CN112069736A - Quasi-electrostatic field coupling communication model optimization method based on improved immune algorithm - Google Patents
Quasi-electrostatic field coupling communication model optimization method based on improved immune algorithm Download PDFInfo
- Publication number
- CN112069736A CN112069736A CN202010951835.0A CN202010951835A CN112069736A CN 112069736 A CN112069736 A CN 112069736A CN 202010951835 A CN202010951835 A CN 202010951835A CN 112069736 A CN112069736 A CN 112069736A
- Authority
- CN
- China
- Prior art keywords
- population
- model
- quasi
- affinity
- antibodies
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000004422 calculation algorithm Methods 0.000 title claims abstract description 54
- 230000008878 coupling Effects 0.000 title claims abstract description 51
- 238000010168 coupling process Methods 0.000 title claims abstract description 51
- 238000005859 coupling reaction Methods 0.000 title claims abstract description 51
- 238000004891 communication Methods 0.000 title claims abstract description 44
- 238000000034 method Methods 0.000 title claims abstract description 43
- 238000005457 optimization Methods 0.000 title claims abstract description 21
- 230000004044 response Effects 0.000 claims abstract description 27
- 230000005684 electric field Effects 0.000 claims abstract description 21
- 238000012546 transfer Methods 0.000 claims abstract description 14
- 239000003990 capacitor Substances 0.000 claims abstract description 11
- 239000004020 conductor Substances 0.000 claims abstract description 8
- 230000005540 biological transmission Effects 0.000 claims abstract description 7
- 238000002474 experimental method Methods 0.000 claims abstract description 5
- 238000012360 testing method Methods 0.000 claims abstract description 5
- 238000010367 cloning Methods 0.000 claims description 24
- 239000002184 metal Substances 0.000 claims description 22
- 229910052751 metal Inorganic materials 0.000 claims description 22
- 239000000427 antigen Substances 0.000 claims description 13
- 108091007433 antigens Proteins 0.000 claims description 13
- 102000036639 antigens Human genes 0.000 claims description 13
- 230000008569 process Effects 0.000 claims description 10
- 108090000623 proteins and genes Proteins 0.000 claims description 10
- 238000001514 detection method Methods 0.000 claims description 7
- 238000004088 simulation Methods 0.000 claims description 5
- 230000002068 genetic effect Effects 0.000 claims description 4
- 230000003044 adaptive effect Effects 0.000 claims description 3
- 230000008859 change Effects 0.000 claims description 3
- 239000013589 supplement Substances 0.000 claims description 3
- 230000006872 improvement Effects 0.000 claims 3
- 238000006243 chemical reaction Methods 0.000 description 3
- 230000007547 defect Effects 0.000 description 3
- 230000000694 effects Effects 0.000 description 3
- 238000005516 engineering process Methods 0.000 description 3
- 230000035772 mutation Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 238000004364 calculation method Methods 0.000 description 2
- 230000003053 immunization Effects 0.000 description 2
- 241000257303 Hymenoptera Species 0.000 description 1
- ATJFFYVFTNAWJD-UHFFFAOYSA-N Tin Chemical compound [Sn] ATJFFYVFTNAWJD-UHFFFAOYSA-N 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 230000005686 electrostatic field Effects 0.000 description 1
- 238000005562 fading Methods 0.000 description 1
- 238000011478 gradient descent method Methods 0.000 description 1
- 230000008076 immune mechanism Effects 0.000 description 1
- 238000002649 immunization Methods 0.000 description 1
- 230000006698 induction Effects 0.000 description 1
- 238000002955 isolation Methods 0.000 description 1
- 238000012804 iterative process Methods 0.000 description 1
- 238000012067 mathematical method Methods 0.000 description 1
- 150000002739 metals Chemical class 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000002245 particle Substances 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 230000000717 retained effect Effects 0.000 description 1
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
- 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
- G06F2111/08—Probabilistic or stochastic CAD
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Evolutionary Computation (AREA)
- Biophysics (AREA)
- Evolutionary Biology (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- Physiology (AREA)
- Computational Linguistics (AREA)
- Mathematical Physics (AREA)
- General Health & Medical Sciences (AREA)
- Genetics & Genomics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Medical Informatics (AREA)
- Computer Hardware Design (AREA)
- Geometry (AREA)
- Measurement Of Resistance Or Impedance (AREA)
Abstract
The invention discloses a quasi-electrostatic field coupling communication model optimization method based on an improved immune algorithm, and belongs to the technical field of quasi-electrostatic field capacitive coupling communication. Firstly, representing the coupling relation and the conduction current loss between conductors in quasi-electrostatic field transmission by using equivalent capacitance and equivalent resistance respectively, and establishing an electric field coupling equivalent model; then analyzing the series-parallel connection relation between the resistor and the capacitor, and establishing a corresponding equivalent circuit model; deducing a transfer function between the input voltage and the output voltage according to the equivalent circuit model; testing the frequency response characteristic of the channel through experiments; taking the mean square error of the frequency response of the equivalent circuit and the frequency response of the experiment as a target function, and adopting an improved immune algorithm to carry out optimal parameter estimation; and finally, obtaining a complete equivalent circuit model through the obtained optimized parameter values. The invention effectively improves the convergence rate and the operation efficiency of the algorithm on the premise of ensuring to solve the optimized parameter value, and has good universality and expandability.
Description
Technical Field
The invention belongs to the technical field of quasi-electrostatic field capacitive coupling communication, and relates to a quasi-electrostatic field coupling communication model optimization method based on an improved immune algorithm.
Background
The quasi-electrostatic field capacitive coupling-based wireless communication is a wireless communication technology which takes electric conductors such as human bodies or metals and the like as channels between transmitting and receiving nodes and transmits data through electrostatic field coupling, is widely applied to the field of body communication such as medical treatment, entertainment and the like at present, can realize low-power-consumption communication, and has good confidentiality. The quasi-electrostatic field capacitive coupling communication technology is applied to wireless data transmission in a closed metal body (such as an automobile body, a space vehicle cabin body and the like), and can realize the strong fading non-line-of-sight transmission, communication under the condition of dense multipath effect of a narrow compact closed metal space and complete metal isolation cross-metal body communication which cannot be realized by the traditional wireless communication technology (such as radio frequency communication, optical communication and the like). However, the model optimization method of the quasi-electrostatic field capacitive coupling communication channel is not mature, and the development thereof faces many problems and needs to be further researched.
At present, the modeling of the quasi-electrostatic field capacitive coupling communication channel mainly includes the following methods: finite element model, statistical channel model, equivalent circuit model and distributed circuit model. The equivalent circuit model is a method which is widely applied at present, wherein the equivalent circuit model is characterized by respectively using equivalent capacitance and equivalent resistance to represent electric field coupling and current loss between objects from the circuit perspective, and establishing an equivalent circuit according to the electric field coupling relation. In the equivalent circuit model, parameters such as resistance and capacitance are usually estimated by adopting a general curve fitting or function optimization algorithm, such as a least square method, a newton method, a gradient descent method and other traditional mathematical methods. However, in an actual quasi-electrostatic field capacitive coupling communication model, as many as dozens or even more parameter values need to be estimated, the parameters need to perform optimization solution on two objective functions of amplitude-frequency and phase-frequency response characteristics of a transfer function, and a general curve fitting or function optimization algorithm is not enough to solve the complex multi-objective and multi-parameter solutions and is easy to converge to a local optimal solution to cause model parameter estimation errors. In recent years, population-based intelligent optimization algorithms are widely applied to solving complex practical engineering problems, such as genetic algorithms, particle swarm algorithms, artificial bee colony algorithms, artificial immune algorithms and the like, but all the algorithms have respective defects, such as slow convergence speed, low operation efficiency, convergence to local optimum along with the reduction of population evolutionary diversity and the like.
In order to overcome the defects, the invention adopts an improved artificial immune algorithm and applies the improved artificial immune algorithm to the estimation problem of the resistance and capacitance parameters of the equivalent circuit model of the capacitance coupling communication channel of the quasi-electrostatic field. In the process of solving the practical problems, the immune algorithm simulates a biological immune mechanism, combines the evolutionary principle of genes and tries to selectively and purposefully utilize some characteristic information or knowledge in the problem to be solved to inhibit the degradation phenomenon in the optimization process. The most obvious advantage is that in the iterative process, on the premise of keeping the optimal individual of the previous generation, the global optimal solution can be converged by the probability 1.
Retrieved by the applicant: the model optimization method of the quasi-electrostatic field capacitive coupling communication channel based on the improved immune algorithm optimal parameter estimation is not disclosed in published publications at home and abroad.
Disclosure of Invention
The invention aims to overcome the defects of the prior art and provide a quasi-electrostatic field capacitive coupling communication channel model optimization method combining improved immune algorithm parameter estimation and an equivalent circuit model in order to solve the problems of model parameter estimation error caused by local convergence, low convergence speed, low operation efficiency and the like.
The method is realized by the following technical scheme:
a quasi-electrostatic field capacitive coupling communication channel model optimization method based on improved immune algorithm optimal parameter estimation is basically implemented as follows:
the method comprises the following steps that firstly, the coupling relation among conductors in the transmission process of a quasi-electrostatic field is represented by equivalent capacitance, the loss of conduction current is represented by equivalent resistance, and an electric field coupling equivalent model is established;
step two, analyzing the series-parallel relation between each resistor and each capacitor according to the electric field coupling equivalent model obtained in the step one, and establishing a corresponding equivalent circuit model;
step three, deducing a transfer function between the input voltage and the output voltage according to the equivalent circuit model obtained in the step two;
step four, building a cross-metal body communication model, and testing the frequency response characteristic of a channel through experiments;
and fifthly, carrying out optimal parameter estimation by adopting an improved immune algorithm to obtain the numerical values of each resistor and each capacitor in the model.
The improved immune algorithm has the following characteristics: (1) the cloning probability and the cloning scale related to the affinity and the concentration are adopted on cloning, so that the excellent degree and diversity of the offspring population are ensured; (2) in order to improve convergence speed and solving precision, a method of combining selective cross and adaptive Gaussian variation is adopted to improve global search capability, and meanwhile, cross and variation probability is related to affinity, so that the randomness of traditional cross and variation is avoided, and errors are reduced; (3) similarity detection is carried out on the population after cross variation, antibodies with extremely high similarity and relatively low affinity are removed, and the diversity of the population is ensured; (4) and the artificial bee colony algorithm is combined, and the population after each generation of updating is further searched by adopting the artificial bee colony algorithm, so that the global search and the local search are combined, and the convergence speed is improved.
The method comprises the following specific steps:
step 401: antigen recognition, namely, adopting a minimum mean square error criterion, taking the mean square error of frequency response characteristics (including amplitude-frequency response characteristics and phase-frequency response characteristics) actually measured by a model and the frequency response characteristics of an equivalent circuit model as an objective function, and taking the two objective functions and a constraint condition as an antigen intrusion system;
step 402: antibody coding, namely coding resistance and capacitance parameters needing to be solved as antibodies, and generating the antibodies with the length L ═ m + n according to the difference of the number of the parameters, wherein the resistance value is coded as R1、R2……RmThe capacitance value is coded as C1、C2……Cn；
Step 403: initializing, randomly generating m in a value rangeN initial values of resistance and N × N initial values of capacitance are generated to generate an initial parent population A (p) with a population size N1,p2,……，pN) For the jth antibody pj＝(a1j,a2j,……，aLj)；
Step 404: calculating the affinity of each antibody and antigen in the genetic algebra k and A (k) of the current parent population, namely the fitness of a target function, selecting the minimum mean square error as an optimal individual, terminating the operation and outputting a result if an ending condition (finding an optimal solution or reaching the maximum iteration number) is met, and otherwise continuing the next step;
step 405: calculating the concentration of each antibody in the current population, and evaluating the excellent degree of the antibodies by integrating the affinity and the concentration to obtain the expected reproduction probability;
step 406: selecting the first i antibodies as a population A1 according to the expected reproduction probability, selecting the first h antibodies according to the affinity to replace the h antibodies with lower affinity in the excellent gene library, and cloning the selected i antibodies according to the cloning probability and the cloning scale related to the affinity and the concentration to form a temporary population B;
step 407: crossing and mutating individuals in the cloned population B to change the affinity of the individuals, so that the overall search of the population is realized;
step 408: carrying out similarity detection on the variant population B, removing the antibodies with extremely high similarity and relatively low affinity, and then retaining the first j antibodies with high affinity to form a population B1;
step 409: randomly generating r new antibodies as a supplement population C1, forming a new population A (k +1) with the antibody populations A1 and B1, improving the population diversity and keeping the population size to be N;
step 410: the new population is further searched by adopting an artificial bee colony algorithm, so that the platform period in the population evolution process is effectively reduced, and the convergence speed is improved;
step 411: repeating the steps 404-410.
And step six, substituting the optimized resistance and capacitance parameter values obtained in the step five into a circuit to obtain a complete equivalent circuit model, and substituting the parameter values into the transfer function in the step three to obtain a transfer function of a communication channel, so that a model of the quasi-electrostatic field capacitance coupling communication channel is established.
Advantageous effects
Compared with the prior art, the method disclosed by the invention has the advantages that the global search and the local search are combined, the global optimal solution of parameters such as resistance and capacitance can be obtained by the probability 1, the convergence speed, the solving precision and the operation efficiency are effectively improved, the accuracy of the quasi-electrostatic field capacitance coupling communication channel model is ensured, and the effects of high convergence speed, small error and high efficiency are achieved.
Drawings
FIG. 1 is a general flow diagram of an embodiment of the present invention;
FIG. 2 is an electric field coupling equivalent model of quasi-electrostatic field capacitive coupling communication in a closed metal body;
FIG. 3 is an equivalent circuit model of the capacitance coupling communication of the quasi-electrostatic field in the closed metal body;
FIG. 4 is an amplitude-frequency response curve of an experimental test channel;
FIG. 5 is a flow chart of an estimation of optimal parameters for an immunization algorithm;
FIG. 6 is a graph of an evolutionary curve of an estimation process of an optimal parameter of an immune algorithm;
fig. 7 is a graph of a fit of the amplitude frequency response.
Detailed Description
The following describes in detail embodiments of the method of the present invention with reference to the accompanying drawings.
As shown in fig. 1, a model optimization method for a quasi-electrostatic field capacitive coupling communication channel based on improved immune algorithm parameter estimation specifically includes the steps of:
the method comprises the steps of firstly, representing the coupling relation among conductors in the transmission process of the quasi-electrostatic field by using equivalent capacitance, representing the loss of conduction current by using equivalent resistance, and establishing an electric field coupling equivalent model.
In this step, the coupling electric field is generated by alternating signals applied to the positive and negative ends of the transmitting electrode, for example, by ANSYS using a sealed metal body isolated in the middleThe rough mutual capacitance values between the conductors obtained by Maxwell electric field simulation determine the negligible capacitance and the non-negligible capacitance, so that: the induced electric field of the sending end is coupled to the forward channel and the backward channel and is also coupled to the metal body; the induction electric field of the receiving end is coupled to the receiving electrode and the metal body; coupling electric fields also exist between the receiving electrode and the metal body as well as between the front channel and the back channel; in addition, the introduced resistance characterizes the conduction current losses between the channel and the metal body and between the electrode and the metal body, in view of the direct contact between the channel and the metal body through the insulating layer and the close coupling between the electrode and the metal body. An electric field coupling equivalent model established according to the distribution characteristics of the electric field in the cabin is shown in FIG. 2, T+And T-Positive and negative terminals, R, of the transmitting electrode, respectively+And R-Respectively the positive and negative ends of a receiving electrode, H represents a metal cabin body, a forward channel is marked by F, a backward channel is marked by B, and the metal cabin body comprises 17 capacitance parameters and 7 resistance parameters.
And step two, analyzing the series-parallel relation between each resistor and each capacitor according to the electric field coupling equivalent model obtained in the step one, and establishing a corresponding equivalent circuit model.
The series-parallel relation of the resistance and the capacitance of the model is arranged, the equivalent circuit model obtained by conversion is shown in figure 3, the capacitance value or the resistance value between each object uses the identification of two related objects as subscripts, for example, the capacitance between the positive end of the transmitting electrode plate and the forward channel is represented as Ct1fA transmission voltage is applied to the capacitor Ct1t2Across a receiving voltage in a capacitor Cr1r2And detecting the two ends.
And step three, deducing a transfer function between the input voltage and the output voltage according to the equivalent circuit model obtained in the step two.
In the step, an equivalent circuit model obtains the functional relation between input voltage and output voltage about frequency, various resistance parameters and capacitance parameters, in a complex frequency domain model of the circuit, a capacitor C is changed into 1/Cs after Laplace conversion, a resistor R is still changed into R after Laplace conversion, all capacitors are changed into impedance with the resistance value of 1/Cs, then the relation between the voltage of an output end and the voltage of an input end is solved according to a kirchhoff voltage law and a kirchhoff current law simultaneous equation set, finally, s in a transfer function is replaced into jw to obtain frequency characteristics, and the transfer function can be deduced to be as follows according to the circuit shown in figure 3:
and step four, building a cross-metal body communication model, and testing the frequency response characteristic of the channel through experiments.
In the step, a metal box body with the size of 2.1m multiplied by 1.0m is built in a laboratory, and the middle of the box body is isolated by a metal plate; a forward and backward channel is established in the box body through tin foil paper; the transmitting electrode and the receiving electrode are respectively arranged in two isolated boxes, and the positive end and the negative end of the transmitting electrode are respectively arranged on a front channel and a rear channel. The signal generator adopts UPS uninterrupted power supply to supply power, is separated from the power supply of the receiving end frequency spectrograph, and provides 1mW power signals for the transmitting electrode; the receiving electrode was connected to a spectrometer and the measured amplitude-frequency response of the channel was shown in figure 4.
And fifthly, carrying out optimal parameter estimation by adopting an improved immune algorithm to obtain the numerical values of each resistor and each capacitor in the model.
The improved immune algorithm has the following characteristics: (1) the cloning probability and the cloning scale related to the affinity and the concentration are adopted on cloning, so that the excellent degree and diversity of the offspring population are ensured; (2) in order to improve convergence speed and solving precision, a method of combining selective cross and adaptive Gaussian variation is adopted to improve global search capability, and meanwhile, cross and variation probability is related to affinity, so that the randomness of traditional cross and variation is avoided, and errors are reduced; (3) similarity detection is carried out on the population after cross variation, antibodies with extremely high similarity and relatively low affinity are removed, and the diversity of the population is ensured; (4) and the artificial bee colony algorithm is combined, and the population after each generation of updating is further searched by adopting the artificial bee colony algorithm, so that the global search and the local search are combined, and the convergence speed is improved.
The algorithm flow chart is shown in fig. 5, and the specific steps are as follows:
step 401: and antigen identification, namely adopting a minimum mean square error criterion, taking the mean square error of frequency response characteristics (including amplitude-frequency response characteristics and phase-frequency response characteristics) actually measured by the model and the frequency response characteristics of the equivalent circuit model as an objective function, and taking the two objective functions and a constraint condition as an antigen intrusion system.
The target functions of the amplitude-frequency response and the phase-frequency response are respectively as follows:
Ariand phiriRespectively the measured received power amplitude and phase, A, of each frequency pointri' and phiri' received power amplitude and phase for each frequency point calculated for the theoretical power transfer function, and n is the number of frequency points.
The constraint condition is the value range of resistance and capacitance parameter values, a cabin model is established through ANSYS Maxwell electric field simulation software to obtain the rough value of the coupling capacitance between conductors, so that the value range of the capacitance in the equivalent circuit is determined, and the value range of the resistance in the equivalent circuit is determined according to the weak current loss, namely:
C∈(0,0.5)nF
R∈(0,5)Ω
step 402: antibody coding, namely coding resistance and capacitance parameters needing to be solved as antibodies, and generating the antibodies with the length L ═ m + n according to the difference of the number of the parameters, wherein the resistance value is coded as R1、R2……RmThe capacitance value is coded as C1、C2……Cn。
Where the resistance value is coded as R1、R2、……、R7The capacitance value is coded as C1、C2、……、C17Each antibody is 24 a in length.
Step 403: initializing, randomly generating m × N initial resistance values and N × N initial capacitance values in a value range, and generating an initial parent population a (p) with a population size N1,p2,……，pN) For the jth antibody pj＝(a1j,a2j,……，aLj)。
Here, 7 × N resistance initial values and 17 × N capacitance initial values are randomly generated.
Step 404: and (5) calculating the affinity of each antibody and the antigen in the genetic algebra k and A (k) of the current parent population, namely the fitness of the objective function, and selecting the optimal individual with the minimum mean square error.
For any antibody v of A (k), the affinity was calculated as follows:
if the end condition is met (the optimal solution is found or the maximum iteration number is reached), the operation is terminated and the result is output, otherwise, the next step is continued.
Step 405: and calculating the concentration of each antibody in the current population, and evaluating the excellent degree of the antibody by integrating the affinity and the concentration to obtain the expected reproduction probability.
Wherein the antibody concentration is:
n is the size of the antibody population, ayv,wIs a determination of the magnitude of the affinity between antibody v and antibody w, axv,wIs an antibody v withAffinity between antibodies w, ξ is a preset affinity threshold, dv,kIs the k gene of antibody v, dw,kIs the kth gene of antibody w, and L is the total length of the antibody.
The expected probability of propagation for antibody v is then:
alpha is the adjusting coefficient of affinity and concentration, and is (0,1), ConvAs concentration of antibody v, affvIs the affinity of the antibody v for the antigen.
Step 406: selecting the first i antibodies as a population A1 according to the expected reproduction probability, selecting the first h antibodies according to the affinity to replace the h antibodies with lower affinity in the excellent gene library, and cloning the selected i antibodies according to the cloning probability and the cloning scale related to the affinity and the concentration to form a temporary population B.
In the step, the antibodies in the parent population are sorted from high to low according to the expected reproduction probability and affinity, and the relationship between the selected antibody quantity i and the population size N is as follows: i/N is 0.2. The method has high cloning probability and cloning scale for the antibody with high affinity and low concentration, on the contrary, the cloning probability and cloning scale for the antibody with low affinity and high concentration are relatively low, the cloning probability is calculated according to the expected reproduction probability, and the calculation method of the cloning scale is as follows:
step 407: and (4) crossing and mutating individuals in the cloned population B to change the affinity of the individuals, so that the global search of the population is realized.
In the step, a method combining selective crossover and self-adaptive Gaussian variation is adopted for global search, and simultaneously, crossover and variation probability are related to antibody affinity. The selective crossing is to cross the antibody and the antibody in the excellent gene library, and to improve the cross probability and cross digit for the antibody with low affinity, and to reduce the cross probability and cross digit for the antibody with high affinity, so as to avoid the randomness of the traditional crossing and improve the convergence rate, wherein the cross probability is as follows:
beta is the cross coefficient, affmaxFor maximum affinity, affminFor minimum affinity, affvIs the affinity of the antibody v for the antigen.
The self-adaptive Gaussian variation is that the random number generated in Gaussian distribution with the mean value of mu and the variance of sigma is adopted to make the antibody gene variation, so that the variation range is in a local area near an individual, the randomness of the traditional variation is avoided, the field range is reduced along with the increase of the iteration number, the convergence rate and the solving precision are improved, and the calculation mode of the Gaussian variation is as follows:
aij' is a mutated gene, aijFor the gene before mutation, N (μ, σ) is Gaussian distribution with mean μ and variance σ, and k is evolution algebra and is the initial value of the domain range.
Meanwhile, the mutation probability is improved for the antibody with high affinity, and is reduced for the antibody with low affinity, wherein the mutation probability is as follows:
gamma is coefficient of variation, affmaxFor maximum affinity, affminFor minimum affinity, affvIs the affinity of the antibody v for the antigen.
Step 408: and (3) carrying out similarity detection on the variant population B, removing the antibodies with extremely high similarity and relatively low affinity, and then retaining the first j antibodies with high affinity to form a population B1.
Before similarity detection, the population B is sorted according to descending order of affinity, then the similarity of each antibody v and 2N-v antibodies ranked behind the antibody v is calculated, and if the similarity is larger than a threshold xi, the antibodies with relatively low affinity with the antigen are removed. The last j antibodies retained were related to the population size N by j/N ═ 0.6.
Step 409: r new antibodies are randomly generated to be used as a supplement population C1, and the antibody populations A1 and B1 form a new population A (k +1), so that the population diversity is improved, and the population size is kept to be N.
This step consisted of immunizing a selected antibody population a1 consisting of i antibodies, immunosuppressed antibody population B1 consisting of j antibodies, and randomly generated antibody population C1 consisting of r new antibodies, where r/N is 0.2, to form a new population a (k + 1).
Step 410: and further searching the new population by adopting an artificial bee colony algorithm, thereby effectively reducing the platform period in the population evolution process and improving the convergence speed.
In the step, the leading bees search the updated population (food source) in the field to generate a new solution (food source), if the affinity of the new solution is higher than that of the original solution, the new solution is used for replacing the original solution, otherwise, the original solution is kept unchanged, and the new solution generation mode in the searching process is as follows:
wherein, aij' is a new solution, aijThe method is to solve the problem for the original solution,
is [ -1,1 [ ]]K is an element of [1, N)]，i∈[1,L]。
Step 411: repeating the steps 404-410.
Through the process, the optimal estimation of the resistance and capacitance parameters in the equivalent circuit can be obtained at a higher convergence speed.
Fig. 6 shows an evolutionary graph in the solution process of the immune algorithm, in which the abscissa is the number of iterations and the ordinate is the objective function value, which reflects that the algorithm obtains higher solution accuracy with fewer iterations.
And step six, substituting the optimized resistance and capacitance parameter values obtained in the step five into a circuit to obtain a complete equivalent circuit model, and substituting the parameter values into the transfer function in the step three to obtain a transfer function of a communication channel, so that a model of the quasi-electrostatic field capacitance coupling communication channel is established.
In the step, the parameter values obtained by the algorithm are substituted into a circuit built by ANSYS Simplorer, and the frequency response of the equivalent circuit model is obtained through simulation. Fig. 7 shows the fitting result of the frequency response of the model to the actual frequency response.
Therefore, the model optimization process of the quasi-electrostatic field capacitive coupling communication channel based on the improved immune algorithm optimal parameter estimation is realized.
Although the embodiments of the present invention have been described in conjunction with the accompanying drawings, it will be apparent to those skilled in the art that various modifications may be made without departing from the principles of the invention and these are considered to fall within the scope of the invention.
Claims (6)
1. A quasi-electrostatic field coupling communication model optimization method based on an improved immune algorithm is characterized by comprising the following steps:
the method comprises the following steps that firstly, the coupling relation among conductors in the transmission process of a quasi-electrostatic field is represented by equivalent capacitance, the loss of conduction current is represented by equivalent resistance, and an electric field coupling equivalent model is established;
step two, analyzing the series-parallel relation between each resistor and each capacitor according to the electric field coupling equivalent model obtained in the step one, and establishing a corresponding equivalent circuit model;
step three, deducing a transfer function between the input voltage and the output voltage according to the equivalent circuit model obtained in the step two;
step four, building a cross-metal body communication model, and testing the frequency response characteristic of a channel through experiments;
fifthly, carrying out optimal parameter estimation by adopting an improved immune algorithm to obtain the numerical values of each resistor and each capacitor in the model;
and step six, substituting the optimized resistance and capacitance parameter values obtained in the step five into a circuit to obtain a complete equivalent circuit model, and substituting the parameter values into the transfer function in the step three to obtain a transfer function of a communication channel, so that a model of the quasi-electrostatic field capacitance coupling communication channel is established.
2. The improved immune algorithm-based quasi-electrostatic field coupling communication model optimization method as claimed in claim 1, wherein the electric field coupling equivalent model in the step one determines a negligible capacitance and a non-negligible capacitance by a rough mutual capacitance value between conductors obtained by ANSYS Maxwell electric field simulation, thereby establishing an equivalent circuit model thereof.
3. The improved immune algorithm-based quasi-electrostatic field coupling communication model optimization method as claimed in claim 1, wherein the artificial immune algorithm in the fifth step comprises an improvement on a traditional artificial immune algorithm, and the improved artificial immune algorithm is combined with the equivalent circuit model.
4. The quasi-electrostatic field coupling communication model optimization method based on the improved immune algorithm as claimed in claim 3, wherein the improvement of the traditional artificial immune algorithm comprises: (1) the cloning probability and the cloning scale related to the affinity and the concentration are adopted on cloning, so that the excellent degree and diversity of the offspring population are ensured; (2) in order to improve convergence speed and solving precision, a method of combining selective cross and adaptive Gaussian variation is adopted to improve global search capability, and meanwhile, cross and variation probability is related to affinity, so that the randomness of traditional cross and variation is avoided, and errors are reduced; (3) similarity detection is carried out on the population after cross variation, antibodies with extremely high similarity and relatively low affinity are removed, and the diversity of the population is ensured; (4) and the artificial bee colony algorithm is combined, and the population after each generation of updating is further searched by adopting the artificial bee colony algorithm, so that the global search and the local search are combined, and the convergence speed is improved.
5. The improved immune algorithm-based quasi-electrostatic field coupling communication model optimization method as claimed in claim 3, wherein the traditional artificial immune algorithm is combined with an equivalent circuit model after the improvement, and the model parameters of the quasi-electrostatic field coupling communication channel are optimized and estimated through evolutionary computation, and the method comprises the following steps:
step 401: antigen recognition, namely, adopting a minimum mean square error criterion, taking the mean square error of frequency response characteristics (including amplitude-frequency response characteristics and phase-frequency response characteristics) actually measured by a model and the frequency response characteristics of an equivalent circuit model as an objective function, and taking the two objective functions and a constraint condition as an antigen intrusion system;
step 402: antibody coding, namely coding resistance and capacitance parameters needing to be solved as antibodies, and generating the antibodies with the length L ═ m + n according to the difference of the number of the parameters, wherein the resistance value is coded as R1、R2……RmThe capacitance value is coded as C1、C2……Cn；
Step 403: initializing, randomly generating m × N initial resistance values and N × N initial capacitance values in a value range, and generating an initial parent population a (p) with a population size N1,p2,……，pN) For the jth antibody pj＝(a1j,a2j,……，aLj)；
Step 404: calculating the affinity of each antibody and antigen in the genetic algebra k and A (k) of the current parent population, namely the fitness of a target function, selecting the minimum mean square error as an optimal individual, terminating the operation and outputting a result if an ending condition (finding an optimal solution or reaching the maximum iteration number) is met, and otherwise continuing the next step;
step 405: calculating the concentration of each antibody in the current population, and evaluating the excellent degree of the antibodies by integrating the affinity and the concentration to obtain the expected reproduction probability;
step 406: selecting the first i antibodies as a population A1 according to the expected reproduction probability, selecting the first h antibodies according to the affinity to replace the h antibodies with lower affinity in the excellent gene library, and cloning the selected i antibodies according to the cloning probability and the cloning scale related to the affinity and the concentration to form a temporary population B;
step 407: crossing and mutating individuals in the cloned population B to change the affinity of the individuals, so that the overall search of the population is realized;
step 408: carrying out similarity detection on the variant population B, removing the antibodies with extremely high similarity and relatively low affinity, and then retaining the first j antibodies with high affinity to form a population B1;
step 409: randomly generating r new antibodies as a supplement population C1, forming a new population A (k +1) with the antibody populations A1 and B1, improving the population diversity and keeping the population size to be N;
step 410: the new population is further searched by adopting an artificial bee colony algorithm, so that the platform period in the population evolution process is effectively reduced, and the convergence speed is improved;
step 411: repeating the steps 404-410.
6. The improved immune algorithm-based quasi-electrostatic field coupling communication model optimization method as claimed in claim 5, wherein the capacitance value range in step 403 is determined according to a coarse capacitance value obtained by ANSYS Maxwell electric field simulation, and a simulation capacitance value is added to the initial population, thereby effectively improving the convergence speed of the algorithm.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010951835.0A CN112069736A (en) | 2020-09-11 | 2020-09-11 | Quasi-electrostatic field coupling communication model optimization method based on improved immune algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010951835.0A CN112069736A (en) | 2020-09-11 | 2020-09-11 | Quasi-electrostatic field coupling communication model optimization method based on improved immune algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112069736A true CN112069736A (en) | 2020-12-11 |
Family
ID=73696466
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010951835.0A Pending CN112069736A (en) | 2020-09-11 | 2020-09-11 | Quasi-electrostatic field coupling communication model optimization method based on improved immune algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112069736A (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101188438A (en) * | 2006-11-21 | 2008-05-28 | 索尼株式会社 | Communication system and communication apparatus |
| CN104123421A (en) * | 2014-07-31 | 2014-10-29 | 西安电子科技大学 | Electrostatic forming film reflecting surface form design method based on mechanical and electrical field coupling |
-
2020
- 2020-09-11 CN CN202010951835.0A patent/CN112069736A/en active Pending
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101188438A (en) * | 2006-11-21 | 2008-05-28 | 索尼株式会社 | Communication system and communication apparatus |
| CN104123421A (en) * | 2014-07-31 | 2014-10-29 | 西安电子科技大学 | Electrostatic forming film reflecting surface form design method based on mechanical and electrical field coupling |
Non-Patent Citations (4)
| Title |
|---|
| GONZALEZ等: "Approaching the Fault Passage Indicator", 《2ND INTERNATIONAL SYMPOSIUM ON ENERGY CHALLENGES MECHANICS》 * |
| ZHU L L: "Research on the resource allocation model for the satellite constellation communication system", 《ADVANCED MATERIALS RESEARCH. TRANS TECH PUBLICATIONS LTD》 * |
| 王永泉等: "基于遗传算法的静电反馈式微传感器多目标优化", 《西安交通大学学报》 * |
| 解禹等: "基于准静电场的体域通信信道随机特性及系统性能仿真", 《科技导报》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111291836B (en) | Method for generating student network model | |
| CN109214575B (en) | Ultrashort-term wind power prediction method based on small-wavelength short-term memory network | |
| CN102413029B (en) | Method for partitioning communities in complex dynamic network by virtue of multi-objective local search based on decomposition | |
| Liu et al. | A fault diagnosis intelligent algorithm based on improved BP neural network | |
| CN113985294A (en) | Method and device for estimating remaining life of battery | |
| CN111143984A (en) | Magnetotelluric two-dimensional inversion method based on genetic algorithm optimization neural network | |
| CN112307667A (en) | Method and device for estimating state of charge of storage battery, electronic equipment and storage medium | |
| Wang et al. | A research on the optimal design of BP neural network based on improved GEP | |
| Yu et al. | Dynamic density clustering ant colony algorithm with filtering recommendation backtracking mechanism | |
| CN115938505A (en) | Drug molecule screening method and system based on fusion of graph neural network block structure and multi-head attention mechanism | |
| CN101893852B (en) | Multi-target modeling method for complex industrial process | |
| CN109167113B (en) | Lithium ion battery charging optimization method | |
| CN110852435A (en) | Neural evolution calculation model | |
| CN112069736A (en) | Quasi-electrostatic field coupling communication model optimization method based on improved immune algorithm | |
| CN109459609B (en) | Distributed power supply frequency detection method based on artificial neural network | |
| CN107276559B (en) | Multi-constraint FIR digital filter generation method of quantum biophysical evolution mechanism | |
| CN113704570B (en) | Large-scale complex network community detection method based on self-supervision learning type evolution | |
| CN114327859B (en) | Source model clustering selection method for large-scale problem agent optimization of cloud computing environment | |
| CN115081323A (en) | Method for solving multi-objective constrained optimization problem and storage medium thereof | |
| CN113379035A (en) | Graph neural network structure optimization searching method, system and storage medium | |
| Hu et al. | Apenas: An asynchronous parallel evolution based multi-objective neural architecture search | |
| CN114885345A (en) | System computing energy efficiency maximization resource allocation method based on differential evolution algorithm | |
| CN112132326A (en) | Social network friend prediction method based on random walk punishment mechanism | |
| CN112487187A (en) | News text classification method based on graph network pooling | |
| CN115081744B (en) | Unit energy consumption prediction method for aluminum electrolysis manufacturing system |
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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20201211 |
|
| WD01 | Invention patent application deemed withdrawn after publication |