CN112883632B - Lithium battery equivalent circuit model parameter identification method based on improved ant colony algorithm - Google Patents
Lithium battery equivalent circuit model parameter identification method based on improved ant colony algorithm Download PDFInfo
- Publication number
- CN112883632B CN112883632B CN202110044991.3A CN202110044991A CN112883632B CN 112883632 B CN112883632 B CN 112883632B CN 202110044991 A CN202110044991 A CN 202110044991A CN 112883632 B CN112883632 B CN 112883632B
- Authority
- CN
- China
- Prior art keywords
- path
- terminal voltage
- lithium battery
- ant colony
- colony 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.)
- Active
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F30/00—Computer-aided design [CAD]
- G06F30/20—Design optimisation, verification or simulation
- G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/3644—Constructional arrangements
- G01R31/3648—Constructional arrangements comprising digital calculation means, e.g. for performing an algorithm
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01R—MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
- G01R31/00—Arrangements for testing electric properties; Arrangements for locating electric faults; Arrangements for electrical testing characterised by what is being tested not provided for elsewhere
- G01R31/36—Arrangements for testing, measuring or monitoring the electrical condition of accumulators or electric batteries, e.g. capacity or state of charge [SoC]
- G01R31/367—Software therefor, e.g. for battery testing using modelling or look-up tables
-
- 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
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/12—Computing arrangements based on biological models using genetic models
- G06N3/126—Evolutionary algorithms, e.g. genetic algorithms or genetic programming
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F2111/00—Details relating to CAD techniques
- G06F2111/06—Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- General Engineering & Computer Science (AREA)
- Computational Linguistics (AREA)
- Molecular Biology (AREA)
- Mathematical Physics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computing Systems (AREA)
- Evolutionary Biology (AREA)
- General Health & Medical Sciences (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Genetics & Genomics (AREA)
- Physiology (AREA)
- Geometry (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Charge And Discharge Circuits For Batteries Or The Like (AREA)
Abstract
The invention discloses a lithium battery equivalent circuit model parameter identification method based on an improved ant colony algorithm, and belongs to the field of electrical engineering. The method comprises the following steps: building a lithium battery equivalent circuit model; an improved ant colony algorithm is introduced for parameter identification, and a reliable ant colony algorithm bi-pheromone concentration calculation index and a path selection method for effectively preventing local optimal solution are provided; and finally, verifying the precision of the parameter identification result through a simulation experiment. The invention solves the problem of low identification precision of the lithium battery equivalent circuit model parameters, and provides the lithium battery equivalent model parameter identification method with high practicability, effectiveness and applicability.
Description
Technical Field
The invention relates to a parameter identification category of a lithium battery, in particular to a lithium battery equivalent circuit model parameter identification method based on an improved ant colony algorithm.
Background
The environmental protection and energy safety problem pushes new energy automobiles, particularly electric automobiles, to the stage of the era, and becomes the focus of attention of all people, and one of the main technical bottlenecks of energy storage element related technology as electric automobiles is a hot problem to be solved at present, and the energy storage element related technology is widely concerned in the world. The lithium battery in the energy storage element becomes a great hot point for research due to the advantages of long cycle service life, light weight, high energy density, no pollution, high cost performance and the like. At present, lithium battery models are mainly divided into electrochemical models, artificial neural network models and equivalent circuit models. Electrochemical models to accurately describe the characteristics of a battery, a large number of parameters are used to simulate the polarization phenomenon of the battery. The calculation amount for building the electrochemical model is large, the process is complex, the method has great limitation, and the method is not suitable for practical engineering application. The artificial neural network model is trained based on a large amount of experimental data to further obtain the relationship between the parameters of the battery model. The method has the disadvantages that a large amount of experimental data is needed to predict the performance of the battery, and the method has high dependence on the historical data of the sample battery. The equivalent circuit model is the most commonly used method in modeling, simulation and engineering practical application, and the common equivalent circuit models include a PNGV model, a second-order RC equivalent model and a third-order RC equivalent model. The most widely studied and applied is the second order RC equivalent model. The accuracy of the third-order RC equivalent model is higher than that of the second-order RC equivalent model, but more parameters need to be identified, and the calculation is too complex. The fractional order model of the lithium battery has higher model precision than a second order RC equivalent model, the calculated amount is smaller than that of the third order RC equivalent model, but the number of parameters to be identified is still large, and on the premise that a more ideal parameter identification value range is not known, the calculated amount of parameter identification of the several lithium battery equivalent circuit models is very large, and the identification precision is difficult to guarantee.
The ant colony algorithm is used as a bionic algorithm to simulate the foraging process of ants in the nature. The ants carry out pheromone communication in the continuous foraging process, and finally find out the shortest path between the food location and the nest. The ant colony algorithm has the main advantages of better robustness and stronger applicability. However, the traditional ant colony algorithm has limited capability of avoiding the local optimal solution, and the algorithm is easy to fall into the local optimal solution without improvement and lose the global optimal solution.
To sum up, at present, the lithium battery equivalent circuit model parameter identification has the following technical problems:
1. on the premise of not knowing a more ideal parameter identification value range, the calculation amount of parameter identification of the lithium battery equivalent circuit model is very large, and the identification precision is difficult to ensure.
2. If the parameter identification is carried out on the lithium battery equivalent circuit model by adopting the traditional ant colony algorithm, the lithium battery equivalent circuit model is easy to fall into a local optimal solution without improvement, and a global optimal solution is lost.
Disclosure of Invention
Aiming at the defects of the prior art, the invention provides a parameter identification method of a lithium battery equivalent circuit model based on an improved ant colony algorithm in order to overcome the technical defect that the parameter identification precision of the existing lithium battery equivalent circuit model is not high.
The technical means adopted by the invention are as follows:
a lithium battery equivalent model parameter identification method based on an improved ant colony algorithm is characterized in that the following logic operation modules are stored in a computer:
a parameter identification module;
a module of a value range optimization method;
a module for calculating an index;
a module of a path selection method;
and (4) performing operation processing on the pulse discharge experiment data of the lithium battery through the modules in sequence, and finally identifying to obtain parameters of the equivalent model of the lithium battery.
In the invention, except for hardware needed when obtaining the pulse discharge experiment data, the other parts of the invention are realized by writing an m function program in an MATLAB environment on a PC/computer and constructing the function/operation module; all modules are completed in the program. Inputting pulse discharge experimental data and ohmic internal resistance R0 of a lithium battery pack of an NR18650-30Q lithium battery pack with the parameters of the single rated capacity of 3AH and the rated voltage of 3.6V, and directly inputting the measured data into a program; the method comprises the steps that pulse discharge experiment data of the lithium battery are obtained through a high-performance battery monitoring system (hardware), the pulse discharge experiment of the lithium battery is that the lithium battery which is fully charged discharges electricity for 20 times by constant current discharge at intervals, the specific result obtained by the experiment is 20 terminal voltage specific values corresponding to the 20 times of lithium battery discharge, and the 20 terminal voltage specific values are input into a program. Ohmic internal resistance R0 is a known value and is directly input into the program; finally identifying to obtain parameters of the lithium battery equivalent model, namely calculating polarization capacitors C1 and C2, fractional orders m and n of the polarization capacitors C1 and C2 and polarization resistors R1 and R2 of the lithium battery fractional order equivalent model through a software program; among the above modules, the parameter identification module is S101 to S108, the value range optimization method module is S102, the index calculation module is S103 and S104, and the path selection method module is S106.
Further comprising: the parameter identification module is used for identifying parameters of the lithium battery equivalent model by adopting an improved ant colony algorithm; the module of the value range optimization method is a method module for optimizing the value range of the parameter to be identified based on the ant colony algorithm; the index calculating module is a calculation module based on the concentration of the double pheromones; the module of the path selection method aims to improve the search of the global optimal solution.
Further: introducing an improved ant colony algorithm to perform parameter identification on the lithium battery equivalent model, and specifically comprising the following steps:
s101, initializing operation; namely, initializing the software program stored in the computer;
s102, setting the value range of each parameter to be identified of the improved ant colony algorithm according to the calculation result of the adaptive genetic algorithm; the improved ant colony algorithm is the software of the invention and is also called as an m-function program; parameters for improving the ant colony algorithm comprise the number X of ant colonies, the number G of reproduction generations, the number P of paths, the univariate short code length M1, the univariate long code length M2, the variable dimension of the short code length to be optimized is N1, the variable dimension of the long code length is N2, the city number C = M1N1+ M2N2, lithium battery pulse discharge experiment data and a set objective function; the ant colony algorithm simulates the process of ants searching for food in biology, the path refers to the corresponding number of the solution obtained by program calculation, the corresponding path obtained by the first ant calculation is 1, the corresponding path obtained by the second ant calculation is 2, and so on.
S103, calculating pheromone concentration of each path by taking the sum of squared terminal voltage errors as an improved ant colony algorithm calculation index, and obtaining a corresponding value V of a local optimal path J (i) (ii) a The path described in this step is set by the program: the path corresponding to the solution calculated by the first ant is 1, the path corresponding to the solution calculated by the second ant is 2, and so on.
S103 and S104 are two different pheromone concentration calculation formulas. Both represent two different pheromone concentration calculation methods. Can be expressed in terms of a first pheromone concentration, a second pheromone;
the sum of the squares of the terminal voltage errors is an evaluation index of the terminal voltage errors, and a formula is expressed as
Wherein i represents the ith pulse discharge, n is the upper limit of the pulse discharge number and is 20,V Sm (i) A terminal voltage model value V of the ith pulse discharge of the m path Tm (i) The true value of terminal voltage V during the ith pulse discharge of the m path Jm (i) The square sum of the terminal voltage errors corresponding to the ith pulse discharge of the m path is obtained; the pheromone concentration is a calculation index of the solution in the ant colony algorithm, the lower the pheromone concentration in S103 is, the better the solution is, and the higher the pheromone concentration in S104 is, the better the solution is. The optimal path in the S103 refers to a path corresponding to a solution that minimizes the sum of squared terminal voltage errors; the optimal path corresponding value in S103 refers to a solution that minimizes the sum of squared terminal voltage errors;
s104, changing the pheromone concentration calculation index for improving the ant colony algorithm into the difference between the maximum terminal voltage error square sum and the terminal voltage square sum; the maximum terminal voltage error sum of squares refers to the maximum value of the sum of squares of the difference between the real terminal voltage and the model terminal voltage.
S105, calculating the transition probability of each path by taking the pheromone concentration calculation index in the step S104 as a reference; the pheromone concentration calculation index is a formula for calculating pheromone concentration, and since the formula for calculating pheromone concentration in S103 is different from that in S104, the distinction mark is such.
S106, selecting a part of ants to select paths according to the pheromone concentration of S103, setting another part of ants to randomly select paths, and obtaining a corresponding value V of a global optimal path K (i) (ii) a One part to the other part are according to
And 1/4X, wherein X refers to the number of ants, X is 200 in the example,
is a number of 150 (wt.),
is 50; v J (i) For locally optimal path correspondence value, V K (i) And the global optimal path corresponds to a value.
S107, calculating V K (i)-V J (i)：
If the result is positive, V is received J (i) A current solution as a solution;
if the result is negative, then V is received k (i) A current solution as a solution;
s108, if the specified algorithm termination condition is met, outputting the current optimal solution as a final output result and finishing iteration;
otherwise, the process returns to step S103.
Preferably, in S108, the termination condition means: the maximum value of the terminal voltage error does not exceed 0.01 or the number of reproduction generations exceeds 50;
further, in step S102, a method for setting a value range of each parameter to be identified: and performing parameter identification on the lithium battery equivalent circuit model for multiple times by adopting a self-adaptive genetic algorithm, and optimizing and improving the value range of each parameter to be identified of the ant colony algorithm according to a parameter identification result with higher precision.
The adaptive genetic algorithm and the improved ant colony algorithm provided by the invention belong to intelligent algorithms, but the adaptive genetic algorithm belongs to a simpler intelligent algorithm, a formula is more complex, and a large number of papers have detailed descriptions on the algorithm, so that the details are not repeated here. The self-adaptive genetic algorithm does not belong to the innovation point of the invention, and the parameters comprise an initial population number X, a reproduction algebra G, a hybridization probability initial value K1, a variation probability initial value K2, a single-variable short code length M1, a single-variable long code length M2, a short code length variable dimension to be optimized N1, a long code length variable dimension N2, ohmic internal resistance R0, lithium battery pulse discharge experimental data and a set target function; the higher precision in the step means that the maximum error of the terminal voltage does not exceed 0.08 when the adaptive genetic algorithm is used for parameter identification of the embodiment of the invention. Theoretically, the error is not more than 5% (in the embodiment of the invention, the maximum error of the terminal voltage is 0.21 when the error is 5%), and the accuracy is higher.
Further, in each of steps S103 and S104, there is one pheromone concentration, i.e., a bimesogenic concentration calculation index.
Further, in step S103, the pheromone density of each path is calculated by using the sum of squared terminal voltage errors as an algorithm calculation index, and the formula is as follows:
wherein i represents the ith pulse discharge, n is the upper limit of the pulse discharge number and is 20,V Sm (i) Is terminal voltage model value V of m-path ith pulse discharge Tm (i) The true value of terminal voltage V during the ith pulse discharge of the m path Jm (i) The square sum of terminal voltage errors corresponding to the ith pulse discharge of the m path is smaller, the concentration is higher, the probability that ants leave the path is higher, the path is better, and V J (i) Corresponding values for the optimal path; the pheromone concentration calculation method is visual and effective and is used for calculating the optimal path.
Additionally, the path is given/specified by a human/write programmer; the real terminal voltage value of the corresponding number is obtained through the explained pulse discharge experiment and is input into the m function program; and the terminal voltage model value is obtained by the m function program according to the corresponding solution of the previous optimal path.
Further, in step S104, the pheromone density of each path is calculated by the difference between the maximum sum of squared terminal voltage errors and the sum of squared terminal voltage errors, and the formula is:
wherein, V S For terminal voltage model value at each moment, V T For the actual value of the terminal voltage at each moment, V Hm (i) And calculating the path transfer probability by using the difference between the maximum terminal voltage error square sum and the terminal voltage square sum corresponding to the ith pulse discharge of the m path, wherein the larger the value is, the higher the concentration is, and the higher the probability that ants leave the path is. Path transition probability is the probability of an ant selecting a path, formula
Comprises the following steps:
wherein R is m (i) Is the path transition probability of m at the time of the ith pulse discharge, V Hm (i) The difference between the maximum sum of the squares of the terminal voltage errors and the sum of the squares of the terminal voltage during the ith pulse discharge when the ant walks along the path m, V H The sum of the squares of the maximum terminal voltage errors and the difference between the squares of the terminal voltages at each pulse.
If the square sum of the end voltage errors in step S103 is directly used as an algorithm calculation index to calculate the pheromone concentration formula of each path to calculate the transition probability, the path transition probability is smaller when the pheromone concentration is larger.
After calculating the pheromone density of each route by converting the difference between the maximum terminal voltage error sum of squares and the terminal voltage sum of squares in step S104, the route transition probability increases as the pheromone density increases.
Further, in step S107, a method for selecting a path that avoids a locally optimal solution:
selecting
Each ant selects a path according to the concentration of S103 pheromone and sets
An ant randomly selects a path, and X can be 200 in general; here, X refers to the total number of ants, which is 200 in the present example,
the number of the ants is 150,
50 ants;
the ants are divided into two groups according to the proportion,
the ants are large groups of ants, and the ant is a large group of ants,
each ant is a small group; the large population selects a path according to the pheromone concentration of S103, and the small population randomly selects a path. The two groups calculate the optimal path corresponding value V at this time by the pheromone concentration calculation index of step S103 K (i) The method not only ensures that the algorithm has the capability of searching the global optimal solution, but also reduces the calculated amount as much as possible and meets the actual requirements of engineering.
To better illustrate the invention, we now turn an angle to end again as follows:
a lithium battery equivalent circuit model parameter identification method based on an improved ant colony algorithm solves the parameter identification problem of a lithium battery equivalent circuit model through the improved ant colony algorithm, is written and finished in a Matlab environment, and is characterized by comprising the following steps:
s101, initializing operation;
setting initial parameters:
the method comprises the steps of setting the number of ant populations as X, the number of reproduction generations as G, the number of paths as P, the univariate short code length as M1, the univariate long code length as M2, setting the variable dimension of the short code length to be optimized as N1, setting the variable dimension of the long code length as N2, setting the city number as C = M1N1+ M2N2, importing pulse discharge experimental data of a single rated capacity of 3AH and rated voltage of 3.6V for a NR18650-30Q lithium battery pack lithium battery, importing ohmic internal resistance R0, and setting a target function of a program.
The parameters to be identified are polarization resistors R1 and R2, polarization capacitors C1 and C2, and fractional orders m and n of C1 and C2. The M1 univariate short code length is the 2-system code length of 4 variables of R1, R2, M and n, and M1=6 in the embodiment of the invention; the M2 univariate long code length is a 2-system code length of 2 variables, i.e., C1 and C2, where M2=12 in the embodiment of the present invention; n1 and N2 are short and long coding length variable dimensions, respectively, N1=4 and N2=2 in the present embodiment; c is the number of cities, namely the total 2-system coding length of all parameters to be identified, in the embodiment of the invention, C = M1N1+ M2N2=12+12=24 is pulse discharge experimental data and ohmic internal resistance R0 of a lithium battery pack of an NR18650-30Q lithium battery with the monomer rated capacity of 3AH and the rated voltage of 3.6V, and the measured data are directly input into a program; the method comprises the steps that pulse discharge experiment data of the lithium battery are obtained through a high-performance battery monitoring system (hardware), the pulse discharge experiment of the lithium battery is that the lithium battery which is fully charged discharges electricity for 20 times at intervals of constant current, the specific result obtained through the experiment is 20 terminal voltage specific values corresponding to 20 times of lithium battery discharge, and the 20 terminal voltage specific values are input into a program. The ohmic internal resistance R0 is a known value and can be directly input into a program.
S102, setting the value range of each parameter to be identified of the improved ant colony algorithm according to the calculation result of the adaptive genetic algorithm.
The method comprises the following steps of performing multiple parameter identification (multiple times refer to 10 times) on a lithium battery equivalent circuit model by adopting a self-adaptive genetic algorithm, setting a value interval of each parameter to be identified of an improved ant colony algorithm in a reasonable range on the basis of multiple parameter identification results (each parameter to be identified refers to polarization capacitors C1 and C2 of a lithium battery fractional order equivalent model, fractional order orders m and n of the polarization capacitors C1 and C2, and polarization resistors R1 and R2), and specifically comprises the following steps:
s103, calculating the pheromone concentration of each path by taking the sum of squared terminal voltage errors as an algorithm calculation index to obtain the corresponding value V of the optimal path J (i)；
The pheromone concentration calculation formula is as follows:
the optimal path correspondence value expression is as follows:
s104, changing the algorithm pheromone concentration calculation index into the difference between the maximum terminal voltage error square sum and the terminal voltage square sum;
the pheromone concentration calculation formula is changed to:
if the square sum of the end voltage errors in step S103 is directly used as an algorithm calculation index to calculate the pheromone concentration formula of each path to calculate the transition probability, the path transition probability is smaller when the pheromone concentration is larger.
After calculating the pheromone density of each route by converting the difference between the maximum terminal voltage error sum of squares and the terminal voltage sum of squares in step S104, the route transition probability increases as the pheromone density increases.
S105, calculating the transition probability of each path by taking the pheromone concentration calculation index in the step S4 as a reference;
the transition probability formula is:
s106, selecting a part of ants to calculate indexes according to the pheromone concentration of S3 to select paths, setting another part of ants to randomly select paths, and obtaining the corresponding value V of the optimal path at the moment K (i)；
S107, calculating V K (i)-V J (i) If the result is positive, then V is received J (i) A current solution as a solution; if the result is negative, V is received K (i) As the current solution of the solution.
And S108, if the specified algorithm termination condition is met, outputting the current optimal solution as a final output result and ending the iteration, otherwise, repeating S3 to S7.
The termination condition is that the maximum value of the terminal voltage error does not exceed 0.01 or the number of reproduction generations exceeds 50.
The beneficial effects of the invention are as follows:
1: the method is novel, has strong universality and can be used for parameter identification of all lithium battery equivalent circuit models.
2: the parameter identification result of the invention has high precision and strong applicability.
Drawings
FIG. 1 is a diagram of a fractional order model of a lithium battery according to the present invention.
Fig. 2 is a flow chart of the lithium battery equivalent model parameter identification based on the improved ant colony algorithm of the present invention.
Fig. 3 is a circular current waveform of PI parameters optimized based on ant colony algorithm.
FIG. 4 is a circular current waveform of PI parameter optimized based on ant colony simulated annealing algorithm in the present invention.
Detailed Description
Taking a more complex lithium battery fractional order model in a lithium battery equivalent model as an example, the parameter identification is carried out on the lithium battery fractional order model by adopting an improved ant colony algorithm.
Fig. 1 is a second-order RC equivalent model of a lithium battery, a fractional order is introduced to form a fractional order model of the lithium battery, C1 and C2 are expressed in a fractional order form, and the expression is as follows:
wherein j is an imaginary unit, w is an angular velocity, Z1 and Z2 represent Z transformation forms of C1 and C2, respectively, and m are fractional orders of C1 and C2, respectively
Fig. 2 is a flowchart of an improved ant colony algorithm, and a method for identifying parameters of a fractional order lithium battery model based on the improved ant colony algorithm includes: the following steps:
the technical means adopted by the invention are as follows:
s1, initializing operation;
setting initial parameters:
in this embodiment, the number of ant populations is 100, the number of generations of reproduction is 50, the number of paths is 100, the length of a univariate short code is 6, the length of a univariate long code is 12, the variable dimension of the short code to be optimized is 4, the variable dimension of the long code is 2, the number of cities is C =24+24=48, the ohmic internal resistance R0=0.0024, and lithium battery pulse discharge experimental data is imported, wherein the data is pulse discharge data of an NR18650-30Q lithium battery pack with a single rated capacity of 3AH and a rated voltage of 3.6V. The objective function of the program is set to be that the maximum value of the terminal voltage error does not exceed 0.01. Ohmic internal resistance R0 is R0 in FIG. 1 of the present invention. The single rated capacity refers to the maximum capacity value of a single lithium battery when the single lithium battery is fully charged when the single lithium battery leaves a factory.
The rated voltage refers to a maximum terminal voltage value when the lithium battery is fully charged when leaving a factory.
And S2, setting the value range of each parameter to be identified of the improved ant colony algorithm according to the calculation result of the adaptive genetic algorithm. The method comprises the following steps of performing 10 times of parameter identification on a lithium battery fractional order model by adopting a self-adaptive genetic algorithm, and setting the value range of each parameter to be identified of the improved ant colony algorithm by taking five times of better parameter identification results, wherein the method specifically comprises the following steps of:
the lower limit of R1 is 0.0020, and the upper limit is 0.0036; the lower limit of R2 is 0.0020, and the upper limit is 0.0036; c1 lower limit 20000 and upper limit 500000; c2 lower limit of 50000 and upper limit of 1500000; the fractional order m and n have a lower limit of 0.80 and an upper limit of 0.99. The six parameters are polarization capacitors C1 and C2, fractional orders m and n of the polarization capacitors C1 and C2, and polarization resistors R1 and R2 in fig. 1, respectively. The upper limit and the lower limit of the parameters are artificially set maximum values and minimum values, a specific value interval can be set, and different value intervals of the parameters are obtained by identifying the parameters of the example through a 10-time adaptive genetic algorithm.
S3, calculating the pheromone concentration of each path by taking the sum of squared terminal voltage errors as an algorithm calculation index to obtain the corresponding value V of the optimal path J (i)；
S4, changing the algorithm pheromone concentration calculation index into the difference between the maximum terminal voltage error square sum and the terminal voltage square sum;
s5, calculating the transition probability of each path by taking the pheromone concentration calculation index in the step S4 as a reference;
s6, selecting 75 ants to select paths according to the S3 pheromone concentration calculation indexes, setting 25 ants to randomly select paths, and obtaining the optimal path corresponding value V k (i)；
S7, calculating V k (i)-V J (i) If the result is positive, then V is received J (i) A current solution as a solution; if the result is negative, V is received k (i) As the current solution of the solution.
And S8, if the maximum value of the terminal voltage error does not exceed 0.01 or the number of the reproduction generations exceeds 50, outputting the current optimal solution as a final output result and ending iteration, otherwise, repeating S3 to S7.
The experimental results of this example are shown in fig. 3, fig. 4, table 1 and table 2. The accuracy of the parameter identification result of the method is verified by two indexes, namely the maximum error and the average error of the terminal voltage. The smaller the two errors, the higher the identification accuracy. The results of the parameter identification of the lithium battery fractional order model by adopting the improved ant colony algorithm are shown in fig. 3, fig. 4, table 1 and table 2.
TABLE 1
| R1(Ω) | R2(Ω) | C1(F) | C2(F) | m | n |
| 0.0023 | 0.0031 | 24120 | 650480 | 0.8853 | 0.8032 |
TABLE 2
| Maximum error of terminal voltage | Average error of terminal voltage | Unit of | |
| Improved ant colony algorithm | 0.0613 | 0.0053 | V |
Table 1 shows specific parameter identification values, and fig. 3, fig. 4 and table 2 show that the method has high parameter identification precision, the maximum error of the terminal voltage and the average error of the terminal voltage are both limited within 0.065, and the method is a novel and effective novel method for identifying parameters of an equivalent model of a lithium battery.
Claims (2)
1. A lithium battery equivalent circuit model parameter identification method based on an improved ant colony algorithm is characterized in that the following logic operation modules are stored in a computer:
a module for parameter identification;
a module of a value range optimization method;
a module for calculating an index;
a module of a path selection method;
the pulse discharge experiment data of the lithium battery are sequentially subjected to operation processing of the modules, and finally the parameters of the equivalent model of the lithium battery are identified and obtained;
the parameter identification module is used for identifying parameters of the lithium battery equivalent model by adopting an improved ant colony algorithm; the method specifically comprises the following steps:
s101, initializing operation;
s102, setting the value range of each parameter to be identified of the improved ant colony algorithm according to the calculation result of the adaptive genetic algorithm;
s103, squaring terminal voltage errorAnd calculating the pheromone concentration of each path as the calculation index of the improved ant colony algorithm to obtain the corresponding value V of the local optimal path J (i) (ii) a The formula is as follows:
wherein i represents the ith pulse discharge, n is the upper limit of the pulse discharge frequency and is 20,V Sm (i) A terminal voltage model value V of the ith pulse discharge of the m path Tm (i) The true value of terminal voltage V during the ith pulse discharge of the m path Jm (i) The square sum of terminal voltage errors corresponding to the ith pulse discharge of the m path is smaller, the concentration is higher, the probability that ants leave the path is higher, the path is better, and V J (i) Corresponding values for the optimal path;
s104, changing the pheromone concentration calculation index of the improved ant colony algorithm into the difference between the maximum terminal voltage error square sum and the terminal voltage square sum; the formula is as follows:
wherein, V S For terminal voltage model value at each moment, V T For the real value of the terminal voltage at each moment, V Hm (i) The difference between the maximum terminal voltage error sum of squares and the terminal voltage sum of squares corresponding to the ith pulse discharge of the m path is calculated, the larger the value is, the higher the concentration is, the higher the probability that ants walk the path is, and the path transfer probability is calculated; the path transfer probability is the probability that an ant selects a certain path, and the formula is as follows:
wherein R is m (i) Is the path transition probability of m at the time of the ith pulse discharge, V Hm (i) The maximum end voltage error of the ith pulse discharge when the ant walks the path mDifference between sum of squared difference and sum of squared terminal voltage, V H The sum of the square sum of the maximum terminal voltage error and the difference of the square sum of the terminal voltage at each pulse is obtained;
s105, calculating the transition probability of each path by taking the pheromone concentration calculation index in the step S104 as a reference;
s106, selecting a part of ants to select paths according to the pheromone concentration of S103, setting another part of ants to randomly select paths, and obtaining a corresponding value V of a global optimal path K (i) (ii) a One part is in accordance with the other part
And
dividing, wherein X refers to the number of ants;
s107, calculating V K (i)-V J (i)：
If the result is positive, then V is received J (i) A current solution as a solution;
if the result is negative, then V is received k (i) A current solution as a solution;
s108, if the specified algorithm termination condition is met, outputting the current optimal solution as a final output result and ending the iteration,
otherwise, the process returns to step S103.
2. The method for identifying parameters of an equivalent circuit model of a lithium battery based on an improved ant colony algorithm as claimed in claim 1,
the module of the value range optimization method is a method module for optimizing the value range of the parameter to be identified based on the ant colony algorithm;
the index calculating module is a calculation module based on the concentration of the double pheromones;
the module of the path selection method aims to improve the search of the global optimal solution.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110044991.3A CN112883632B (en) | 2021-01-13 | 2021-01-13 | Lithium battery equivalent circuit model parameter identification method based on improved ant colony algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110044991.3A CN112883632B (en) | 2021-01-13 | 2021-01-13 | Lithium battery equivalent circuit model parameter identification method based on improved ant colony algorithm |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112883632A CN112883632A (en) | 2021-06-01 |
| CN112883632B true CN112883632B (en) | 2022-12-02 |
Family
ID=76045806
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110044991.3A Active CN112883632B (en) | 2021-01-13 | 2021-01-13 | Lithium battery equivalent circuit model parameter identification method based on improved ant colony algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112883632B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113671378B (en) * | 2021-07-12 | 2022-04-08 | 南通大学 | Fractional order theory-based lithium ion battery modeling and parameter identification method |
| CN115102200B (en) * | 2022-07-19 | 2023-04-25 | 贵州大学 | Energy storage system model identification method and system |
| CN116819341A (en) * | 2023-07-04 | 2023-09-29 | 上海玫克生储能科技有限公司 | Lithium battery acceleration parameter identification method and system and electronic equipment |
| CN118133874B (en) * | 2024-01-19 | 2024-10-15 | 深圳极数充物联技术有限公司 | Energy storage battery state detection system based on predation model |
Family Cites Families (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US9536192B2 (en) * | 2014-06-23 | 2017-01-03 | International Business Machines Corporation | Solving vehicle routing problems using evolutionary computing techniques |
| CN105389623A (en) * | 2015-10-22 | 2016-03-09 | 东南大学 | Thermal process model parameter identification method adopting improved ant colony algorithm |
| CN107677960B (en) * | 2017-09-21 | 2020-01-10 | 国网福建省电力有限公司 | Synchronous generator parameter identification method based on PMU measurement |
| CN108776612A (en) * | 2018-04-11 | 2018-11-09 | 深圳大学 | A kind of cloud computing method for allocating tasks, device, equipment and storage medium |
| 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 |
| CN110909858B (en) * | 2019-11-27 | 2023-06-02 | 安徽大学 | Batch scheduling method based on double pheromone ant colony algorithm |
| CN110932274A (en) * | 2019-12-18 | 2020-03-27 | 辽宁工业大学 | Power system measurement and load parameter analysis and identification method |
| CN111914981B (en) * | 2020-05-31 | 2022-11-08 | 吉林大学 | Improved PI model identification method based on particle swarm optimization-ant swarm optimization parallel cross algorithm |
-
2021
- 2021-01-13 CN CN202110044991.3A patent/CN112883632B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN112883632A (en) | 2021-06-01 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112883632B (en) | Lithium battery equivalent circuit model parameter identification method based on improved ant colony algorithm | |
| CN110568359B (en) | Lithium battery residual life prediction method | |
| Gill et al. | Training back propagation neural networks with genetic algorithm for weather forecasting | |
| CN107909220A (en) | Electric heating load prediction method | |
| CN111861013B (en) | Power load prediction method and device | |
| CN113702843B (en) | Lithium battery parameter identification and SOC estimation method based on suburb optimization algorithm | |
| CN108960491A (en) | Method for forecasting photovoltaic power generation quantity based on RBF neural | |
| CN116449218B (en) | Lithium battery health state estimation method | |
| CN109215344B (en) | Method and system for urban road short-time traffic flow prediction | |
| CN116774088A (en) | Lithium ion battery health state estimation method based on multi-objective optimization | |
| CN106327357A (en) | Load identification method based on improved probabilistic neural network | |
| CN114707292B (en) | Analysis method for voltage stability of distribution network containing electric automobile | |
| CN113743538A (en) | Intelligent building energy consumption prediction method, equipment and medium based on IPSO-BP neural network | |
| CN113591299B (en) | Method and system for judging comprehensive performance of different types of micro-channel radiators | |
| CN110929464A (en) | Storage battery parameter identification method based on improved dragonfly algorithm | |
| CN105608295A (en) | Multi-objective evolutionary algorithm (MOEA) and radial basis function (RBF) neural network optimization modeling method of coking furnace pressure | |
| CN115356635B (en) | Identification method for lithium battery equivalent circuit model parameters | |
| CN115542165A (en) | Method for predicting health state of lithium ion battery | |
| CN114781277A (en) | Supercapacitor residual service life prediction method based on LSTM-SVR | |
| CN112990776B (en) | Distribution network equipment health degree evaluation method | |
| CN112098869B (en) | Self-adaptive electric vehicle SOC estimation method based on big data | |
| CN117150334B (en) | Lithium battery multi-condition prediction method and device based on optimized BiLSTM neural network | |
| CN113762591A (en) | Short-term electric quantity prediction method and system based on GRU and multi-core SVM counterstudy | |
| Xu et al. | Short-term electricity consumption forecasting method for residential users based on cluster classification and backpropagation neural network | |
| CN116702598A (en) | Training method, device, equipment and storage medium for building achievement prediction model |
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 |