CN109635506A - The photovoltaic cell model parameter identification method of adaptive chaos tree and seed algorithm - Google Patents
The photovoltaic cell model parameter identification method of adaptive chaos tree and seed algorithm Download PDFInfo
- Publication number
- CN109635506A CN109635506A CN201910024120.8A CN201910024120A CN109635506A CN 109635506 A CN109635506 A CN 109635506A CN 201910024120 A CN201910024120 A CN 201910024120A CN 109635506 A CN109635506 A CN 109635506A
- Authority
- CN
- China
- Prior art keywords
- tree
- photovoltaic cell
- seed
- parameter
- algorithm
- 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.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/004—Artificial life, i.e. computing arrangements simulating life
- G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Computational Linguistics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Computer Hardware Design (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Geometry (AREA)
- Data Mining & Analysis (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Photovoltaic Devices (AREA)
Abstract
The invention discloses the photovoltaic cell model parameter identification methods of a kind of adaptive chaos tree and seed algorithm.Include the following steps: output voltage and output current data that photovoltaic cell 1) is obtained by execute-in-place or experiment;2) objective function for searching for the mean square deviation for the electric current that the actual current of photovoltaic cell and identification model obtain as adaptive chaos tree and seed algorithm optimizing;3) set algorithm operating parameter;4) it runs adaptive chaos tree and seed algorithm recognizes photovoltaic cell unknown parameter, by minimizing objective function.The identification estimated value of unknown parameter in model is obtained, identified parameters are substituted into kinetic model, forms mathematical model.The modeling method of the invention has the characteristics that realization is simple, low optimization accuracy is high, fast convergence rate, is also applied for the parameter identification of other complex process models.
Description
Technical field
The present invention relates to the photovoltaic cell model parameter identification methods of a kind of adaptive chaos tree and seed algorithm.
Background technique
Energy crisis, environmental pollution, climate change and fossil energy exhaustion problem are the significant challenge of facing mankind, the sun
Hot spot can be had become as a kind of important renewable and clean energy resource, the utilization of solar energy and the research of photovoltaic cell characteristic.
Scholars propose the photovoltaic cell model of different description I-V curves.I-V curve is the statement of photovoltaic cell characteristic, mould
Shape parameter is the reflection of system intrinsic characteristic.Parameter by recognizing photovoltaic cell model can optimize photovoltaic cell system, have
Help design photovoltaic cell and assess the performance of photovoltaic cell, and the variation by analyzing these parameters can also study photovoltaic
The reason of cell malfunctions.Therefore the parameter identification of photovoltaic cell model has for studying and improving solar cell properties
Important realistic meaning.
I-V characteristic equation is that a complexity surmounts nonlinear function, directly can not solve design parameter by simple computation.
The parameter identification method of photovoltaic cell is broadly divided into analytic method and numerical method at present.Analytic method is to utilize mathematical method
By I-V characteristic equation simplification, the analytic value of each parameter is found out by numerical fitting.Although with the method approximate solution of mathematical analysis
The method simple, intuitive of parameter, but its parameter error acquired is larger, is not suitable for when model accuracy is more demanding;Numerical method
Including being based on Deterministic Methods and two kinds of intelligent optimization algorithm, Deterministic Methods such as Newton method, general gradient method etc. are for initial value
It is very sensitive, and can only generally search local optimum.
Method for parameter estimation based on intelligent optimization algorithm mainly joins photovoltaic cell using intelligent optimization algorithm
Number identification.Intelligent optimization algorithm has many advantages, such as not depending on plant characteristic, calculates simple and global search.In recent years, many intelligence
Energy optimization algorithm is used in the parameter identification of photovoltaic cell model, such as particle swarm algorithm, genetic algorithm, differential evolution are calculated
Method, artificial bee colony algorithm etc..There is more significant advantage in precision and reliability based on the parameter Estimation of intelligent optimization algorithm,
But most of intelligent optimization algorithms there is fall into local optimum and with the number of iterations increase search efficiency decline the defects of,
These defects constrain further increasing for parameter identification precision.
Tree and seed algorithm (Tree-Seed algorithm) are a kind of novel heuristic values, have part
The advantages that search capability is strong, fast convergence rate, but that there are ability of searching optimum is weaker, easily falls into office for basic tree and seed algorithm
The disadvantages of portion is optimal, algorithm later period search efficiency declines.Adaptive chaos tree proposed by the present invention and seed algorithm, for basic
Defect existing for algorithm generates initial population using chaotic maps, and according to the characteristics of algorithmic statement process and population at individual is poor
It is different, it realizes that algorithm parameter is adaptive, improves the ability of searching optimum and search efficiency of algorithm.The present invention is by the adaptive of proposition
Chaos tree and seed algorithm are for achieving ideal result in photovoltaic cell identification of Model Parameters.
Summary of the invention
In view of the above-mentioned deficiencies in the prior art, it is an object of the present invention to provide the light of a kind of adaptive chaos tree and seed algorithm
Lie prostrate battery model parameter identification method.
A kind of photovoltaic cell model parameter identification method of adaptive chaos tree and seed algorithm, this method include following step
It is rapid:
Step 1: the output voltage and output current data of photovoltaic cell model are obtained by on-the-spot test or experiment;
Step 2: establishing photovoltaic cell I-V characteristic equation shown in following formula, determine parameter to be identified;
Wherein, VLAnd ILIt is the output voltage and output electricity of the photovoltaic cell obtained by execute-in-place or experiment respectively
Stream, IphIt is photogenerated current, unit A;IsdIt is diode reverse saturation current, unit is μ A;RSIt is photovoltaic cell series connection resistance,
Unit is Ω;RshIt is photovoltaic cell parallel resistance, unit Ω;N is diode quality factor;Q is electron charge, is 1.608
×10-19C;K is Boltzmann constant, is 1.380 × 10-23J/K;T is the absolute temperature of battery, unit K;Ginseng to be identified
Number is Iph Isd RS RshAnd n;
Step 3: the value range of input experimental data and each parameter to be identified is based on adaptive chaos tree for constructing
With the photovoltaic cell model parameter identification method of seed algorithm;
Step 4: the operating parameter of adaptive chaos tree and seed algorithm, including woodlot scale N, maximum number of iterations are set
GmaxWith algorithm termination rules;
Step 5: running adaptive chaos tree and seed algorithm to the photogenerated current I in photovoltaic cell modelph, diode it is anti-
To saturation current Isd, cell series resistance RS, battery parallel resistance RshIt is distinguished with five unknown parameters of diode quality factor n
Know, by minimizing objective function, obtains the estimated value of unknown parameter;
Step 6: the parameter that step 5 identification obtains being substituted into photovoltaic cell I-V characteristic equation, obtains the identification of photovoltaic cell
Model.
Further, the specific steps of the step 5 are as follows:
Step 5-1: utilizing chaotic maps initialization population, generated at random in parameter optimization space 5 it is different initial
It is worth [x01,x02,x03,x04,x05], it carries out n times Tent chaotic maps and generates the initial woodlot location matrix that N row 5 arranges, every a line table
Show the possibility solution of one group of photovoltaic cell I-V model parameter, Tent chaotic maps formula is as follows:
As 0 < a < 1 and 0≤x≤1, system is in chaos state;The initial woodlot location matrix T generated is as follows:
Step 5-2: adaptive transformation is made to the control parameter ST in tree and seed algorithm, adaptive transformation formula is as follows:
ST=0.05+0.45 × exp (- 30 × (G/Gmax)5)
In formula, G is current iteration number, GmaxIt is maximum number of iterations;
Step 5-3: adaptive transformation is made to the seed amount that each tree in tree and seed algorithm generates, adaptive transformation is public
Formula is as follows:
In formula, nsiIt is the seed number that i-th tree generates, seed number ns related with population scale sizeiIn population scale
Between 10% to 25%;Weight WiIt is calculated by following formula:
In formula, FiIt is the fitness value of i-th tree;
Step 5-4: ns is generated around i-th treeiA seed carries out local search, and generates in [0, a 1] section
Random number rand:
If rand < ST, new seed is generated according to the position of optimal tree and random tree, such as following formula:
Sij=Tij+α×L(stepij)×(Bj-Tij), i=1,2 ... N, j=1,2 ... 5
If rand >=ST, new seed is generated according to the position of present tree and random tree, such as following formula:
Sij=Tij+φij×(Trj-Tij), i=1,2 ... N, j=1,2 ... 5
In formula: TijIt is the jth dimension variable of i-th tree position, SijIt is the jth dimension change for the new seed position that i-th tree generates
Amount, BjIt is the jth dimension variable of optimal tree position, TrjIt is the jth dimension variable of a random tree position in woodlot, φijBe -1 to 1 it
Between random number;α is the zoom factor of Lay dimension flight step-length, stepijIt is the Lay dimension flight step-length of i-th tree jth dimension variable,
stepijCalculation formula it is as follows:
In formula, β is setting parameter, 1≤β≤3;L(stepij) be to Lay dimension flight step-length clipping operation, such as following formula:
In formula, lbAnd ubFor the bound of clipping operation;
Step 5-5: the seed that fitness is optimal in seed will be generated and be compared with the fitness of present tree, if more
It is excellent, then the position set originally is substituted, otherwise, the position of present tree remains unchanged;
Step 5-6: updating the state of current woodlot, and algorithm iteration number G increases by 1;Simultaneously Population Regeneration optimal solution is recorded,
In minimum problem, population optimal solution calculates such as following formula:
B=min { Fi, i=1,2 ... N
Step 5-7: step 5-2,5-3,5-4,5-5,5-6 are repeated, until meeting algorithm termination rules;
Step 5-8: using the history optimal solution found as the identified parameters of photovoltaic cell I-V model.
Further, in the step 5-1, a value 0.501.
Further, in the step 5-4, β=1.5, lbAnd ubValue is respectively -5 and 5, and α value is 0.2.
Further, the algorithm termination rules in the step 4 are as follows: algorithm number of run reaches maximum number of iterations Gmax。
Adaptive chaos tree proposed by the present invention and seed algorithm are reflected for defect existing for rudimentary algorithm using chaos
It penetrates generation initial population, and according to the characteristics of algorithmic statement process and population at individual difference, realizes that algorithm parameter is adaptive, improve
The ability of searching optimum and search efficiency of algorithm.Photovoltaic electric proposed by the present invention based on adaptive chaos tree and seed algorithm
Pool model parameter identification method, obtained model can accurately reflect the characteristic of photovoltaic cell model.The modeling method of the invention
Have the characteristics that realize that simple, low optimization accuracy is high, fast convergence rate, is also applied for the parameter identification of other complex process models.
Detailed description of the invention
Fig. 1 is single diode model of photovoltaic cell;
Fig. 2 is the photovoltaic cell model parameter estimation method flow diagram of adaptive chaos tree and seed algorithm;
Fig. 3 is the model output data of photovoltaic cell electric current and the comparison figure of experimental data.
Specific embodiment
The present invention is further elaborated and is illustrated with reference to the accompanying drawings and detailed description.
A kind of photovoltaic cell model parameter identification method of adaptive chaos tree and seed algorithm, includes the following steps:
Step 1: the output voltage and output current data of photovoltaic cell model are obtained by on-the-spot test or experiment;
Step 2: establishing photovoltaic cell I-V characteristic equation shown in following formula, determine parameter to be identified;
Wherein, VLAnd ILIt is the output voltage and output electricity of the photovoltaic cell obtained by execute-in-place or experiment respectively
Stream, IphIt is photogenerated current, unit A;IsdIt is diode reverse saturation current, unit is μ A;RSIt is photovoltaic cell series connection resistance,
Unit is Ω;RshIt is photovoltaic cell parallel resistance, unit Ω;N is diode quality factor;Q is electron charge, is 1.608
×10-19C;K is Boltzmann constant, is 1.380 × 10-23J/K;T is the absolute temperature of battery, unit K;Ginseng to be identified
Number is Iph Isd RS RshAnd n;
Step 3: the value range of input experimental data and each parameter to be identified is based on adaptive chaos tree for constructing
With the photovoltaic cell model parameter identification method of seed algorithm;
Step 4: the operating parameter of adaptive chaos tree and seed algorithm, including woodlot scale N, maximum number of iterations are set
GmaxWith algorithm termination rules.Algorithm termination rules may be configured as: algorithm number of run reaches maximum number of iterations Gmax。
Step 5: running adaptive chaos tree and seed algorithm to the photogenerated current I in photovoltaic cell modelph, diode it is anti-
To saturation current Isd, cell series resistance RS, battery parallel resistance RshIt is distinguished with five unknown parameters of diode quality factor n
Know, by minimizing objective function, obtains the estimated value of unknown parameter;
The specific steps of step 5 are as follows:
Step 5-1: utilizing chaotic maps initialization population, generated at random in parameter optimization space 5 it is different initial
It is worth [x01,x02,x03,x04,x05], it carries out n times Tent chaotic maps and generates the initial woodlot location matrix that N row 5 arranges, every a line table
Show the possibility solution of one group of photovoltaic cell I-V model parameter, Tent chaotic maps formula is as follows:
As 0 < a < 1 and 0≤x≤1, system is in chaos state, and a can be with value for 0.501 herein;What is generated is initial
Woodlot location matrix T is as follows:
Step 5-2: adaptive transformation is made to the control parameter ST in tree and seed algorithm, adaptive transformation formula is as follows:
ST=0.05+0.45 × exp (- 30 × (G/Gmax)5)
In formula, G is current iteration number, GmaxIt is maximum number of iterations, ST parameter is to control population in tree and seed algorithm
The value of the key parameter of mode of evolution, ST is bigger, and algorithm local search ability is stronger, and convergence rate is faster, and the value of ST subtracts
Small, algorithm the convergence speed is slack-off but ability of searching optimum becomes strong;Self-adapted ST parameter reduces as the number of iterations increases, can be with
It keeps the diversity of population early period and improves the convergence precision in later period;
Step 5-3: adaptive transformation is made to the seed amount that each tree in tree and seed algorithm generates, adaptive transformation is public
Formula is as follows:
In formula, nsiIt is the seed number that i-th tree generates, seed number ns related with population scale sizeiIn population scale
Between 10% to 25%;Through experimental analysis, seed number nsiWhen between 10% to the 25% of population scale, the search of algorithm is imitated
Rate is preferable;Adaptive seed number, which sets fitness more preferably, can be generated more seeds, improves Evolution of Population efficiency;Power
Weight WiIt is calculated by following formula:
In formula, FiIt is the fitness value of i-th tree;
Step 5-4: ns is generated around i-th treeiA seed carries out local search, and generates in [0, a 1] section
Random number rand:
If rand < ST, new seed is generated according to the position of optimal tree and random tree, such as following formula:
Sij=Tij+α×L(stepij)×(Bj-Tij), i=1,2 ... N, j=1,2 ... 5
If rand >=ST, new seed is generated according to the position of present tree and random tree, such as following formula:
Sij=Tij+φij×(Trj-Tij), i=1,2 ... N, j=1,2 ... 5
In formula: TijIt is the jth dimension variable of i-th tree position, SijIt is the jth dimension change for the new seed position that i-th tree generates
Amount, BjIt is the jth dimension variable of optimal tree position, TrjIt is the jth dimension variable of a random tree position in woodlot, φijBe -1 to 1 it
Between random number;α is the zoom factor of Lay dimension flight step-length, stepijIt is the Lay dimension flight step-length of i-th tree jth dimension variable,
stepijCalculation formula it is as follows:
In formula, β is setting parameter, and 1≤β≤3, value is β=1.5 herein;L(stepij) it is that flight step-length is tieed up to Lay
Clipping operation, such as following formula:
In formula, lbAnd ubFor the bound of clipping operation;lbAnd ubValue is respectively -5 and 5, and α value is 0.2.
Step 5-5: the seed that fitness is optimal in seed will be generated and be compared with the fitness of present tree, if more
It is excellent, then the position set originally is substituted, otherwise, the position of present tree remains unchanged;
Step 5-6: updating the state of current woodlot, and algorithm iteration number G increases by 1;Simultaneously Population Regeneration optimal solution is recorded,
In minimum problem, population optimal solution calculates such as following formula:
B=min { Fi, i=1,2 ... N
Step 5-7: step 5-2,5-3,5-4,5-5,5-6 are repeated, until meeting algorithm termination rules;
Step 5-8: using the history optimal solution found as the identified parameters of photovoltaic cell I-V model.
Step 6: the parameter that step 5 identification obtains being substituted into photovoltaic cell I-V characteristic equation, obtains the identification of photovoltaic cell
Model.
It is subsequent that photovoltaic cell system optimization is carried out based on the identification model or pre- using the I-V characteristic equation acquired
The output power of side photovoltaic cell.
Below by way of a specific embodiment, present invention is further described in detail:
Embodiment:
The utilization of solar energy and the research of photovoltaic cell characteristic have become hot spot, domestic as research is continuous deeply
Outer scholar proposes the photovoltaic cell model of different description I-V curves.I-V curve be photovoltaic cell characteristic macroscopical description its
In parameter be model intrinsic characteristic reflection.I-V equation can be not only determined by recognizing photovoltaic cell parameter, using acquiring
The pre- side photovoltaic cell of I-V equation output power.Therefore the identification of photovoltaic cell inner parameter is carried out for studying and improving
Its characteristic is significantly.
Accurate photovoltaic cell model is obtained, the output voltage and output electric current pair of photovoltaic cell can be accurately calculated
It should be related to.Photovoltaic cell model is photovoltaic cell list diode model shown in FIG. 1 in the present embodiment.Photovoltaic cell list diode
The I-V characteristic equation of model is shown below:
Wherein, VLAnd ILIt is by execute-in-place or to test the output voltage of the photovoltaic cell obtained and export electric current,
IphIt is photogenerated current (A), IsdIt is diode reverse saturation current (μ A), RSIt is cell series resistance (Ω), RshIt is battery parallel connection
Resistance (Ω), n are diode quality factor, and q is electron charge (1.608 × 10-19C), K be Boltzmann constant (1.380 ×
10-23J/K), T is the absolute temperature (K) of battery.
In a model, Iph Isd RS RshIt is the photovoltaic cell list diode die shape parameter of 5 estimations with n, it can be by sample
Data estimation.
Photovoltaic cell list diode model method for parameter estimation step is as shown in Fig. 2, detailed process is as follows:
Step 1: 26 groups of photovoltaic cell data are determined by experiment, comprising: VL、IL,T.The present embodiment data are from document
Oliva D,El Aziz M A,Hassanien A E.Parameter estimation of photovoltaic cells
using an improved chaotic whale optimization algorithm[J].Applied Energy,
2017,200:141-154。
Step 2: setting optimization object function are as follows:
Wherein, N is sample size, x={ Iph,Isd,RS,Rsh, n },It is the output voltage in c group data
With output electric current,It is given by:
Fitness function when formula (2) is searched for as adaptive chaos tree and seed algorithm optimizing;
Step 3: chaos intialization population: setting population scale N=20, maximum number of iterations Gmax=1000, it is sought in parameter
5 initial value [x for having fine difference are generated in excellent space at random01,x02,x03,x04,x05], it carries out n times Tent chaotic maps and produces
The original chaotic woodlot location matrix that raw N row 5 arranges, every a line indicate the possibility solution of one group of photovoltaic cell I-V model parameter, Tent
Chaotic maps formula is as follows:
As 0 < a < 1,0≤x≤1, system is in chaos state, a value 0.501.The initial population position square of generation
Battle array is as follows:
Step 4: the termination rules of set algorithm are as follows: algorithm number of run reaches maximum number of iterations Gmax;
Step 5: using the position of each of population individual as the I-V characteristic equation of photovoltaic cell list diode model
One group of unknown parameter to be estimated, and optimization object function value corresponding to this group of parameter is calculated by formula (2), as individual
Fitness value.Record and save the state of woodlot optimum individual;
Step 6: the self adaptive control parameter ST of setting tree and seed algorithm, such as following formula:
ST=0.05+0.45 × exp (- 30 × (G/Gmax)5) (6)
In formula (6), G is current iteration number, GmaxIt is maximum number of iterations;
Step 7: the adaptive seed parameter ns of setting tree and seed algorithm, such as following formula:
In formula (7), nsiIt is the seed number W that i-th tree generatesiIt is calculated by following formula:
In formula (8), FiIt is the fitness value of i-th tree.
Step 8: each tree generates seed and carries out local search around tree:
If rand < ST, new seed is generated according to the position of optimal tree and random tree, such as following formula:
Sij=Tij+α×L(stepij)×(Bj-Tij), i=1,2 ... N, j=1,2 ... 5 (9)
If rand >=ST, new seed is generated according to the position of present tree and random tree, such as following formula:
Sij=Tij+φij×(Trj-Tij), i=1,2 ... N, j=1,2 ... 5 (10)
In formula (9), (10): TijIt is the jth dimension variable of i-th tree position, SijIt is the new seed position that i-th tree generates
Jth tie up variable, BjIt is the jth dimension variable of optimal tree position, TrjIt is the jth dimension variable of a random tree position in woodlot, φij
It is the random number between -1 to 1.α is the zoom factor of Lay dimension flight step-length, stepijIt is that i-th Lay dimension for setting jth dimension variable flies
Row step-length, stepijCalculation formula it is as follows:
In formula (11), β (1≤β≤3) is setting parameter, β=1.5.L(stepij) it is to be grasped to the clipping of Lay dimension flight step-length
Make, such as following formula:
In formula (12), lbAnd ubFor the bound of clipping operation, value is respectively -5 and 5, and α value is 0.2.
Step 9: the optimal seed of fitness generated in seed is compared with the fitness for the tree for being currently generated seed,
If more excellent, the position set originally is substituted, otherwise, the position of present tree remains unchanged.
Step 10: updating the state of current population, algorithm iteration number G increases by 1, records simultaneously Population Regeneration optimal solution, kind
Group's optimal solution calculates such as following formula:
B=min { Fi, i=1,2 ... N (13)
Step 11: 6~step 10 of algorithm steps is repeated, until meeting algorithm stop criterion;Most by the history found
Identified parameters of the excellent solution as photovoltaic cell I-V model.
According to the above method, the estimates of parameters for obtaining photovoltaic cell list diode model is following (table 1):
The photovoltaic cell list diode model estimates of parameters of the adaptive chaos tree of table 1 and seed algorithm
Unknown parameter | Iph(A) | Isd(μA) | RS(Ω) | Rsh(Ω) | n |
Estimated value | 0.76078 | 0.3229 | 0.03638 | 53.7120 | 1.4812 |
The parameter identification result for offering estimation with original text compares, as a result following (table 2):
2 distinct methods of table recognize photovoltaic cell I-V equation model parameter comparison result
Unknown parameter | A | Ea/J | a | b | c | RMSE |
Inventive algorithm | 0.76078 | 0.3229 | 0.03638 | 53.7120 | 1.4812 | 9.8602e-4 |
Newton method | 0.7608 | 0.3223 | 0.0364 | 53.7634 | 1.4837 | 9.6960e-3 |
CWOA | 0.76077 | 0.3239 | 0.03636 | 53.7987 | 1.4812 | 9.8604e-4 |
Newton method, the result of CWOA method identified parameters are from Oliva D, El Aziz M A, Hassanien A
E.Parameter estimation of photovoltaic cells using an improved chaotic whale
Optimization algorithm [J] .Applied Energy, 2017,200:141-154, RMSE are that formula (2) defines
Optimization object function.It can be seen that from the comparison result of table 2 for same experimental data, using adaptive chaos tree and kind
The photovoltaic cell I-V model of subalgorithm identified parameters has higher precision.
The parameter value that adaptive chaos tree and seed algorithm are estimated is substituted into photovoltaic cell model, corresponding I-V is obtained
Recognize model.The I-V data that the I-V equation curve and experiment for recognizing model obtain are as shown in Figure 3.The results show that the present invention mentions
The photovoltaic cell model parameter identification method based on adaptive chaos tree and seed algorithm out, obtained model can be accurately
Reflect the characteristic of photovoltaic cell model.
Claims (5)
1. a kind of photovoltaic cell model parameter identification method of adaptive chaos tree and seed algorithm, which is characterized in that including such as
Lower step:
Step 1: the output voltage and output current data of photovoltaic cell model are obtained by on-the-spot test or experiment;
Step 2: establishing photovoltaic cell I-V characteristic equation shown in following formula, determine parameter to be identified;
Wherein, VLAnd ILIt is the output voltage and output electric current of the photovoltaic cell obtained by execute-in-place or experiment, I respectivelyph
It is photogenerated current, unit A;IsdIt is diode reverse saturation current, unit is μ A;RSIt is photovoltaic cell series connection resistance, unit
For Ω;RshIt is photovoltaic cell parallel resistance, unit Ω;N is diode quality factor;Q is electron charge, is 1.608 × 10-19C;K is Boltzmann constant, is 1.380 × 10-23J/K;T is the absolute temperature of battery, unit K;Parameter to be identified is
Iph Isd RS RshAnd n;
Step 3: the value range of input experimental data and each parameter to be identified, for constructing based on adaptive chaos tree and kind
The photovoltaic cell model parameter identification method of subalgorithm;
Step 4: the operating parameter of adaptive chaos tree and seed algorithm, including woodlot scale N, maximum number of iterations G are setmax
With algorithm termination rules;
Step 5: running adaptive chaos tree and seed algorithm to the photogenerated current I in photovoltaic cell modelph, diode reversely satisfies
With electric current Isd, cell series resistance RS, battery parallel resistance RshIt is recognized with five unknown parameters of diode quality factor n,
By minimizing objective function, the estimated value of unknown parameter is obtained;
Step 6: the parameter that step 5 identification obtains being substituted into photovoltaic cell I-V characteristic equation, obtains the identification mould of photovoltaic cell
Type.
2. the photovoltaic cell model parameter identification method of adaptive chaos tree and seed algorithm according to claim 1,
It is characterized in that the specific steps of the step 5 are as follows:
Step 5-1: chaotic maps initialization population is utilized, generates 5 different initial values at random in parameter optimization space
[x01,x02,x03,x04,x05], it carries out n times Tent chaotic maps and generates the initial woodlot location matrix that N row 5 arranges, every a line indicates
The possibility solution of one group of photovoltaic cell I-V model parameter, Tent chaotic maps formula are as follows:
As 0 < a < 1 and 0≤x≤1, system is in chaos state;The initial woodlot location matrix T generated is as follows:
Step 5-2: adaptive transformation is made to the control parameter ST in tree and seed algorithm, adaptive transformation formula is as follows:
ST=0.05+0.45 × exp (- 30 × (G/Gmax)5)
In formula, G is current iteration number, GmaxIt is maximum number of iterations;
Step 5-3: the seed amount generated to each tree in tree and seed algorithm makees adaptive transformation, and adaptive transformation formula is such as
Under:
In formula, nsiIt is the seed number that i-th tree generates, seed number ns related with population scale sizeiThe 10% of population scale
To between 25%;Weight WiIt is calculated by following formula:
In formula, FiIt is the fitness value of i-th tree;
Step 5-4: ns is generated around i-th treeiA seed carries out local search, and generates random in [0, a 1] section
Number rand:
If rand < ST, new seed is generated according to the position of optimal tree and random tree, such as following formula:
Sij=Tij+α×L(stepij)×(Bj-Tij), i=1,2 ... N, j=1,2 ... 5
If rand >=ST, new seed is generated according to the position of present tree and random tree, such as following formula:
Sij=Tij+φij×(Trj-Tij), i=1,2 ... N, j=1,2 ... 5
In formula: TijIt is the jth dimension variable of i-th tree position, SijIt is the jth dimension variable for the new seed position that i-th tree generates, Bj
It is the jth dimension variable of optimal tree position, TrjIt is the jth dimension variable of a random tree position in woodlot, φijIt is between -1 to 1
Random number;α is the zoom factor of Lay dimension flight step-length, stepijIt is the Lay dimension flight step-length of i-th tree jth dimension variable, stepij
Calculation formula it is as follows:
In formula, β is setting parameter, 1≤β≤3;L(stepij) be to Lay dimension flight step-length clipping operation, such as following formula:
In formula, lbAnd ubFor the bound of clipping operation;
Step 5-5: will generate the seed that fitness is optimal in seed and be compared with the fitness of present tree, if more excellent,
The position set originally is substituted, otherwise, the position of present tree remains unchanged;
Step 5-6: updating the state of current woodlot, and algorithm iteration number G increases by 1;Simultaneously Population Regeneration optimal solution is recorded, minimum
In value problem, population optimal solution calculates such as following formula:
B=min { Fi, i=1,2 ... N
Step 5-7: step 5-2,5-3,5-4,5-5,5-6 are repeated, until meeting algorithm termination rules;
Step 5-8: using the history optimal solution found as the identified parameters of photovoltaic cell I-V model.
3. the photovoltaic cell identification of Model Parameters side of a kind of adaptive chaos tree and seed algorithm according to claim 2
Method, which is characterized in that in the step 5-1, a value 0.501.
4. the photovoltaic cell identification of Model Parameters side of a kind of adaptive chaos tree and seed algorithm according to claim 2
Method, which is characterized in that in the step 5-4, β=1.5, lbAnd ubValue is respectively -5 and 5, and α value is 0.2.
5. the photovoltaic cell identification of Model Parameters side of a kind of adaptive chaos tree and seed algorithm according to claim 1
Method, which is characterized in that the algorithm termination rules in the step 4 are as follows: algorithm number of run reaches maximum number of iterations Gmax。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910024120.8A CN109635506B (en) | 2019-01-10 | 2019-01-10 | Photovoltaic cell model parameter identification method of self-adaptive chaotic tree and seed algorithm |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910024120.8A CN109635506B (en) | 2019-01-10 | 2019-01-10 | Photovoltaic cell model parameter identification method of self-adaptive chaotic tree and seed algorithm |
Publications (2)
Publication Number | Publication Date |
---|---|
CN109635506A true CN109635506A (en) | 2019-04-16 |
CN109635506B CN109635506B (en) | 2021-04-13 |
Family
ID=66060465
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201910024120.8A Active CN109635506B (en) | 2019-01-10 | 2019-01-10 | Photovoltaic cell model parameter identification method of self-adaptive chaotic tree and seed algorithm |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109635506B (en) |
Cited By (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110287540A (en) * | 2019-05-29 | 2019-09-27 | 江苏大学 | A kind of photovoltaic cell parameter identification method based on elite masses' differential evolution algorithm |
CN110829491A (en) * | 2019-10-25 | 2020-02-21 | 国网甘肃省电力公司电力科学研究院 | Grid-connected photovoltaic power generation system parameter identification method based on transient disturbance |
CN110929464A (en) * | 2019-11-20 | 2020-03-27 | 燕山大学 | Storage battery parameter identification method based on improved dragonfly algorithm |
CN111191375A (en) * | 2020-01-04 | 2020-05-22 | 温州大学 | Photovoltaic cell parameter identification method based on improved Harris eagle optimization algorithm |
CN111625998A (en) * | 2020-05-29 | 2020-09-04 | 华中科技大学 | Method for optimizing structure of laminated solar cell |
CN111814399A (en) * | 2020-07-08 | 2020-10-23 | 温州大学 | Model parameter optimization extraction method and measurement data prediction method for solar photovoltaic cell system |
CN111859796A (en) * | 2020-07-14 | 2020-10-30 | 温州大学 | Harris eagle photovoltaic model parameter optimization method based on longitudinal and transverse intersection and NM type |
CN112084703A (en) * | 2020-08-18 | 2020-12-15 | 温州大学 | Photovoltaic cell system model parameter identification method based on variant shuffling frog leaping algorithm |
CN112507613A (en) * | 2020-12-01 | 2021-03-16 | 湖南工程学院 | Second-level ultra-short-term photovoltaic power prediction method |
CN112526243A (en) * | 2020-12-28 | 2021-03-19 | 南京航空航天大学 | Sectional type direct current arc noise model, parameter optimization and identification method |
CN113868996A (en) * | 2021-10-19 | 2021-12-31 | 青岛科技大学 | Photovoltaic solar model parameter estimation method based on hierarchical Newton identification algorithm |
WO2022105060A1 (en) * | 2020-11-20 | 2022-05-27 | 温州大学 | Method for identifying parameters of photovoltaic cell having single diode structure |
CN115392139A (en) * | 2022-10-27 | 2022-11-25 | 西北工业大学 | Fuel cell parameter identification method |
CN117117976A (en) * | 2023-10-25 | 2023-11-24 | 广东电网有限责任公司中山供电局 | Double-fed induction wind driven generator parameter identification method and device |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106202914A (en) * | 2016-07-07 | 2016-12-07 | 国网青海省电力公司 | Based on the photovoltaic cell parameter identification method improving particle cluster algorithm |
CN106447024A (en) * | 2016-08-31 | 2017-02-22 | 上海电机学院 | Particle swarm improved algorithm based on chaotic backward learning |
CN107103154A (en) * | 2017-05-17 | 2017-08-29 | 南京南瑞继保电气有限公司 | A kind of photovoltaic module model parameter identification method |
CN107563489A (en) * | 2017-07-25 | 2018-01-09 | 华南理工大学 | Photovoltaic array powerinjected method method under local shades based on Chaos particle swarm optimization algorithm |
-
2019
- 2019-01-10 CN CN201910024120.8A patent/CN109635506B/en active Active
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106202914A (en) * | 2016-07-07 | 2016-12-07 | 国网青海省电力公司 | Based on the photovoltaic cell parameter identification method improving particle cluster algorithm |
CN106447024A (en) * | 2016-08-31 | 2017-02-22 | 上海电机学院 | Particle swarm improved algorithm based on chaotic backward learning |
CN107103154A (en) * | 2017-05-17 | 2017-08-29 | 南京南瑞继保电气有限公司 | A kind of photovoltaic module model parameter identification method |
CN107563489A (en) * | 2017-07-25 | 2018-01-09 | 华南理工大学 | Photovoltaic array powerinjected method method under local shades based on Chaos particle swarm optimization algorithm |
Non-Patent Citations (1)
Title |
---|
MUSTAFA SERVET KIRAN: "Tree-seed algorithm for continuous optimization", 《EXPERT SYSTEMS WITH APPLICATIONS》 * |
Cited By (24)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110287540A (en) * | 2019-05-29 | 2019-09-27 | 江苏大学 | A kind of photovoltaic cell parameter identification method based on elite masses' differential evolution algorithm |
CN110287540B (en) * | 2019-05-29 | 2023-06-13 | 江苏大学 | Photovoltaic cell parameter identification method based on elite crowd differential evolution algorithm |
CN110829491A (en) * | 2019-10-25 | 2020-02-21 | 国网甘肃省电力公司电力科学研究院 | Grid-connected photovoltaic power generation system parameter identification method based on transient disturbance |
CN110829491B (en) * | 2019-10-25 | 2023-08-22 | 国网甘肃省电力公司电力科学研究院 | Grid-connected photovoltaic power generation system parameter identification method based on transient disturbance |
CN110929464B (en) * | 2019-11-20 | 2022-06-03 | 燕山大学 | Storage battery parameter identification method based on improved dragonfly algorithm |
CN110929464A (en) * | 2019-11-20 | 2020-03-27 | 燕山大学 | Storage battery parameter identification method based on improved dragonfly algorithm |
CN111191375A (en) * | 2020-01-04 | 2020-05-22 | 温州大学 | Photovoltaic cell parameter identification method based on improved Harris eagle optimization algorithm |
CN111191375B (en) * | 2020-01-04 | 2023-06-02 | 温州大学 | Photovoltaic cell parameter identification method based on improved Harris eagle optimization algorithm |
CN111625998A (en) * | 2020-05-29 | 2020-09-04 | 华中科技大学 | Method for optimizing structure of laminated solar cell |
CN111814399A (en) * | 2020-07-08 | 2020-10-23 | 温州大学 | Model parameter optimization extraction method and measurement data prediction method for solar photovoltaic cell system |
CN111814399B (en) * | 2020-07-08 | 2024-03-19 | 温州大学 | Model parameter optimization extraction method and measurement data prediction method of solar photovoltaic cell system |
CN111859796A (en) * | 2020-07-14 | 2020-10-30 | 温州大学 | Harris eagle photovoltaic model parameter optimization method based on longitudinal and transverse intersection and NM type |
CN112084703B (en) * | 2020-08-18 | 2023-08-25 | 温州大学 | Method for identifying parameters of photovoltaic cell system model based on variant shuffling frog-leaping algorithm |
CN112084703A (en) * | 2020-08-18 | 2020-12-15 | 温州大学 | Photovoltaic cell system model parameter identification method based on variant shuffling frog leaping algorithm |
WO2022105060A1 (en) * | 2020-11-20 | 2022-05-27 | 温州大学 | Method for identifying parameters of photovoltaic cell having single diode structure |
CN112507613B (en) * | 2020-12-01 | 2022-05-10 | 湖南工程学院 | Second-level ultra-short-term photovoltaic power prediction method |
CN112507613A (en) * | 2020-12-01 | 2021-03-16 | 湖南工程学院 | Second-level ultra-short-term photovoltaic power prediction method |
CN112526243B (en) * | 2020-12-28 | 2021-10-26 | 南京航空航天大学 | Sectional type direct current arc noise model, parameter optimization and identification method |
CN112526243A (en) * | 2020-12-28 | 2021-03-19 | 南京航空航天大学 | Sectional type direct current arc noise model, parameter optimization and identification method |
CN113868996A (en) * | 2021-10-19 | 2021-12-31 | 青岛科技大学 | Photovoltaic solar model parameter estimation method based on hierarchical Newton identification algorithm |
CN113868996B (en) * | 2021-10-19 | 2024-05-07 | 青岛科技大学 | Photovoltaic solar model parameter estimation method based on hierarchical Newton identification algorithm |
CN115392139A (en) * | 2022-10-27 | 2022-11-25 | 西北工业大学 | Fuel cell parameter identification method |
CN117117976A (en) * | 2023-10-25 | 2023-11-24 | 广东电网有限责任公司中山供电局 | Double-fed induction wind driven generator parameter identification method and device |
CN117117976B (en) * | 2023-10-25 | 2024-03-05 | 广东电网有限责任公司中山供电局 | Double-fed induction wind driven generator parameter identification method and device |
Also Published As
Publication number | Publication date |
---|---|
CN109635506B (en) | 2021-04-13 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109635506A (en) | The photovoltaic cell model parameter identification method of adaptive chaos tree and seed algorithm | |
CN109873610B (en) | Photovoltaic array fault diagnosis method based on IV characteristic and depth residual error network | |
CN110597240B (en) | Hydroelectric generating set fault diagnosis method based on deep learning | |
CN106485075B (en) | Photovoltaic model parameter identification method based on eagle strategy and self-adaptive NM simplex | |
CN104951834A (en) | LSSVM (least squares support vector machine) wind speed forecasting method based on integration of GA (genetic algorithm) and PSO (particle swarm optimization) | |
CN106126863B (en) | Photovoltaic cell parameter identification method based on artificial fish-swarm and the algorithm that leapfrogs | |
CN112380765A (en) | Photovoltaic cell parameter identification method based on improved balance optimizer algorithm | |
CN110334870B (en) | Photovoltaic power station short-term power prediction method based on gated cyclic unit network | |
CN105574615A (en) | Spatial correlation and genetic algorithm (GA) based wind power forecast method for wavelet-BP neural network | |
CN110929451A (en) | Fuel cell single voltage consistency prediction method | |
CN111859796A (en) | Harris eagle photovoltaic model parameter optimization method based on longitudinal and transverse intersection and NM type | |
CN116840720A (en) | Fuel cell remaining life prediction method | |
CN110852017A (en) | Hydrogen fuel cell fault diagnosis method based on particle swarm optimization and supporting vector machine | |
CN109961173A (en) | A kind of intelligent Forecasting for distributed generation resource generated output | |
Liao et al. | An improved differential evolution to extract photovoltaic cell parameters | |
CN113177673A (en) | Air conditioner cold load prediction optimization method, system and equipment | |
CN109543269B (en) | Photovoltaic model parameter identification method based on mixed trust interval reflection | |
CN115995810A (en) | Wind power prediction method and system considering weather fluctuation self-adaptive matching | |
CN106786499B (en) | Based on the short-term wind power forecast method for improving AFSA optimization ELM | |
CN106100582B (en) | Based on the recursive least-squares photovoltaic cell parameter identification method with forgetting factor | |
CN117374941A (en) | Photovoltaic power generation power prediction method based on neural network | |
CN113595132A (en) | Photovoltaic online parameter identification method based on global maximum power point tracking and hybrid optimization algorithm | |
CN117540638A (en) | Equivalent modeling method, medium and system for photovoltaic power station | |
CN112836876A (en) | Power distribution network line load prediction method based on deep learning | |
CN117272184A (en) | ISSA-RF algorithm-based photovoltaic power generation system fault classification method |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |