particle are updated；
Step S6, when calculating individual adaptive optimal control degree of each particle in current iteration number of times with a front iterations
Individual adaptive optimal control degree ratio, the ratio and predetermined threshold are compared, if described certain particle ratio is less than pre-
Determine threshold value, then judge the particle search success；
Step S7, calculates the Euclidean distance of the position of the successful particle of search to the population optimal location, and to all
Search for the Euclidean distance corresponding to successful particle to average, obtain distance threshold；
Whether step S8, the position for judging each particle is less than the distance to the Euclidean distance of the population optimal location
Threshold value, some particles if so, then adjusted the distance in threshold value carry out mutation operation；
Whether step S9, judge current iteration number of times less than the iterations for setting, if it is not, then execution step S10；
Step S10, output population is mapped as the population optimal location to support in current population optimal location
Penalty factor and Radial basis kernel function radius g in vector machine；
Step S11, is trained according to the penalty factor and Radial basis kernel function radius g to SVMs.
The beneficial effect of the embodiment of the present invention：On the one hand the embodiment of the present invention adjusts inertia weight w according to particle fitness,
It is achieved thereby that the self-adaptative adjustment of inertia weight w, increased the diversity of inertia weight, preferably balance PSO algorithms are global
On the other hand exploring ability and local search ability, by using searching for threshold value that successful particle calculates as variation
Condition, can better control over the opportunity of particle variations, and particle after variation, jump out the ability of locally optimal solution and carried by particle
Rise, be more beneficial for finding the optimal value of parameter C and g, ultimately help the classification accuracy for improving SVM algorithm.
Description of the drawings
Fig. 1 is optimization method flow chart of optimized algorithm of the present invention based on population to SVMs；
Fig. 2 is the use particle individuality adaptive optimal control degree mean value and individual adaptive optimal control degree inertia weight distribution map of the present invention；
Fig. 2 a are the inertia weight values for using particle individuality adaptive optimal control degree mean value to obtain；
Fig. 2 b are inertia weight values acquired when being spent using particle individuality adaptive optimal control；
Fig. 3 is the schematic diagram of the non-definitive variation particle of population of the present invention；
Fig. 4 is the schematic diagram that the population of the present invention has determined that variation particle；
Fig. 5 is based on original in optimization method and prior art based on AIWPSO algorithms to svm classifier model of the invention
Beginning particle cluster algorithm is to the optimization method of svm classifier model in classification accuracy and the temporal comparative result figure of classification；
Fig. 5 a are based on original grain in optimization method and prior art based on AIWPSO algorithms to svm classifier model of the invention
Swarm optimization to the optimization method of svm classifier model classification accuracy comparative result figure；
Fig. 5 b are based on original in optimization method and prior art based on AIWPSO algorithms to svm classifier model of the invention
Beginning particle cluster algorithm is to the optimization method of svm classifier model in the comparative result figure of time of classifying.
Specific embodiment
To make the object, technical solutions and advantages of the present invention of greater clarity, with reference to specific embodiment and join
According to accompanying drawing, the present invention is described in more detail.It should be understood that these descriptions are simply exemplary, and it is not intended to limit this
Bright scope.Additionally, in the following description, the description to known features and technology is eliminated, to avoid unnecessarily obscuring this
The concept of invention.
Fig. 1 is stream of optimized algorithm of the present invention based on population to the first embodiment of the optimization method of SVMs
Cheng Tu.
As shown in figure 1, a kind of optimization method of optimized algorithm based on population to SVMs, comprises the following steps
S1 to S10：
Step S1, initializes to each parameter of population, and the parameter includes the population scale of population, iteration time
Number, search space dimension, the maximum of hunting zone, the minimum of a value of hunting zone, the speed of each particle, position in population
Put, the self-teaching factor and social learning's factor.
Before each parameter to population is initialized, the parameters for setting population, setting population rule are needed
Mould is s (i.e. the population includes s particle), maximum iteration time is T, search space dimension is D, the minimum of hunting zone
It is worth for popmin, hunting zone maximum be popmax, the speed for setting each particle in population is x, society as v, position
Studying factors c1, the self-teaching factor be c2。
The speed of each particle in population is wherein set as v, is also included, set maximal rate as Vmax, minimum speed be
Vmin.After the completion of above parameter setting, start to initialize population, so-called initialization enters above parameters
Row assignment so that each gain of parameter initial value.
Here to population it should be noted that carry out initializing including the speed to each particle in population and position
Put imparting when carrying out assignment is random value.Specifically, the speed of each particle is initialized based on following formula (1)；It is based on
Following formula (2) is initialized to the position of each particle, wherein, rand () is the random number between interval [0,1].
V=rand () formula (1)
The formulas of x=200rand () -100 (2)
Step S2, brings the position initial value of each particle after initialization into fitness function, obtains each particle
Fitness.
Based on each particle, the initial value of its position obtained after population initialization is brought into fitness function, obtain
To the fitness of each particle.Specifically, population after initialization, an initial position that each particle is obtained (i.e. position
Initial value).The penalty factor and radial direction initial value of these positions being mapped as in SVMs (svm classifier model)
Base kernel function radius g, is trained, based on formula (3) according to penalty factor and Radial basis kernel function radius g to SVMs
Obtain fitness.
Wherein, n is training set total sample number, and r is the correct number of samples of classification, and F is fitness.In particle cluster algorithm pair
In the optimization of SVMs, fitness F is the classification accuracy for training the svm classifier model for finishing to training set, and classification is accurate
Really the bigger explanation classifying quality of rate is better.Reach and export after the iterations of setting fitness (being exactly here classification accuracy)
The maximum corresponding population optimal location of that group, and population optimal location is mapped as parameter C and g, then population is to SVM's
Parameter C and g optimizing are finished.
Step S3, according to the fitness of each particle, calculates the personal best particle of each particle, individual adaptive optimal control degree
And population optimal location, the population adaptive optimal control degree of population.
Wherein, in optimization of the particle cluster algorithm to SVMs, the fitness in particle cluster algorithm is exactly svm classifier
Classification accuracy of the model (SVMs) to training set, so individual adaptive optimal control degree is each particle in whole iteration mistake
The value of the fitness maximum obtained in journey；Population adaptive optimal control degree is that all particles are individual in whole iterative process in population
Maximum in body adaptive optimal control degree；Personal best particle is the position corresponding to the particle of individual adaptive optimal control degree；Population is most
Position of the excellent position corresponding to the particle of population adaptive optimal control degree.
It should be noted that all particles in population are during iteration, and per iteration once, each search space
The position of the particle in dimension all can change once.For population scale be s, iterations be t, search space dimension for D
For population, if particle iteration is once, then for each particle, have D positional value, this D positional value band
Enter to be obtained in fitness function fitness of the particle in current iteration.If particle iteration t time, by particle each
Positional value in iteration is substituted into fitness function, then obtain t fitness.From the work that selective value in this t fitness is maximum
The individual adaptive optimal control degree for being the particle in whole iterative process, the corresponding position of the individual adaptive optimal control degree is the particle
Personal best particle.After the individual adaptive optimal control degree of each particle determines, then compare the individual adaptive optimal control degree of s particle,
The therefrom maximum population adaptive optimal control degree as the population of selective value, the corresponding position of population adaptive optimal control degree as should
The population optimal location of population.
Step S4, based on population adaptive optimal control degree and individual adaptive optimal control degree inertia weight is obtained.
Specifically, population adaptive optimal control degree and individual adaptive optimal control degree are substituted into formula (4) and is calculated, obtain inertia power
Weight.
Wherein, what i was characterized is that particle is i-th, and t characterizes iteration to t time, and w characterizes inertia weight, wiT () characterizes i-th
Particle iteration is to inertia weight value when t time, and it is optimum to population when t time that fitness (gbest) (t) characterizes population iteration
Fitness, fitness (pbest)iT () characterizes i-th particle iteration to individual adaptive optimal control degree when t time.
Fig. 2 is distribution map of the inertia weight of the particle iteration 20 times of population 20 in plane coordinate system.
Wherein, transverse axis represents iterations, and the longitudinal axis represents particle inertia weighted value, and Fig. 2 a are to use particle individual optimum suitable
Response mean value fitness (pbest)averageThe inertia weight value of acquirement, Fig. 2 b are to use particle individuality adaptive optimal control degree
fitness(pbest)iWhen acquired inertia weight value.
As shown in Figure 2 a, inertia weight high concentration, almost all of particle correspondence identical inertia weight value.Such as Fig. 2 b
Shown, each iteration of correspondence, the inertia weight distribution of particle is wider, (0.5,1.5) between.It can therefore be seen that making
With particle individuality adaptive optimal control degree fitness (pbest)iCompare using particle individuality adaptive optimal control degree mean value fitness
(pbest)average, the inertia weight value of acquirement more has diversity, so can ensure that particle in global search and local
Search has the division of labor, so that algorithm obtains effectively balance between global exploring ability and local search ability.
Step S5, based on inertia weight, the self-teaching factor, social learning's factor, each particle personal best particle
With the population optimal location of population, speed and the position of each particle are updated.
Based on inertia weight, the self-teaching factor, social learning's factor, the personal best particle of each particle and population
Population optimal location calculated, obtain the speed of each particle and the updated value of position, and by each grain after initialization
The speed of son and the initial value of position replace with the updated value.
Specifically, step S5 comprises the following steps S51-S52：
Step S51, by inertia weight, the self-teaching factor, social learning's factor, the personal best particle of each particle and
The population optimal location of population substitutes into formula (5) and is calculated, and obtains the speed after particle updates.
vij(t+1)=wvij(t)+c1r1[pbestij(t)-xij(t)]+c2r2[gbestj(t)-xij(t)] formula (5)
Step S52, by inertia weight, the self-teaching factor, social learning's factor, the personal best particle of each particle and
The population optimal location of population substitutes into formula (6) and is calculated, and obtains the position after particle updates.
xij(t+1)=xij(t)+vij(t+1) formula (6)
Wherein, what i was characterized is that particle is i-th, and what j was characterized is the jth dimension of particle, and x characterizes the position of particle, t tables
Iteration is levied to t time, w characterizes inertia weight, vijT () is characterized in the speed of jth dimension space when i-th particle iterates to the t time,
vij(t+1) characterize when i-th particle iterates to the t+1 time in the speed of jth dimension space, xijT () characterizes i-th particle iteration
To when the t time in the position of jth dimension space, xij(t+1) characterize when i-th particle iterates to the t+1 time in the position of jth dimension space
Put, pbestijT () is characterized when i-th particle iterates to the t time in the personal best particle of jth dimension space, gbestjT () characterizes
When population iterates to the t time population jth dimension space population optimal location, c1For social learning's factor, c2For self
Practise the factor, r1And r2For the random number in interval [0,1].
Step S6, when calculating individual adaptive optimal control degree of each particle in current iteration number of times with a front iterations
Individual adaptive optimal control degree ratio, the ratio and predetermined threshold are compared, if described certain particle ratio is less than pre-
Determine threshold value, then judge the particle search success.
Specifically, set predetermined threshold as 1, if individual adaptive optimal control degree of certain particle in current iteration number of times with it is front
The ratio of individual adaptive optimal control degree during an iteration number of times is less than 1, then judge the particle search success, if certain particle is being worked as
Individual adaptive optimal control degree during front iterations is equal to 1 with the ratio of individual adaptive optimal control degree during a front iterations, then
Judge the particle search failure.Further, it is possible to it is 1 to arrange the successful characterization value of particle search, the sign of particle search failure
It is worth for 0, judges whether each particle is searched for successfully based on following formula (8).
Wherein, SS (i, t)=1 represents i-th particle search success, and SS (i, t)=0 represents that i-th particle search loses
Lose.I-th particle iteration is represented to individual adaptive optimal control degree when t time,For
The i particles iteration is to individual adaptive optimal control degree when t-1 time.
Step S7, calculates the Euclidean distance of the position of the successful particle of search to the population optimal location, and to all
Search for the Euclidean distance corresponding to successful particle to average, obtain distance threshold.
Due to having judged whether each particle is searched for successfully in step S6, according to the judged result of each particle, with regard to energy
All quantity for searching for successful particle in the population are counted, the number based on the successful particle of all search in the population
Amount can calculate the search success rate of the population.
Specifically, the position of i-th particle in the successful particle of search is calculated to population optimal location based on following formula (9)
Euclidean distance：
Wherein, distiCharacterize the Euclidean distance of the position of i-th particle to population optimal location, gbestjCharacterize jth dimension
Population optimal location, xijThe position of the jth dimension of i-th particle is characterized, D characterizes search space dimension.
The successful particle of all search is calculated to the mean value of the Euclidean distance of population optimal location based on following formula (10)：
Wherein, distaverageCharacterize mean value (i.e. distance threshold)；M characterizes the number of the successful particle of search.
Whether step S8, judge the position of each particle to the Euclidean distance of population optimal location less than described apart from threshold
Value, some particles if so, then adjusted the distance in threshold value carry out mutation operation.
Judge whether the position of each particle is less than apart from threshold to the Euclidean distance of population optimal location based on following formula (11)
Value.
Wherein, distiCharacterize the Euclidean distance of the position of i-th particle to population optimal location, mutiRepresent and judge knot
Really.If muti=1 represents that the particle is fallen into distance threshold, muti=0 represents that the particle is not fallen within distance threshold.
It should be noted that after it is determined that whether particle is fallen into distance threshold, it is not right to carry out during mutation operation
The all particles fallen into distance threshold carry out mutation operation, because there may exist optimum in all particles in distance threshold
Position, so needing selected section particle variations when variation, retains a part of particle and maintains as former state.
The some particles are the particle or 1/3rd particle of half." half " " 1/3rd " are an experiences
Value, it is also possible to select other values, in a preferred embodiment of the invention, the half particle in threshold value of adjusting the distance enters row variation behaviour
Make.
The some particles adjusted the distance in threshold value carry out mutation operation based on formula (7), obtain each grain in some particles
Position of the son after variation.
Pop (i)=(popmax-popmin)·rand()+popmin (7)
Wherein, position of i-th particle of pop (i) signs after variation, popmaxCharacterize the hunting zone of population most
Big value, popminThe minimum of a value of the hunting zone of population is characterized, rand () is the random number in interval [0,1].
Fig. 3 is the non-definitive variation particle schematic diagram of population.As shown in figure 3, the little solid dot of black represents particle, band in figure
The solid dot for having circle represents the population optimal location that current search is arrived.
Fig. 4 is the schematic diagram that population has determined variation particle.As shown in figure 4, the little solid dot of black represents grain in figure
Son, the solid dot with circle represents the population optimal location that current search is arrived, distaverageFor distance threshold.
Whether step S9, judge current iteration number of times less than the iterations for setting, if it is not, then execution step S10.
Step S10, output population is mapped as the population optimal location to support in current population optimal location
Penalty factor and Radial basis kernel function radius g in vector machine (svm classifier model).
Step S11, is trained according to the penalty factor and Radial basis kernel function radius g to SVMs.
Further, described method, wherein, in step S11, according to the penalty factor and Radial basis kernel function radius g
After being trained to SVMs, also include：
The fitness of each particle when obtaining current iteration number of times, and return to step S3.
After training is finished, the fitness of each particle is obtained, and the fitness is substituted into step S3.
C is penalty factor in SVMs, characterizes the tolerance to classification error.G is characterized in SVMs
The radius of Radial basis kernel function (Radial Basis Function).
Beneficial effects of the present invention are illustrated below by way of experimental data.
Illustrate the beneficial effect of the algorithm being optimized to population in the present invention (in order to state by experimental data first
Convenience, hereinafter referred to as AIWPSO algorithms).
Inventor used 11 test functions shown in table 1 below in an experiment to test prior art in five kinds of particles
Group's innovatory algorithm (CPSO algorithms, RPSO algorithms, LDPSO algorithms, NLDPSO algorithms, APSO algorithms) calculates with the AIWPSO of the present invention
Optimizing situation of the method to test function.This 11 test functions include unimodal function, Solving Multimodal Function.
Table 1
The relevant information of 11 functions shown in table 1 is as shown in the following Table 2.Wherein, the global optimum in table 2 is test
The minimum of a value that function can be got, (CPSO algorithms, RPSO algorithms, LDPSO algorithms, NLDPSO algorithms, APSO are calculated any of the above algorithm
Method, AIWPSO algorithms) if to the optimizing result of test function closer to global optimum, showing that the low optimization accuracy of algorithm is higher.
| Test function title | Search space dimension | Hunting zone | Global optimum position | Global optimum | |
| f1 | Sphere | 30 | [-100,100]D | [0,…,0]D | 0 |
| f2 | Schwefel P2.22 | 30 | [-10,10]D | [0,…,0]D | 0 |
| f3 | Rosenbrock | 30 | [-30,30]D | [1,…,1]D | 0 |
| f4 | Noisy Quadric | 30 | [-1.28,1.28]D | [0,…,0]D | 0 |
| f5 | Rastrigin | 30 | [-5.12,5.12]D | [0,…,0]D | 0 |
| f6 | Griewank | 30 | [-600,600]D | [0,…,0]D | 0 |
| f7 | Ackley | 30 | [-32,32]D | [0,…,0]D | 0 |
| f8 | Rotated hyper ellipsoid | 30 | [-100,100]D | [0,…,0]D | 0 |
| f9 | Rotated Rastrigin | 30 | [-5,5]D | [0,…,0]D | 0 |
| f10 | Rotated Griewank | 30 | [-600,600]D | [0,…,0]D | 0 |
| f11 | Shifted Rotated Rastrigin | 30 | [-600,600]D | [0,…,0]D | -330 |
Table 2
In order to reduce the impact that random error is brought, to five kinds of population innovatory algorithms in prior art in this experiment
(CPSO algorithms, RPSO algorithms, LDPSO algorithms, NLDPSO algorithms, APSO algorithms) uses phase with the AIWPSO algorithms of the present invention
Same test function independent operating 30 times, the minimum of a value of each algorithm optimizing result, mean value, standard deviation are as shown in table 3 below.
Table 3
As shown in table 3, wherein, Min is represented by taking the minimum of a value in result after algorithm independent operating 30 times, at this
In test, as a result represent that algorithm low optimization accuracy is higher closer to the global optimum in table 2；Mean is represented to algorithm independent operating
To this 30 results averageds after 30 times；SD represents the standard deviation of this 30 results, and standard deviation embodies the stability of algorithm,
The less explanation algorithm of standard deviation is more stable.AIWPSO algorithms based on the present invention are improved with five kinds of populations of the prior art and calculated
Method (CPSO algorithms, RPSO algorithms, LDPSO algorithms, NLDPSO algorithms, APSO algorithms) is compared, and the AIWPSO based on the present invention is calculated
Method to the minimum of a value of the optimizing result of test function closer to respective function in table 2 global optimum, and the mark of optimizing result
Quasi- difference is less, therefore the AIWPSO algorithms low optimization accuracy of the present invention is high, algorithm performance is stable.
Prove that the AIWPSO algorithms of invention improve the classification accuracy of svm classifier model below by way of experimental data.
This experiment adopts UCI machine learning storehouse (UCI Machine Learning Repository：http://
Archive.ics.uci.edu/ml/ the data set in) assesses different disaggregated models.UCI machine learning storehouse is University of California
The database for machine learning that Irving branch school (University of California Irvine) proposes, UCI data sets
It is a conventional standard testing data set.Wherein, the data set used in the present invention includes heart disease data set
(Statlog), diabetes data collection (Diabetes), thoracic surgery data set (Thoracic Surgery), breast cancer data set
(Breast Cancer), liver diseases data set (Liver Disorders) totally 5 data sets, the tool of 5 data sets of the above
Body information is referring to table 4 below.
| Dataset name | Sample size | Number of features | Classification number | Training set number | Test set number |
| Heart disease data set | 270 | 10 | 2 | 150 | 120 |
| Diabetes data collection | 768 | 5 | 2 | 500 | 268 |
| Thoracic surgery data set | 470 | 9 | 2 | 300 | 170 |
| Breast cancer data set | 699 | 9 | 2 | 500 | 199 |
| Liver diseases data set | 345 | 4 | 2 | 200 | 145 |
Table 4
In raw data set information, comprising sex, the two features of age, refer to not as characteristic of division in Classification and Identification
Mark.For other characteristic informations in data set, this experiment has used Statistical Identifying Method to enter the ga s safety degree of characteristic index
Row differentiates that, by statistical check, the classification indicators that there is significant difference between group could be used as characteristic of division.
The optimization method (abbreviation PSO-SVM) based on predecessor group algorithm to svm classifier model is compared by experiment
With the present invention based on AIWPSO algorithms to the classification accuracy of the optimization method (abbreviation AIWPSO-SVM) of svm classifier model and
The classification time.The feature tag of normal person is 1, and the feature tag of patient is 2.Experiment porch is to associate M490PC, 32
Windows7 operating systems, Intel's Duo i5 three generations's processors, the calculating frequency of CPU is 2.50GHz, and running memory is 4GB,
Software version is MATLAB R2013b.Tested using LIBSVM kits.Experimental result is as shown in Figure 5.
Fig. 5 is based on original in optimization method and prior art based on AIWPSO algorithms to svm classifier model of the invention
Beginning particle cluster algorithm is to the optimization method of svm classifier model in classification accuracy and the temporal comparative result figure of classification.
Wherein, Fig. 5 a are in the optimization method and prior art based on AIWPSO algorithms to svm classifier model of the present invention
Based on predecessor group algorithm to the optimization method of svm classifier model classification accuracy comparative result figure.
Fig. 5 b are based on original in optimization method and prior art based on AIWPSO algorithms to svm classifier model of the invention
Beginning particle cluster algorithm is to the optimization method of svm classifier model in the comparative result figure of time of classifying.
Statlog, Diabetes, Thoracic Surgery, Breast as shown in Fig. 5 a, 5b, on transverse axis
Cancer, Liver Disorders represents that respectively use in the present invention 5 data sets (are corresponding in turn to heart disease data set, sugar
The sick data set of urine, thoracic surgery data set, breast cancer data set, liver diseases data set), wherein the part with shade is existing skill
Optimization method of the predecessor group algorithm to svm classifier model is based in art, shadeless part is that the present invention is based on
Optimization method of the AIWPSO algorithms to svm classifier model.The longitudinal axis in Fig. 5 a is classification accuracy axle, and the longitudinal axis in Fig. 5 b is to divide
Class time shaft (unit:Second).
Can draw from Fig. 5 a, Fig. 5 b：1, for classification accuracy：AIWPSO algorithms based on the present invention are than existing
Classification accuracy in technology based on predecessor group's algorithm optimization svm classifier model is high；2, for Riming time of algorithm：This
Although the AIWPSO algorithms of invention in prior art relative to being based on, and original particle cluster algorithm run time is longer, difference is not
Greatly.So, the present invention significantly improves the classification accuracy of svm classifier model on the basis of the long time is not lost.
It should be appreciated that the above-mentioned specific embodiment of the present invention is used only for exemplary illustration or explains the present invention's
Principle, and be not construed as limiting the invention.Therefore, that what is done in the case of without departing from the spirit and scope of the present invention is any
Modification, equivalent, improvement etc., should be included within the scope of the present invention.Additionally, claims of the present invention
Whole changes in the equivalents for being intended to fall into scope and border or this scope and border and
Modification.
Claims (10)
1. optimization method of a kind of optimized algorithm based on population to SVMs, it is characterised in that include：
Step S1, initializes to each parameter of population, the population scale of the parameter including population, iterations,
Search space dimension, the maximum of hunting zone, the minimum of a value of hunting zone, the speed of each particle in population, position, from
My Studying factors and social learning's factor；
Step S2, brings the position initial value of each particle after initialization into fitness function, obtains the adaptation of each particle
Degree；
Step S3, according to the fitness of each particle, calculate the personal best particle of each particle, individual adaptive optimal control degree and
The population optimal location of population, population adaptive optimal control degree；
Step S4, based on population adaptive optimal control degree and individual adaptive optimal control degree inertia weight is calculated；
Step S5, based on inertia weight, the self-teaching factor, social learning's factor, the personal best particle of each particle and grain
The population optimal location of subgroup, updates speed and the position of each particle；
Step S6, calculate individual adaptive optimal control degree of each particle in current iteration number of times with during a front iterations
The ratio of body adaptive optimal control degree, the ratio and predetermined threshold are compared, if described certain particle ratio is less than predetermined threshold
Value, then judge the particle search success；
Step S7, calculates the Euclidean distance of the position of the successful particle of search to the population optimal location, and to all search
Successfully the Euclidean distance corresponding to particle is averaged, and obtains distance threshold；
Whether step S8, judge the position of each particle to the Euclidean distance of the population optimal location less than described apart from threshold
It is worth, if so, the particle then to being less than distance threshold carries out mutation operation；
Whether step S9, judge current iteration number of times less than the iterations for setting；
Step S10, if current iteration number of times exports the current population optimal location of population less than the iterations of setting,
And the penalty factor and Radial basis kernel function radius g being mapped as the population optimal location in SVMs；
Step S11, is trained according to the penalty factor and Radial basis kernel function radius g to SVMs.
2. method according to claim 1, wherein, in step S11, according to the penalty factor and Radial basis kernel function
After radius g is trained to SVMs, also include：
Obtain the fitness of each particle, and return to step S3.
3. method according to claim 1, wherein, in step S1, carrying out initialization to population is included to particle
The speed of each particle in group and position are initialized, and initialized mode is that the speed to particle and position give at random
Value.
4. method according to claim 3, wherein,
The speed of each particle is initialized based on formula (1)；
V=rand () formula (1)
The position of each particle is initialized based on formula (2)；
The formulas of x=200rand () -100 (2)
Wherein, rand () is the random number between [0,1].
5. method according to claim 1, wherein, the step 2 is based on each particle, by it after population initialization
The initial value of the position for obtaining brings fitness function into, obtains the fitness of each particle and includes：
Step 21, by the initial value of the position of each particle in population the penalty factor in SVMs and footpath are mapped as
To base kernel function radius g；
Step 22, is trained, based on formula (3) according to the penalty factor and Radial basis kernel function radius g to SVMs
Obtain fitness；
Wherein, n is training set total sample number, and r is the correct number of samples of classification, and F is fitness.
6. the method according to any one of claim 1-5, wherein, in step S3,
Individual adaptive optimal control degree is the value of the fitness maximum that each particle is obtained in whole iterative process；
Population adaptive optimal control degree is the maximum in population in individual adaptive optimal control degree of all particles in whole iterative process
Value；
Personal best particle is the position corresponding to the particle of individual adaptive optimal control degree；
Position of the population optimal location corresponding to the particle of population adaptive optimal control degree.
7. the method according to any one of claim 1-5, wherein, step S4, based on population adaptive optimal control degree and
Body adaptive optimal control degree obtains inertia weight, including：
Population adaptive optimal control degree and individual adaptive optimal control degree are substituted into formula (4) and calculated, inertia weight is obtained；
Wherein, what i was characterized is that particle is i-th, and t characterizes iteration to t time, and w characterizes inertia weight；
wiT () characterizes i-th particle iteration to inertia weight value when t time；
Fitness (gbest) (t) characterizes population iteration to population adaptive optimal control degree when t time；
fitness(pbest)iT () characterizes i-th particle iteration to individual adaptive optimal control degree when t time.
8. the method according to any one of claim 1-5, wherein, step S5 includes：
Step S51, by inertia weight, the self-teaching factor, social learning's factor, the personal best particle of each particle and particle
The population optimal location of group substitutes into formula (5) and is calculated, and obtains the speed after particle updates；
vij(t+1)=wvij(t)+c1r1[pbestij(t)-xij(t)]+c2r2[gbestj(t)-xij(t)] formula (5)
Step S52, by inertia weight, the self-teaching factor, social learning's factor, the personal best particle of each particle and particle
The population optimal location of group substitutes into formula (6) and is calculated, and obtains the position after particle updates；
xij(t+1)=xij(t)+vij(t+1) formula (6)
Wherein, what i was characterized is that particle is i-th, and what j was characterized is the jth dimension of particle；
X characterizes the position of particle, and t characterizes iteration to t time, and w characterizes inertia weight；
vijT () is characterized when i-th particle iterates to the t time in the speed of jth dimension space；
vij(t+1) characterize when i-th particle iterates to the t+1 time in the speed of jth dimension space；
xijT () is characterized when i-th particle iterates to the t time in the position of jth dimension space；
xij(t+1) characterize when i-th particle iterates to the t+1 time in the position of jth dimension space；
pbestijT () is characterized when i-th particle iterates to the t time in the personal best particle of jth dimension space；
gbestjT () characterizes population optimal location of the population in jth dimension space when population iterates to the t time；
c1For social learning's factor, c2For the self-teaching factor；
r1And r2For the random number in interval [0,1].
9. the method according to any one of claim 1-5, wherein, step S8 includes：
The some particles adjusted the distance in threshold value carry out mutation operation based on formula (7), obtain each particle in some particles and exist
Position after variation；
Pop (i)=(popmax-popmin)·rand()+popmin(7)；
Wherein, position of i-th particle of pop (i) signs after variation；
popmaxCharacterize the maximum of the hunting zone of population；
popminCharacterize the minimum of a value of the hunting zone of population；
Rand () is the random number in interval [0,1].
10. the method according to any one of claim 1-5, wherein, in step S8, some particles for half particle
Or 1/3rd particle.
Priority Applications (2)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610916399.7A CN106682682A (en) | 2016-10-20 | 2016-10-20 | Method for optimizing support vector machine based on Particle Swarm Optimization |
| PCT/CN2017/070894 WO2018072351A1 (en) | 2016-10-20 | 2017-01-11 | Method for optimizing support vector machine on basis of particle swarm optimization algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610916399.7A CN106682682A (en) | 2016-10-20 | 2016-10-20 | Method for optimizing support vector machine based on Particle Swarm Optimization |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN106682682A true CN106682682A (en) | 2017-05-17 |
Family
ID=58840147
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610916399.7A Pending CN106682682A (en) | 2016-10-20 | 2016-10-20 | Method for optimizing support vector machine based on Particle Swarm Optimization |
Country Status (2)
| Country | Link |
|---|---|
| CN (1) | CN106682682A (en) |
| WO (1) | WO2018072351A1 (en) |
Cited By (31)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107103357A (en) * | 2017-05-23 | 2017-08-29 | 沈阳航空航天大学 | A kind of new dandelion algorithm |
| CN107247844A (en) * | 2017-06-10 | 2017-10-13 | 福州大学 | The minimum tree algorithms of X architecture Steiner based on adaptive PSO and mixing switching strategy |
| CN108364030A (en) * | 2018-03-20 | 2018-08-03 | 东北大学 | A kind of multi-categorizer model building method based on three layers of dynamic particles group's algorithm |
| CN108363838A (en) * | 2018-01-18 | 2018-08-03 | 上海电力学院 | Temperature effect forecast method in electrostatic precipitator based on ATPSO-SVM models |
| CN108539571A (en) * | 2018-04-08 | 2018-09-14 | 上海交通大学 | A kind of fast automatic mode locking method covering multimode pulse recognition |
| CN108594290A (en) * | 2018-05-02 | 2018-09-28 | 成都理工大学 | A kind of spectral line modification method |
| CN108615069A (en) * | 2018-03-25 | 2018-10-02 | 哈尔滨工程大学 | A kind of optimized calculation method based on improved adaptable quanta particle swarm optimization |
| CN108629155A (en) * | 2018-05-14 | 2018-10-09 | 浙江大学 | A kind of leukaemia cancer cell detector that parameter is optimal |
| CN109150873A (en) * | 2018-08-16 | 2019-01-04 | 武汉虹旭信息技术有限责任公司 | Malice domain name detection system and method based on PSO_SVM optimization algorithm |
| CN109739959A (en) * | 2018-11-30 | 2019-05-10 | 东软集团股份有限公司 | Method and device used in being calculated in topic association |
| CN110070458A (en) * | 2019-03-15 | 2019-07-30 | 福建商学院 | The method for manufacturing Dynamic Scheduling |
| CN110096927A (en) * | 2018-01-30 | 2019-08-06 | 西安交通大学 | Contactor diagnostic method and diagnostic system based on particle group optimizing support vector machines |
| CN110390419A (en) * | 2019-05-20 | 2019-10-29 | 重庆大学 | Freeway toll station method for predicting based on PSO-LSSVM model |
| CN110728231A (en) * | 2019-10-10 | 2020-01-24 | 华东理工大学 | Sleep staging method based on improved particle swarm algorithm and twin support vector machine |
| CN111047102A (en) * | 2019-12-18 | 2020-04-21 | 江南大学 | Express delivery distribution route optimization method based on elite-driven particle swarm algorithm |
| CN111210075A (en) * | 2020-01-07 | 2020-05-29 | 国网辽宁省电力有限公司朝阳供电公司 | Lightning stroke transmission line fault probability analysis method based on combined classifier |
| CN111275078A (en) * | 2020-01-13 | 2020-06-12 | 南京航空航天大学 | Optimization method of support vector machine for part image recognition |
| CN111643321A (en) * | 2020-04-30 | 2020-09-11 | 北京精密机电控制设备研究所 | Exoskeleton joint angle prediction method and system based on sEMG signals |
| CN111681258A (en) * | 2020-06-12 | 2020-09-18 | 上海应用技术大学 | Hybrid enhanced intelligent trajectory prediction method and device based on hybrid wolf optimization SVM |
| CN111709584A (en) * | 2020-06-18 | 2020-09-25 | 中国人民解放军空军研究院战略预警研究所 | Radar networking optimization deployment method based on artificial bee colony algorithm |
| CN111717217A (en) * | 2020-06-30 | 2020-09-29 | 重庆大学 | Driver intention identification method based on probability correction |
| CN111736618A (en) * | 2020-06-28 | 2020-10-02 | 清华大学 | Unmanned motorcycle steering control parameter setting method and device |
| CN111950604A (en) * | 2020-07-27 | 2020-11-17 | 江苏大学 | Image identification and classification method of multi-classification support vector machine based on minimum reconstruction error search reduction and particle swarm optimization |
| CN112308229A (en) * | 2020-11-26 | 2021-02-02 | 西安邮电大学 | Dynamic multi-objective evolution optimization method based on self-organizing mapping |
| CN112365117A (en) * | 2020-09-03 | 2021-02-12 | 中交西安筑路机械有限公司 | Pavement structure performance calculation method based on optimized support vector machine |
| CN113570555A (en) * | 2021-07-07 | 2021-10-29 | 温州大学 | Two-dimensional segmentation method of multi-threshold medical image based on improved grasshopper algorithm |
| CN113759722A (en) * | 2021-09-13 | 2021-12-07 | 桂林电子科技大学 | Parameter optimization method for active disturbance rejection controller of unmanned aerial vehicle |
| CN115222007A (en) * | 2022-05-31 | 2022-10-21 | 复旦大学 | Improved particle swarm parameter optimization method for glioma multitask integrated network |
| CN115412671A (en) * | 2022-08-29 | 2022-11-29 | 特斯联科技集团有限公司 | Camera shutter artificial intelligence adjusting method and system for monitoring moving object |
| CN115880572A (en) * | 2022-12-19 | 2023-03-31 | 江苏海洋大学 | Forward-looking sonar target identification method based on asynchronous learning factor |
| CN111950604B (en) * | 2020-07-27 | 2024-05-14 | 江苏大学 | Image recognition classification method of multi-classification support vector machine based on minimum reconstruction error search dimension reduction and particle swarm optimization |
Families Citing this family (224)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108763926B (en) * | 2018-06-01 | 2021-11-12 | 中国电子技术标准化研究院 | Industrial control system intrusion detection method with safety immunity capability |
| CN109063242B (en) * | 2018-06-20 | 2022-11-11 | 中国人民解放军国防科技大学 | Guidance tool error identification method based on particle swarm optimization |
| CN110689156B (en) * | 2018-07-04 | 2023-03-07 | 新智数字科技有限公司 | Universal energy station optimization method and device |
| CN109086497B (en) * | 2018-07-16 | 2023-06-02 | 中国科学院宁波材料技术与工程研究所 | Method for developing potential energy force field of metal and alloy based on particle swarm algorithm |
| CN109087367B (en) * | 2018-07-27 | 2022-09-27 | 西安航空学院 | High-spectrum image rapid compressed sensing reconstruction method based on particle swarm optimization |
| CN108984946B (en) * | 2018-08-03 | 2023-03-10 | 安徽大学 | Power network key node identification method based on multi-objective optimization algorithm |
| CN109146984B (en) * | 2018-08-14 | 2022-11-22 | 西安航空学院 | Particle swarm optimization-based hyperspectral image sparse decomposition method |
| CN109190270B (en) * | 2018-09-12 | 2022-12-27 | 北京化工大学 | APSO-BP-based double-counterweight-disc automatic balance control method |
| CN109766562B (en) * | 2018-09-27 | 2023-04-07 | 中原工学院 | Cycloidal gear tooth profile modification method based on genetic algorithm and particle swarm combined algorithm |
| CN109766988A (en) * | 2018-09-28 | 2019-05-17 | 中国人民解放军空军工程大学 | Target cluster dividing method based on chaos ant lion optimization algorithm |
| CN109238715B (en) * | 2018-10-31 | 2024-01-26 | 合肥工业大学 | Bearing fault signal enhancement method and system |
| CN109635999B (en) * | 2018-11-06 | 2023-06-20 | 华中科技大学 | Hydropower station scheduling method and system based on particle swarm-bacterial foraging |
| CN109508779B (en) * | 2018-11-09 | 2023-10-13 | 重庆化工职业学院 | Energy-saving control method for municipal street lamp |
| CN109657274B (en) * | 2018-11-16 | 2023-08-29 | 广东省建筑设计研究院 | Suspended dome cable force optimization method based on particle swarm optimization algorithm in building structure |
| CN109726456A (en) * | 2018-12-14 | 2019-05-07 | 重庆大学 | Magnetic resonance based on Chaos particle swarm optimization algorithm couples spiral winding optimum design method |
| CN111381600B (en) * | 2018-12-28 | 2022-11-04 | 陕西师范大学 | UUV path planning method based on particle swarm optimization |
| CN109635880B (en) * | 2019-01-08 | 2023-06-27 | 浙江大学 | Coal mining machine fault diagnosis system based on robust self-adaptive algorithm |
| CN109713665B (en) * | 2019-01-12 | 2023-06-23 | 湖北鄂电德力电气有限公司 | Minimum collision set algorithm suitable for multiple multiphase faults of power distribution network |
| CN109873810B (en) * | 2019-01-14 | 2022-07-19 | 湖北工业大学 | Network fishing detection method based on goblet sea squirt group algorithm support vector machine |
| CN109921472B (en) * | 2019-03-11 | 2022-11-04 | 上海电力学院 | Power system equivalent inertia evaluation method based on particle swarm optimization algorithm |
| CN109936141A (en) * | 2019-03-28 | 2019-06-25 | 广州番禺职业技术学院 | A kind of Economic Dispatch method and system |
| CN110110753B (en) * | 2019-04-03 | 2023-08-25 | 河南大学 | Effective mixed characteristic selection method based on elite flower pollination algorithm and ReliefF |
| CN110008634B (en) * | 2019-04-19 | 2023-04-18 | 华北水利水电大学 | Method and system for determining parameters of bi-quad generalized integrator frequency-locked loop |
| CN110111275B (en) * | 2019-04-29 | 2022-11-29 | 武汉工程大学 | Method and system for signal noise reduction and computer storage medium |
| CN110097119A (en) * | 2019-04-30 | 2019-08-06 | 西安理工大学 | Difference secret protection support vector machine classifier algorithm based on dual variable disturbance |
| CN110111001B (en) * | 2019-05-06 | 2023-07-28 | 广东工业大学 | Site selection planning method, device and equipment for electric vehicle charging station |
| CN110210087B (en) * | 2019-05-20 | 2022-11-11 | 中国科学院光电技术研究所 | Light beam jitter model parameter real-time identification method based on particle swarm optimization |
| CN110211638B (en) * | 2019-05-28 | 2023-03-24 | 河南师范大学 | Gene selection method and device considering gene correlation |
| CN110175413B (en) * | 2019-05-29 | 2024-01-19 | 国网上海市电力公司 | Power distribution network reconstruction method and device based on R2 index multi-target particle swarm algorithm |
| CN110288634A (en) * | 2019-06-05 | 2019-09-27 | 成都启泰智联信息科技有限公司 | A kind of method for tracking target based on Modified particle swarm optimization algorithm |
| 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 |
| CN110276140B (en) * | 2019-06-26 | 2023-01-06 | 温州大学 | Method for predicting response time of electromagnet |
| CN110334026B (en) * | 2019-07-03 | 2023-03-24 | 浙江理工大学 | CS-SPSO algorithm-based combined test case generation method |
| CN110309604A (en) * | 2019-07-05 | 2019-10-08 | 江苏师范大学 | A kind of machine components profile data predication method |
| CN110321638A (en) * | 2019-07-05 | 2019-10-11 | 江苏师范大学 | A kind of destilling tower number of plates acquisition methods based on Gilliland correlation |
| CN110309613B (en) * | 2019-07-09 | 2022-10-04 | 大连海事大学 | Design and optimization method of tunnel excavation step method based on BIM |
| CN110308649B (en) * | 2019-07-11 | 2022-10-14 | 东南大学 | PID parameter optimization method based on PSO-SOA fusion algorithm and applied to industrial process control |
| CN110348642A (en) * | 2019-07-12 | 2019-10-18 | 国网四川省电力公司信息通信公司 | A kind of optimization method of the support vector machines for volume forecasting |
| CN110389902A (en) * | 2019-07-18 | 2019-10-29 | 江苏科技大学 | Software reliability method for parameter estimation based on artificial bee colony Hybrid Particle Swarm |
| CN111008549B (en) * | 2019-08-07 | 2024-01-26 | 哈尔滨工程大学 | UUV platform DVL signal distortion reconstruction method based on sample entropy and IFOA-GRNN |
| CN110674915B (en) * | 2019-09-18 | 2022-11-25 | 东北大学 | Irregular pipeline defect inversion method based on improved particle swarm optimization |
| CN110738726B (en) * | 2019-09-27 | 2023-04-18 | 华南理工大学 | Robot vision-guided three-dimensional object reconstruction method based on octree |
| CN110852344A (en) * | 2019-09-27 | 2020-02-28 | 武汉船舶职业技术学院 | Intelligent substation network fault classification based method |
| CN110728001B (en) * | 2019-09-29 | 2023-08-04 | 温州大学 | Engineering optimization method based on multi-strategy enhancement Harisk hawk algorithm |
| CN110750756B (en) * | 2019-10-01 | 2023-06-20 | 深圳市行健自动化股份有限公司 | Real-time on-line instrument checksum diagnosis method through optimal support vector machine algorithm |
| CN110598804B (en) * | 2019-10-14 | 2023-05-09 | 安徽理工大学 | Improved FastSLAM method based on clustering and membrane calculation |
| CN110955865B (en) * | 2019-10-18 | 2023-12-29 | 浙江工业大学 | Data envelope analysis DEA method based on particle filtering |
| CN110765706B (en) * | 2019-10-23 | 2024-03-01 | 扬州大学 | Aerofoil unsteady stall aerodynamic coefficient modeling method based on OHNGBM (1, 1) |
| CN110990940B (en) * | 2019-10-28 | 2023-03-24 | 西北工业大学 | Wing assembly positioning layout design method based on MSVR |
| CN110909856B (en) * | 2019-11-13 | 2023-04-18 | 西安工业大学 | Improved fruit fly optimization method for mechanical roundness error evaluation |
| CN111046527A (en) * | 2019-11-18 | 2020-04-21 | 山东科技大学 | Battery equivalent parameter identification method based on coevolution particle swarm algorithm |
| CN111079208B (en) * | 2019-11-20 | 2024-01-23 | 杭州电子科技大学 | Particle swarm algorithm-based CAD model surface corresponding relation identification method |
| CN110956641A (en) * | 2019-11-20 | 2020-04-03 | 南京拓控信息科技股份有限公司 | Train wheel tread image segmentation method based on chemical reaction optimization |
| CN111242971B (en) * | 2019-12-03 | 2023-05-02 | 西安电子科技大学 | Target tracking method based on improved double-center particle swarm optimization algorithm |
| CN110942205B (en) * | 2019-12-05 | 2022-12-06 | 国网安徽省电力有限公司 | Short-term photovoltaic power generation power prediction method based on HIMVO-SVM |
| CN111127139B (en) * | 2019-12-06 | 2023-06-27 | 成都理工大学 | Mixed recommendation algorithm based on ProbS and HeatS calculation mode improvement |
| CN111159857B (en) * | 2019-12-13 | 2024-02-13 | 天津大学 | Two-dimensional transient temperature field reconstruction method for sonic nozzle pipe wall |
| CN111222284B (en) * | 2019-12-27 | 2023-05-26 | 中国大唐集团科学技术研究院有限公司西北电力试验研究院 | Method for integrally and flexibly measuring primary air quantity of inlet of medium-speed coal mill unit |
| CN111080035A (en) * | 2019-12-31 | 2020-04-28 | 芜湖哈特机器人产业技术研究院有限公司 | Global path planning method based on improved quantum particle swarm optimization algorithm |
| CN111259600B (en) * | 2020-01-19 | 2023-07-28 | 西北大学 | Optimization efficiency method for improving automatic well position optimization |
| CN111325238B (en) * | 2020-01-21 | 2023-06-09 | 全球能源互联网研究院有限公司 | Phase noise compensation method and system |
| CN111325308B (en) * | 2020-02-14 | 2023-03-28 | 集美大学 | Nonlinear system identification method |
| CN111353582B (en) * | 2020-02-19 | 2022-11-29 | 四川大学 | Particle swarm algorithm-based distributed deep learning parameter updating method |
| CN111428748B (en) * | 2020-02-20 | 2023-06-27 | 重庆大学 | HOG feature and SVM-based infrared image insulator identification detection method |
| CN111310902B (en) * | 2020-02-24 | 2023-09-29 | 石家庄铁道大学 | Training method of neural network model, slope displacement prediction method and related devices |
| CN111368892B (en) * | 2020-02-27 | 2024-01-30 | 合肥工业大学 | Electric energy quality disturbance efficient identification method for generalized S transformation and SVM |
| CN111371607B (en) * | 2020-02-28 | 2022-09-16 | 大连大学 | Network flow prediction method for optimizing LSTM based on decision-making graying algorithm |
| CN111383710A (en) * | 2020-03-13 | 2020-07-07 | 闽江学院 | Gene splice site recognition model construction method based on particle swarm optimization gemini support vector machine |
| CN111429419B (en) * | 2020-03-19 | 2023-04-07 | 国网陕西省电力公司电力科学研究院 | Insulator contour detection method based on hybrid ant colony algorithm |
| CN111488208B (en) * | 2020-03-22 | 2023-10-31 | 浙江工业大学 | Bian Yun collaborative computing node scheduling optimization method based on variable-step-size bat algorithm |
| CN111444649B (en) * | 2020-03-24 | 2022-10-18 | 成都理工大学 | Slope system reliability analysis method based on intensity reduction method |
| CN111525547B (en) * | 2020-03-24 | 2023-06-16 | 云南电网有限责任公司临沧供电局 | Low-voltage intelligent treatment method based on optimal reactive compensation |
| CN111462157B (en) * | 2020-03-31 | 2023-04-07 | 西安工程大学 | Infrared image segmentation method based on genetic optimization threshold method |
| CN111563920B (en) * | 2020-04-15 | 2023-04-07 | 西安工程大学 | 3D color point cloud registration method based on global optimization and multi-constraint condition iteration |
| CN111695233B (en) * | 2020-04-20 | 2023-03-28 | 安徽博微长安电子有限公司 | Array element failure correction method based on improved whale optimization algorithm |
| CN111580145B (en) * | 2020-04-27 | 2022-07-22 | 山东大学 | Dynamic measurement method for accelerator dosage |
| CN111487995B (en) * | 2020-04-30 | 2023-04-07 | 湖南科技大学 | Multi-target search cooperation method for group unmanned aerial vehicle based on three-dimensional simplified virtual model |
| CN111597651B (en) * | 2020-04-30 | 2023-05-02 | 上海工程技术大学 | Rolling bearing performance degradation evaluation method based on HWPSO-SVDD model |
| CN111695290B (en) * | 2020-05-14 | 2024-04-09 | 天津大学 | Short-term runoff intelligent forecasting mixed model method suitable for changing environment |
| CN113704949B (en) * | 2020-05-21 | 2023-10-13 | 北京机械设备研究所 | Method for establishing nonlinear model of electric steering engine based on particle swarm optimization algorithm |
| CN111639695B (en) * | 2020-05-26 | 2024-02-20 | 温州大学 | Method and system for classifying data based on improved drosophila optimization algorithm |
| CN111627495B (en) * | 2020-06-01 | 2023-03-14 | 集美大学 | Method for judging species value of population |
| CN113764047A (en) * | 2020-06-05 | 2021-12-07 | 中国石油天然气股份有限公司 | Propylene polymerization quality on-line measuring system |
| CN111814839B (en) * | 2020-06-17 | 2023-09-01 | 合肥工业大学 | Template matching method of longicorn group optimization algorithm based on self-adaptive variation |
| CN111679685B (en) * | 2020-06-19 | 2023-04-07 | 中国人民解放军国防科技大学 | Unmanned aerial vehicle total energy based flight control method and device |
| CN111812041A (en) * | 2020-06-29 | 2020-10-23 | 重庆邮电大学 | Portable water body COD (chemical oxygen demand) measuring system and method |
| CN111860622B (en) * | 2020-07-03 | 2023-12-22 | 北京科技大学 | Clustering method and system applied to programming field big data |
| CN111898725A (en) * | 2020-07-07 | 2020-11-06 | 西安建筑科技大学 | Air conditioning system sensor fault detection method and device and electronic equipment |
| CN111930435B (en) * | 2020-07-13 | 2023-04-28 | 兰州理工大学 | Task unloading decision method based on PD-BPSO technology |
| CN111709494B (en) * | 2020-07-13 | 2023-05-26 | 哈尔滨工业大学 | Novel hybrid-optimized image stereo matching method |
| CN111815061B (en) * | 2020-07-17 | 2023-07-11 | 河北工业大学 | Task priority dividing method for solving overhead travelling crane scheduling process based on SVM |
| CN111832507B (en) * | 2020-07-20 | 2024-04-09 | 安徽大学 | Wheat scab remote sensing identification method based on wheat head spectral information |
| CN111950615B (en) * | 2020-07-31 | 2023-12-05 | 武汉烽火技术服务有限公司 | Network fault feature selection method based on tree species optimization algorithm |
| CN111880140A (en) * | 2020-08-03 | 2020-11-03 | 中北大学 | RSSI-based wireless sensor network arc triangle positioning method |
| CN112085059B (en) * | 2020-08-06 | 2023-10-20 | 温州大学 | Breast cancer image feature selection method based on improved sine and cosine optimization algorithm |
| CN112116952B (en) * | 2020-08-06 | 2024-02-09 | 温州大学 | Gene selection method of gray wolf optimization algorithm based on diffusion and chaotic local search |
| CN111950622B (en) * | 2020-08-10 | 2023-08-15 | 中国平安人寿保险股份有限公司 | Behavior prediction method, device, terminal and storage medium based on artificial intelligence |
| CN111951291B (en) * | 2020-08-13 | 2024-02-06 | 哈尔滨商业大学 | Infrared image edge detection method based on multi-structure morphology and FODPSO mixed processing |
| CN112017733B (en) * | 2020-08-24 | 2022-11-11 | 郑州大学 | Particle swarm algorithm-based high polymer slurry parameter identification method |
| CN112016663B (en) * | 2020-08-24 | 2022-11-11 | 郑州大学 | Polymer slurry parameter identification method based on group intelligent optimization algorithm |
| CN112052933B (en) * | 2020-08-31 | 2022-04-26 | 浙江工业大学 | Particle swarm optimization-based safety testing method and repairing method for deep learning model |
| CN112257897B (en) * | 2020-09-17 | 2024-03-22 | 华北电力大学 | Electric vehicle charging optimization method and system based on improved multi-target particle swarm |
| CN112101814B (en) * | 2020-09-25 | 2024-04-16 | 吴俊江 | Oil-gas engineering classification method and system based on weighted fuzzy clustering algorithm |
| CN112100893B (en) * | 2020-09-25 | 2024-04-02 | 西安交通大学 | Discontinuous domain optimization method for finite element calculation |
| CN111985144B (en) * | 2020-09-27 | 2023-07-18 | 江西师范大学 | IDW interpolation method for multi-parameter collaborative optimization of geometrics data |
| CN112346010B (en) * | 2020-09-28 | 2022-06-10 | 中国人民解放军海军航空大学 | Dual-computer passive positioning method based on scale difference and time difference |
| CN112181867B (en) * | 2020-09-29 | 2022-07-26 | 西安电子科技大学 | On-chip network memory controller layout method based on multi-target genetic algorithm |
| CN112308288A (en) * | 2020-09-29 | 2021-02-02 | 百维金科(上海)信息科技有限公司 | Particle swarm optimization LSSVM-based default user probability prediction method |
| CN112200353B (en) * | 2020-09-30 | 2022-06-17 | 重庆师范大学 | Support vector machine weather prediction method based on improved quantum optimization algorithm |
| CN112183884A (en) * | 2020-10-19 | 2021-01-05 | 河南工业大学 | Grain storage quality prediction method and device |
| CN112230678B (en) * | 2020-10-29 | 2023-07-14 | 皖江工学院 | Three-dimensional unmanned aerial vehicle path planning method and system based on particle swarm optimization |
| CN112199897B (en) * | 2020-11-02 | 2022-10-14 | 国网重庆市电力公司电力科学研究院 | Particle swarm algorithm-based improved GIS equipment abnormal sound vibration identification method |
| CN112328364B (en) * | 2020-11-05 | 2022-07-08 | 北京理工大学 | Computing-intensive cloud workflow scheduling method based on farmland fertility algorithm |
| CN112330164B (en) * | 2020-11-09 | 2022-06-03 | 国网电力科学研究院武汉南瑞有限责任公司 | Data quality management system and method based on message bus |
| CN114547954A (en) * | 2020-11-24 | 2022-05-27 | 中国移动通信集团浙江有限公司 | Logistics distribution center site selection method and device and computer equipment |
| CN112558119B (en) * | 2020-11-30 | 2023-10-10 | 中航机载系统共性技术有限公司 | Satellite selection method based on self-adaptive BFO-PSO |
| CN112801127A (en) * | 2020-12-09 | 2021-05-14 | 西安华谱电力设备制造有限公司 | Cable partial discharge defect identification method based on oscillation waves |
| CN112446435B (en) * | 2020-12-10 | 2023-12-05 | 长春理工大学 | City data classification method and system |
| CN112487816B (en) * | 2020-12-14 | 2024-02-13 | 安徽大学 | Named entity identification method based on network classification |
| CN112669169B (en) * | 2020-12-15 | 2024-04-30 | 国网辽宁省电力有限公司阜新供电公司 | Short-term photovoltaic power prediction device and method |
| CN112668446A (en) * | 2020-12-24 | 2021-04-16 | 常州大学 | Flower pollination algorithm-based method for monitoring wear state of micro milling cutter by optimizing SVM (support vector machine) |
| CN112668078B (en) * | 2020-12-24 | 2022-05-17 | 青岛理工大学 | Method for identifying damage of rusted reinforced concrete beam after fire disaster |
| CN112667876B (en) * | 2020-12-24 | 2024-04-09 | 湖北第二师范学院 | Opinion leader group identification method based on PSOTVCF-Kmeans algorithm |
| CN112763988B (en) * | 2020-12-24 | 2023-12-26 | 西安电子科技大学 | Anti-interference waveform design method based on self-adaptive binary particle swarm genetic algorithm |
| CN112765845B (en) * | 2021-01-04 | 2024-03-05 | 华东理工大学 | Sensor array optimization method for damage positioning of pressure vessel with connecting pipe |
| CN112668248B (en) * | 2021-01-20 | 2023-12-26 | 中国建筑土木建设有限公司 | Method and system for dispatching optimization calculation theoretical model of concrete truck |
| CN112668247B (en) * | 2021-01-20 | 2023-12-26 | 中国建筑土木建设有限公司 | Construction channel thickness optimization design method and system |
| CN112800682B (en) * | 2021-02-04 | 2022-10-04 | 中国长江三峡集团有限公司 | Feedback optimization fan blade fault monitoring method |
| CN112992291B (en) * | 2021-02-04 | 2023-07-18 | 中国科学院沈阳自动化研究所 | High-temperature electrical grade magnesium oxide powder batching optimization method |
| CN112784811B (en) * | 2021-02-09 | 2023-06-23 | 西安科技大学 | Global optimal ultrasonic signal denoising method |
| CN112862055B (en) * | 2021-02-11 | 2024-01-12 | 西北工业大学 | Cluster behavior quantitative analysis method considering consistency and density of clustered objects |
| CN113034554B (en) * | 2021-02-27 | 2024-03-29 | 西北大学 | Whale optimized broken warrior body fragment registration method based on chaos reverse learning |
| CN113141317B (en) * | 2021-03-05 | 2022-09-30 | 西安电子科技大学 | Streaming media server load balancing method, system, computer equipment and terminal |
| CN112884368B (en) * | 2021-03-23 | 2022-11-01 | 合肥工业大学 | Multi-target scheduling method and system for minimizing delivery time and delay of high-end equipment |
| CN113112130B (en) * | 2021-03-23 | 2022-09-30 | 合肥工业大学 | High-end equipment manufacturing process quality on-line monitoring method and system |
| CN113011589B (en) * | 2021-03-29 | 2024-03-08 | 湖北工业大学 | Co-evolution-based hyperspectral image band selection method and system |
| CN113435101B (en) * | 2021-04-01 | 2023-06-30 | 国网内蒙古东部电力有限公司 | Particle swarm optimization-based power failure prediction method for support vector machine |
| CN113051771B (en) * | 2021-04-09 | 2024-03-26 | 中国科学院苏州生物医学工程技术研究所 | Triple quadrupole mass spectrometer parameter optimization method and system based on particle swarm optimization |
| CN113011680A (en) * | 2021-04-16 | 2021-06-22 | 西安建筑科技大学 | Power load prediction method and system |
| CN113376536A (en) * | 2021-04-22 | 2021-09-10 | 安徽锐能科技有限公司 | Data-driven high-precision lithium battery SOC (State of Charge) joint estimation method and system |
| CN113239761B (en) * | 2021-04-29 | 2023-11-14 | 广州杰赛科技股份有限公司 | Face recognition method, device and storage medium |
| CN113269426A (en) * | 2021-05-18 | 2021-08-17 | 哈尔滨工程大学 | Dormitory distribution method based on heuristic improved particle swarm optimization |
| CN113158495B (en) * | 2021-05-21 | 2024-04-26 | 华东理工大学 | Method and system for generating molecular sieve process conditions based on particle swarm optimization |
| CN113327322B (en) * | 2021-05-26 | 2022-02-18 | 国勘数字地球(北京)科技有限公司 | Inversion method and device of surface wave frequency dispersion curve and computer readable storage medium |
| CN113376541B (en) * | 2021-06-04 | 2023-06-06 | 辽宁工程技术大学 | Lithium ion battery health state prediction method based on CRJ network |
| CN113505914B (en) * | 2021-06-17 | 2023-05-26 | 广东工业大学 | Injection molding machine hydraulic system energy consumption prediction method, system and equipment for optimizing SVM |
| CN113435304B (en) * | 2021-06-23 | 2023-09-19 | 西安交通大学 | Method, system, device and storage medium for extracting torsional vibration information of torsional vibration signal |
| CN113393451B (en) * | 2021-06-25 | 2024-03-29 | 江南大学 | Defect detection method based on automatic machine learning |
| CN113283573B (en) * | 2021-06-25 | 2024-03-29 | 江南大学 | Automatic searching method for optimal structure of convolutional neural network |
| CN113407895B (en) * | 2021-06-29 | 2023-01-20 | 广东电网有限责任公司 | Flash bird repelling optimal frequency selection method and system based on simulated annealing algorithm |
| CN113821317B (en) * | 2021-07-02 | 2023-08-11 | 华侨大学 | Side cloud cooperative microservice scheduling method, device and equipment |
| CN113353582A (en) * | 2021-07-02 | 2021-09-07 | 东北大学 | Conveyor belt carrier roller equipment fault detection method based on audio information and PSO-MSVM |
| CN113643335B (en) * | 2021-07-13 | 2024-04-12 | 西北大学 | Three-dimensional image registration method based on EDF-DE model and application |
| CN113759333B (en) * | 2021-07-14 | 2024-04-02 | 中国人民解放军空军预警学院 | Wind turbine multipath echo jiggle parameter estimation method based on whale optimization algorithm |
| CN113806992A (en) * | 2021-07-23 | 2021-12-17 | 任长江 | Optimization method based on convection-dispersion theory |
| CN113642613B (en) * | 2021-07-23 | 2023-10-10 | 温州大学 | Medical disease feature selection method based on improved goblet sea squirt swarm algorithm |
| CN113590587A (en) * | 2021-07-30 | 2021-11-02 | 湘潭大学 | Offline position fingerprint database construction method based on self-adaptive simulated annealing-particle swarm-kriging interpolation algorithm |
| CN113707216A (en) * | 2021-08-05 | 2021-11-26 | 北京科技大学 | Infiltration immune cell proportion counting method |
| CN113690933B (en) * | 2021-08-06 | 2022-11-29 | 合肥工业大学 | Grid-connected inverter phase-locked loop parameter identification method |
| CN113673015B (en) * | 2021-08-10 | 2023-08-25 | 石家庄铁道大学 | Intelligent system construction and parameter identification method for beam column end plate connection node optimization design |
| CN113807486B (en) * | 2021-08-23 | 2023-09-26 | 南京邮电大学 | Multi-robot area coverage method based on improved particle swarm algorithm |
| CN113951881B (en) * | 2021-08-23 | 2023-06-16 | 南方医科大学南方医院 | Remote nursing monitoring system |
| CN113887691B (en) * | 2021-08-24 | 2022-09-16 | 杭州电子科技大学 | Whale evolution system and method for service combination problem |
| CN113852909A (en) * | 2021-08-26 | 2021-12-28 | 广州杰赛科技股份有限公司 | Node positioning method, device, equipment and storage medium of wireless network |
| CN113642220B (en) * | 2021-08-26 | 2023-09-22 | 江苏科技大学 | Ship welding process optimization method based on RBF and MOPSO |
| CN113873428A (en) * | 2021-08-31 | 2021-12-31 | 广州杰赛科技股份有限公司 | Wireless network node positioning method, device, equipment and medium |
| CN113673136A (en) * | 2021-09-06 | 2021-11-19 | 长安大学 | Method, system and equipment for predicting safe thickness of top plate of karst tunnel |
| CN113761689B (en) * | 2021-09-13 | 2024-03-26 | 河北工程大学 | Multi-parameter coupling optimization method and system for aerodynamic engine |
| CN113887692A (en) * | 2021-09-15 | 2022-01-04 | 中南大学 | Research method of controlled particle group based on group activity sensing |
| CN113869514B (en) * | 2021-09-17 | 2024-01-16 | 中林信达（北京）科技信息有限责任公司 | Multi-knowledge integration and optimization method based on genetic algorithm |
| CN113848709A (en) * | 2021-09-17 | 2021-12-28 | 昌吉学院 | Boiler drum water level control method based on improved quantum particle swarm algorithm |
| CN113781002B (en) * | 2021-09-18 | 2023-07-21 | 北京航空航天大学 | Low-cost workflow application migration method based on agent model and multiple group optimization in cloud edge cooperative network |
| CN113687190A (en) * | 2021-09-22 | 2021-11-23 | 云南民族大学 | Distributed power supply containing power distribution network fault positioning method based on SABSO algorithm |
| CN114118130B (en) * | 2021-09-28 | 2024-04-05 | 西安交通大学 | Transformer partial discharge mode identification method, system, medium and equipment |
| CN113988125A (en) * | 2021-10-25 | 2022-01-28 | 西安交通大学 | Torsional vibration signal instantaneous frequency extraction method based on improved synchronous compression transformation |
| CN114090967A (en) * | 2021-10-25 | 2022-02-25 | 广州大学 | APT (android package) organization tracing and tracing method and system based on PSO-MSVM (Power System-Mobile virtual machine) |
| CN113987806B (en) * | 2021-10-29 | 2024-04-26 | 吉林大学 | Atmosphere mode optimization method based on proxy model |
| CN113916860B (en) * | 2021-11-02 | 2023-04-25 | 淮阴工学院 | Pesticide residue type identification method based on fluorescence spectrum |
| CN114065594B (en) * | 2021-11-30 | 2024-04-02 | 西安交通大学 | Single-post insulator electrical performance optimization method for GIS based on neural network model |
| CN114200960A (en) * | 2021-12-02 | 2022-03-18 | 杭州电子科技大学 | Unmanned aerial vehicle cluster search control optimization method for improving sparrow algorithm based on tabu table |
| CN114169098B (en) * | 2021-12-07 | 2024-01-19 | 西安交通大学 | Advanced safe injection box parameter optimization method based on predator strategy and particle swarm optimization |
| CN114459501B (en) * | 2021-12-10 | 2023-08-25 | 华中光电技术研究所（中国船舶重工集团公司第七一七研究所） | Automatic calibration method and device for inertial navigation installation errors |
| CN114244720A (en) * | 2021-12-17 | 2022-03-25 | 湘潭大学 | Multi-controller deployment method based on improved particle swarm algorithm in SDN environment |
| CN114019985B (en) * | 2021-12-20 | 2023-12-22 | 中国海洋大学 | Unmanned rudder direction control design method based on fractional order PID and particle swarm algorithm |
| CN114528907B (en) * | 2021-12-31 | 2023-04-07 | 北京交通大学 | Industrial abnormal data detection method and device |
| CN114415510A (en) * | 2022-01-17 | 2022-04-29 | 江西理工大学 | Magnetic suspension train speed tracking method |
| CN114599004B (en) * | 2022-01-28 | 2024-01-05 | 北京邮电大学 | Base station layout method and device |
| CN114422952B (en) * | 2022-01-29 | 2024-05-03 | 南京邮电大学 | Indoor fingerprint positioning method based on improved LSSVR |
| CN114665971B (en) * | 2022-03-21 | 2023-10-13 | 北京理工大学 | Method for generating multi-mode superimposed beam for improving communication capacity |
| CN114757084B (en) * | 2022-04-07 | 2024-03-05 | 西北工业大学 | Dynamic wave beam hopping method for broadband satellite communication system |
| CN114580306B (en) * | 2022-04-24 | 2022-07-29 | 锦浪科技股份有限公司 | Flyback transformer design method based on improved PSO algorithm |
| CN114947825A (en) * | 2022-04-29 | 2022-08-30 | 吉林大学 | Lower limb prosthesis continuous motion recognition method based on PSOGWO-SVM |
| CN114792071B (en) * | 2022-05-18 | 2023-08-25 | 西安理工大学 | Drainage pump station optimal scheduling method based on machine learning technology |
| CN114936577B (en) * | 2022-05-23 | 2024-03-26 | 大连大学 | Mixed image blind separation method based on improved lion group algorithm |
| CN114896736B (en) * | 2022-05-27 | 2024-04-09 | 太原理工大学 | Anchor rod drill carriage drill arm positioning control method and system based on improved particle swarm optimization |
| CN114899844B (en) * | 2022-06-13 | 2023-06-20 | 华能国际电力股份有限公司德州电厂 | Primary frequency modulation control system of small-disturbance thermal power generating unit |
| CN115081325B (en) * | 2022-06-21 | 2024-03-15 | 桂林电子科技大学 | Lens antenna multi-objective optimization method based on particle swarm and genetic hybrid algorithm |
| CN115021816A (en) * | 2022-07-04 | 2022-09-06 | 吉林大学 | Power distribution method of visible light communication system based on improved goblet sea squirt group algorithm |
| CN115146389B (en) * | 2022-07-19 | 2024-03-12 | 江西理工大学 | Permanent magnet magnetic levitation train dynamics feature modeling method |
| CN115208703B (en) * | 2022-09-16 | 2022-12-13 | 北京安帝科技有限公司 | Industrial control equipment intrusion detection method and system of fragment parallelization mechanism |
| CN116307021B (en) * | 2022-10-08 | 2024-03-22 | 中国大唐集团科学技术研究总院有限公司 | Multi-target energy management method of new energy hydrogen production system |
| CN115876229B (en) * | 2022-10-14 | 2023-06-20 | 哈尔滨理工大学 | Novel encoder angle zero-crossing jump point suppression method and device based on particle swarm |
| CN115375204B (en) * | 2022-10-25 | 2023-02-03 | 中国人民解放军陆军装甲兵学院 | Vehicle-mounted intelligent micro-grid performance evaluation method |
| CN115696354B (en) * | 2022-10-26 | 2024-04-19 | 金陵科技学院 | High-speed rail mobile communication system network coverage method based on improved particle swarm |
| CN115640918B (en) * | 2022-12-26 | 2023-04-07 | 电子科技大学中山学院 | Cable temperature anomaly prediction method, device, medium and equipment |
| CN116307533B (en) * | 2023-02-21 | 2023-10-20 | 珠江水利委员会珠江水利科学研究院 | Reservoir group flood control dispatching intelligent method, system and medium |
| CN115865762A (en) * | 2023-03-03 | 2023-03-28 | 上海人工智能网络系统工程技术研究中心有限公司 | Spatial information network flow prediction method and scheduling system based on SDN framework |
| CN116070151B (en) * | 2023-03-17 | 2023-06-20 | 国网安徽省电力有限公司超高压分公司 | Ultra-high voltage direct current transmission line fault detection method based on generalized regression neural network |
| CN116152316B (en) * | 2023-04-17 | 2023-07-07 | 山东省工业技术研究院 | Image registration method based on self-adaptive parameter particle swarm algorithm |
| CN116341390B (en) * | 2023-05-11 | 2023-11-17 | 西安现代控制技术研究所 | Global search rapid convergence multi-constraint trajectory optimization method |
| CN116562331B (en) * | 2023-05-19 | 2023-11-21 | 石家庄铁道大学 | Method for optimizing SVM by improving reptile search algorithm and application thereof |
| CN116543848B (en) * | 2023-07-05 | 2023-09-29 | 潍坊学院 | Mixture component quantification method based on parallel factor and particle swarm optimization algorithm |
| CN116882279B (en) * | 2023-07-07 | 2024-04-12 | 西南科技大学 | Experiment design optimization method and device for power supply |
| CN117033965A (en) * | 2023-08-11 | 2023-11-10 | 湖北工业大学 | Biological vaccine data characteristic selection method, device, equipment and medium |
| CN116756469B (en) * | 2023-08-22 | 2023-10-31 | 中之力搏建设工程有限公司 | Outdoor lighting lamp optimization management system |
| CN116822567B (en) * | 2023-08-28 | 2023-11-21 | 山东省科学院海洋仪器仪表研究所 | Optimization method for evaporation waveguide prediction model parameters |
| CN117407775A (en) * | 2023-09-15 | 2024-01-16 | 三峡大学 | Nondestructive detection method for loss of foundation bolt and nut of power transmission tower based on INGO-SVM |
| CN117113795B (en) * | 2023-10-23 | 2024-01-26 | 之江实验室 | Method for optimizing parameters of magnetically constrained charged particle imaging system |
| CN117114144B (en) * | 2023-10-24 | 2024-01-26 | 青岛农业大学 | Rice salt and alkali resistance prediction method and system based on artificial intelligence |
| CN117198418B (en) * | 2023-11-07 | 2024-02-13 | 威海百合生物技术股份有限公司 | Polysaccharide extraction process parameter optimization method and system |
| CN117196418B (en) * | 2023-11-08 | 2024-02-02 | 江西师范大学 | Reading teaching quality assessment method and system based on artificial intelligence |
| CN117584136B (en) * | 2024-01-18 | 2024-03-29 | 泰山学院 | Robot fault detection method and system based on artificial intelligence |
| CN117725685A (en) * | 2024-02-05 | 2024-03-19 | 中汽研汽车检验中心（天津）有限公司 | Multi-objective optimization method and equipment for vehicle operability |
| CN117911197A (en) * | 2024-03-20 | 2024-04-19 | 国网江西省电力有限公司电力科学研究院 | Photovoltaic addressing and volume-fixing method and system based on improved multi-target particle swarm algorithm |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US8276106B2 (en) * | 2009-03-05 | 2012-09-25 | International Business Machines Corporation | Swarm intelligence for electrical design space modeling and optimization |
| CN105319071B (en) * | 2015-09-21 | 2017-11-07 | 天津大学 | Diesel Engine Fuel System Fault Diagnosis method based on least square method supporting vector machine |
| CN105159096B (en) * | 2015-10-10 | 2017-08-29 | 北京邮电大学 | A kind of redundancy space manipulator joint moment optimization method based on particle cluster algorithm |
| CN105574231A (en) * | 2015-11-27 | 2016-05-11 | 上海交通大学 | Storage battery surplus capacity detection method |
-
2016
- 2016-10-20 CN CN201610916399.7A patent/CN106682682A/en active Pending
-
2017
- 2017-01-11 WO PCT/CN2017/070894 patent/WO2018072351A1/en active Application Filing
Cited By (44)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107103357A (en) * | 2017-05-23 | 2017-08-29 | 沈阳航空航天大学 | A kind of new dandelion algorithm |
| CN107247844A (en) * | 2017-06-10 | 2017-10-13 | 福州大学 | The minimum tree algorithms of X architecture Steiner based on adaptive PSO and mixing switching strategy |
| CN108363838A (en) * | 2018-01-18 | 2018-08-03 | 上海电力学院 | Temperature effect forecast method in electrostatic precipitator based on ATPSO-SVM models |
| CN108363838B (en) * | 2018-01-18 | 2021-10-08 | 上海电力学院 | Temperature effect prediction method in electrostatic dust collector based on ATPSO-SVM model |
| CN110096927A (en) * | 2018-01-30 | 2019-08-06 | 西安交通大学 | Contactor diagnostic method and diagnostic system based on particle group optimizing support vector machines |
| CN108364030A (en) * | 2018-03-20 | 2018-08-03 | 东北大学 | A kind of multi-categorizer model building method based on three layers of dynamic particles group's algorithm |
| CN108364030B (en) * | 2018-03-20 | 2019-08-20 | 东北大学 | A kind of multi-categorizer model building method based on three layers of dynamic particles group's algorithm |
| CN108615069A (en) * | 2018-03-25 | 2018-10-02 | 哈尔滨工程大学 | A kind of optimized calculation method based on improved adaptable quanta particle swarm optimization |
| CN108539571A (en) * | 2018-04-08 | 2018-09-14 | 上海交通大学 | A kind of fast automatic mode locking method covering multimode pulse recognition |
| CN108594290A (en) * | 2018-05-02 | 2018-09-28 | 成都理工大学 | A kind of spectral line modification method |
| CN108629155A (en) * | 2018-05-14 | 2018-10-09 | 浙江大学 | A kind of leukaemia cancer cell detector that parameter is optimal |
| CN109150873A (en) * | 2018-08-16 | 2019-01-04 | 武汉虹旭信息技术有限责任公司 | Malice domain name detection system and method based on PSO_SVM optimization algorithm |
| CN109739959A (en) * | 2018-11-30 | 2019-05-10 | 东软集团股份有限公司 | Method and device used in being calculated in topic association |
| CN109739959B (en) * | 2018-11-30 | 2021-02-26 | 东软集团股份有限公司 | Method and device used in topic association calculation |
| CN110070458A (en) * | 2019-03-15 | 2019-07-30 | 福建商学院 | The method for manufacturing Dynamic Scheduling |
| CN110390419A (en) * | 2019-05-20 | 2019-10-29 | 重庆大学 | Freeway toll station method for predicting based on PSO-LSSVM model |
| CN110728231A (en) * | 2019-10-10 | 2020-01-24 | 华东理工大学 | Sleep staging method based on improved particle swarm algorithm and twin support vector machine |
| CN110728231B (en) * | 2019-10-10 | 2023-03-28 | 华东理工大学 | Sleep staging method based on improved particle swarm algorithm and twin support vector machine |
| CN111047102A (en) * | 2019-12-18 | 2020-04-21 | 江南大学 | Express delivery distribution route optimization method based on elite-driven particle swarm algorithm |
| CN111210075A (en) * | 2020-01-07 | 2020-05-29 | 国网辽宁省电力有限公司朝阳供电公司 | Lightning stroke transmission line fault probability analysis method based on combined classifier |
| CN111210075B (en) * | 2020-01-07 | 2023-05-12 | 国网辽宁省电力有限公司朝阳供电公司 | Lightning transmission line fault probability analysis method based on combined classifier |
| CN111275078A (en) * | 2020-01-13 | 2020-06-12 | 南京航空航天大学 | Optimization method of support vector machine for part image recognition |
| CN111643321A (en) * | 2020-04-30 | 2020-09-11 | 北京精密机电控制设备研究所 | Exoskeleton joint angle prediction method and system based on sEMG signals |
| CN111681258A (en) * | 2020-06-12 | 2020-09-18 | 上海应用技术大学 | Hybrid enhanced intelligent trajectory prediction method and device based on hybrid wolf optimization SVM |
| CN111709584A (en) * | 2020-06-18 | 2020-09-25 | 中国人民解放军空军研究院战略预警研究所 | Radar networking optimization deployment method based on artificial bee colony algorithm |
| CN111709584B (en) * | 2020-06-18 | 2023-10-31 | 中国人民解放军空军研究院战略预警研究所 | Radar networking optimization deployment method based on artificial bee colony algorithm |
| CN111736618A (en) * | 2020-06-28 | 2020-10-02 | 清华大学 | Unmanned motorcycle steering control parameter setting method and device |
| CN111736618B (en) * | 2020-06-28 | 2021-08-10 | 清华大学 | Unmanned motorcycle steering control parameter setting method and device |
| CN111717217A (en) * | 2020-06-30 | 2020-09-29 | 重庆大学 | Driver intention identification method based on probability correction |
| CN111717217B (en) * | 2020-06-30 | 2022-11-08 | 重庆大学 | Driver intention identification method based on probability correction |
| CN111950604B (en) * | 2020-07-27 | 2024-05-14 | 江苏大学 | Image recognition classification method of multi-classification support vector machine based on minimum reconstruction error search dimension reduction and particle swarm optimization |
| CN111950604A (en) * | 2020-07-27 | 2020-11-17 | 江苏大学 | Image identification and classification method of multi-classification support vector machine based on minimum reconstruction error search reduction and particle swarm optimization |
| CN112365117A (en) * | 2020-09-03 | 2021-02-12 | 中交西安筑路机械有限公司 | Pavement structure performance calculation method based on optimized support vector machine |
| CN112308229A (en) * | 2020-11-26 | 2021-02-02 | 西安邮电大学 | Dynamic multi-objective evolution optimization method based on self-organizing mapping |
| CN112308229B (en) * | 2020-11-26 | 2023-11-24 | 西安邮电大学 | Dynamic multi-objective evolution optimization method based on self-organizing map |
| CN113570555A (en) * | 2021-07-07 | 2021-10-29 | 温州大学 | Two-dimensional segmentation method of multi-threshold medical image based on improved grasshopper algorithm |
| CN113570555B (en) * | 2021-07-07 | 2024-02-09 | 温州大学 | Two-dimensional segmentation method of multi-threshold medical image based on improved grasshopper algorithm |
| CN113759722A (en) * | 2021-09-13 | 2021-12-07 | 桂林电子科技大学 | Parameter optimization method for active disturbance rejection controller of unmanned aerial vehicle |
| CN113759722B (en) * | 2021-09-13 | 2024-03-29 | 桂林电子科技大学 | Unmanned aerial vehicle active disturbance rejection controller parameter optimization method |
| CN115222007A (en) * | 2022-05-31 | 2022-10-21 | 复旦大学 | Improved particle swarm parameter optimization method for glioma multitask integrated network |
| CN115222007B (en) * | 2022-05-31 | 2023-06-20 | 复旦大学 | Improved particle swarm parameter optimization method for colloid rumen multitasking integrated network |
| CN115412671B (en) * | 2022-08-29 | 2023-04-07 | 特斯联科技集团有限公司 | Camera shutter artificial intelligence adjustment method and system for monitoring moving object |
| CN115412671A (en) * | 2022-08-29 | 2022-11-29 | 特斯联科技集团有限公司 | Camera shutter artificial intelligence adjusting method and system for monitoring moving object |
| CN115880572A (en) * | 2022-12-19 | 2023-03-31 | 江苏海洋大学 | Forward-looking sonar target identification method based on asynchronous learning factor |
Also Published As
| Publication number | Publication date |
|---|---|
| WO2018072351A1 (en) | 2018-04-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106682682A (en) | Method for optimizing support vector machine based on Particle Swarm Optimization | |
| Sun et al. | Improved monarch butterfly optimization algorithm based on opposition-based learning and random local perturbation | |
| CN105868775A (en) | Imbalance sample classification method based on PSO (Particle Swarm Optimization) algorithm | |
| CN105930856A (en) | Classification method based on improved DBSCAN-SMOTE algorithm | |
| CN110263673A (en) | Human facial expression recognition method, apparatus, computer equipment and storage medium | |
| CN103886330A (en) | Classification method based on semi-supervised SVM ensemble learning | |
| CN111986811A (en) | Disease prediction system based on big data | |
| De Amorim | Constrained clustering with minkowski weighted k-means | |
| CN110909773B (en) | Client classification method and system based on adaptive particle swarm | |
| WO2020114108A1 (en) | Clustering result interpretation method and device | |
| Kianmehr et al. | Fuzzy clustering-based discretization for gene expression classification | |
| CN103678512A (en) | Data stream merge sorting method under dynamic data environment | |
| Zhang et al. | A Splitting Criteria Based on Similarity in Decision Tree Learning. | |
| CN104091038A (en) | Method for weighting multiple example studying features based on master space classifying criterion | |
| Murty et al. | Automatic clustering using teaching learning based optimization | |
| WO2021189830A1 (en) | Sample data optimization method, apparatus and device, and storage medium | |
| Liu et al. | Possible world based consistency learning model for clustering and classifying uncertain data | |
| Zhou et al. | Region purity-based local feature selection: A multi-objective perspective | |
| CN114817543A (en) | Text clustering method based on contrast learning and dynamic adjustment mechanism | |
| CN110378389A (en) | A kind of Adaboost classifier calculated machine creating device | |
| Li et al. | Classifier subset selection based on classifier representation and clustering ensemble | |
| Tian et al. | Elephant search algorithm on data clustering | |
| Rizk et al. | A local mixture based SVM for an efficient supervised binary classification | |
| Li et al. | Divide-and-conquer ensemble self-training method based on probability difference | |
| CN114821157A (en) | Multi-modal image classification method based on hybrid model network |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20170517 |