CN109793514A - Mix the EEG signals coding and decoding methods for-Volterra model that leapfrogs - Google Patents
Mix the EEG signals coding and decoding methods for-Volterra model that leapfrogs Download PDFInfo
- Publication number
- CN109793514A CN109793514A CN201811612253.9A CN201811612253A CN109793514A CN 109793514 A CN109793514 A CN 109793514A CN 201811612253 A CN201811612253 A CN 201811612253A CN 109793514 A CN109793514 A CN 109793514A
- Authority
- CN
- China
- Prior art keywords
- eeg signals
- frog
- formula
- group
- mould
- 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
Landscapes
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
A kind of EEG signals coding and decoding methods mixing the-Volterra model that leapfrogs construct prediction model by being pre-processed to the EEG signals of input, with Volterra modeling method, determine chaos EEG signals prediction model and encode, decoding step forms.The present invention is used and is improved the existing mixing method of leapfroging, preemphasis, adding window, framing pretreatment are carried out to the EEG signals of input, establish EEG signals prediction model, determine the parameter in EEG signals prediction model, complete the coding of EEG signals, the physics biological meaning reflected according to prediction model parameter of analytic model.The present invention utilizes the chaos feature of EEG signals, rapidly and accurately realizes and is encoded, predicted to EEG signals, and with step, simple, easy to accomplish, high accuracy for examination, can be used for encoding EEG signals, decode.
Description
Technical field
The invention belongs to electroencephalogramsignal signal analyzing technical fields, and in particular to Chaotic time series forecasting model.
Background technique
EEG signals reflect the bioelectrical activity of cerebral nervous system, studies have found that, EEG signals time series is non-
Linear, and show as apparent chaotic characteristic.In recent years, with the development of chaology and further grinding to EEG signals
Study carefully, the prediction model using chaotic characteristic building EEG signals becomes a kind of important method for studying EEG signals.It is most of
It is all directly to use Volterra modeling method that researcher, which constructs a Nonlinear Prediction Models:
Its phase space reconfiguration process is cumbersome, and need to be established on the basis of EEG signals chaotic characteristic using evolution algorithm
EEG signals Chaotic time series forecasting models.Existing evolution algorithm efficiency is lower, to particular problem without specific aim.Standard
Mixing leapfrog stage Search are as follows:
D=r (Xb-Xw)
X′w=Xw+D,||D||≤Dmax
This rule limits region of search of the mould because of evolution, not only reduces convergence rate, and easily leads to precocious receipts
It holds back.
Summary of the invention
Technical problem to be solved by the present invention lies in the above-mentioned prior art is overcome, provide a kind of step it is simple,
Easy to accomplish, speed is fast, accuracy rate is high mixing leapfrogs the EEG signals coding and decoding methods of-Volterra model.
Solving technical solution used by above-mentioned technical problem is to comprise the steps of:
(1) EEG signals of input are pre-processed
In the EEG signals of input, the uniform frame of waveform is found as analysis frame, progress preemphasis, adding window, framing are pre-
Processing.
Above-mentioned adding window is carried out using formula (1) window function:
N is limited positive integer in formula.
(2) prediction model is constructed with Volterra modeling method
By the information of step (1) analysis frame, EEG signals prediction model is established by formula (2):
U (n-i τ) is the analysis frame signal of input in formula, and m is that the memory span of Chaotic time series forecasting model is limited
Positive integer, h1(i) and h2(i, j) is undetermined coefficient, and u (n-i τ) is the n-th-i τ sample of correspondence analysis frame, and n-i τ is step
(1) the sample serial number of analysis frame in, u (n-j τ) are the n-th-j τ sample of corresponding analysis frame, and n-j τ is analysis in step (1)
The sample serial number of frame, τ are delay times for limited positive integer, and j, n are limited positive integer.
(3) it determines EEG signals prediction model and encodes
The EEG signals of analysis frame in step (1) are respectively obtained into optimal memory span m with Cao method and mutual information method
And delay time T, undetermined coefficient h corresponding in EEG signals formula (2) is determined with the mixing method of leapfrogingi(i), undetermined coefficient h2(i,
J), by optimal memory span m, delay time T, undetermined coefficient hi(i) and h2(i, j) substitutes into formula (2), obtains EEG signals
Coding.
(4) it decodes
By optimal memory span m, delay time T, the undetermined coefficient h of the EEG signals of extractioni(i) and h2(i, j) generation
Enter formula (2), obtain the prediction model to induction signal, according to the data having after encoding, is conventionally decoded.
Mixing in step of the present invention (3) leapfrogs method step are as follows:
1) frog group is generated
It is random to generate the initial frog group of R frog composition, the fitness function value F of every frog is determined with formula (3)i(t):
Y in formulaiFor the EEG signals value of input,For model predication value, L is that the length of prediction data is limited just whole
Number, is denoted as X for the maximum frog of fitness value in populationg, frog group individual is by the arrangement of fitness value descending:
{X1,X2,XR, XiIndicate i-th frog.
2) mould is generated because of group
Mould is generated because of group by formula (4) are as follows:
Mk={ Xk+Q(l-1)∈P|1≤l≤R},1≤k≤Q (4)
P is the set of frog group, M in formulakIt is k-th of mould because of the set of group frog, Xk+M(l-1)For the frog in frog group, Q is divided into
Mould because group number be limited positive integer.
3) evolution mould because
To MkThe interior frog successively presses formula (5), formula (6), formula (7) and evolves:
D=rc (Xb-Xw)+W (5)
W=[r1ω1,max,r2ω2,max,,rSωS,max]T (6)
S is space dimensionality, random number of the r between (0,1), a constant of the c between [1,2], r in formulai(1≤i≤
S) the random number between [- 1,1], ωi,max(1≤i≤S) is the maximum perception of i-th dimension search space and not knowing for movement
Property, D is evolution step-length, DTIt is the transposition of D, DmaxIt is that the frog allows to change position maximum value, XbAnd XwMould is respectively indicated because in group
The frog with the frog and worst adaptive value that are preferably adapted to value;Mould is because the frog optimal in group is by formula (8) update:
When evolution number is equal to setting the number of iterations 500~2000 times, mould is completed because evolving.
4) hybrid guided mode is because of group
By the mould after the completion of evolution because the frog in group mixes at random, new frog group P ' is obtained, if hybrid guided mode is because of group number i etc.
When the global mixed iteration number 1000 times or detection input signal brain electricity and predicted value of setting obtain error less than 0.1, output
As a result optimal undetermined coefficient h is obtainedi(i) and h2(i, j), otherwise by 1) to frog group sequence and repetition step 2) and step
3), iteration again.
Since the present invention is using the existing mixing method of leapfroging is improved, pre-add is carried out to the EEG signals of input
Weight, adding window, framing pretreatment, establish EEG signals prediction model, determine the parameter in EEG signals prediction model, complete brain
The coding of electric signal is conventionally decoded according to the data having after encoding.The present invention utilizes the spy of EEG signals
Point rapidly and accurately realizes and is encoded, decoded to EEG signals, and it is excellent to have that step is simple, easy to accomplish, accuracy rate is high etc.
Point can be used for encoding EEG signals, decode.
Detailed description of the invention
Fig. 1 is process flow chart of the invention.
Fig. 2 is the waveform diagram for the sample EEG signals that embodiment 1 inputs the 23rd channel.
Specific embodiment
The present invention is described in more detail with reference to the accompanying drawings and examples, but the present invention is not limited to following embodiment party
Formula.
Embodiment 1
To choose collected preceding 500 samples in the 23rd channel in MMSPG brain wave acquisition database as EEG signals
For (as shown in Figure 2), the EEG signals coding and decoding methods step (as shown in Figure 1) for-Volterra model that leapfrogs is mixed
It is as follows:
(1) EEG signals of input are pre-processed
Collected preceding 500 samples in the 23rd channel are chosen in MMSPG brain wave acquisition database as EEG signals,
In 500 EEG signals, the uniform frame of waveform is found as analysis frame, preemphasis, adding window, framing pretreatment is carried out, carries out pre-
It aggravates and framing pretreatment is the conventional method of this field.Adding window is carried out using formula (1) window function:
N is limited positive integer in formula.
(2) prediction model is constructed with Volterra modeling method
By the information of step (1) analysis frame, EEG signals prediction model is established by formula (2):
U (n-i τ) is the analysis frame signal of input in formula, and m is that the memory span of Chaotic time series forecasting model is limited
Positive integer, h1(i) and h2(i, j) is undetermined coefficient, and u (n-i τ) is the n-th-i τ sample of correspondence analysis frame, and n-i τ is step
(1) the sample serial number of analysis frame in, u (n-j τ) are the n-th-j τ sample of corresponding analysis frame, and n-j τ is analysis in step (1)
The sample serial number of frame, τ are delay times for limited positive integer, and j, n are limited positive integer.
(3) it determines EEG signals prediction model and encodes
The EEG signals of analysis frame in step (1) are respectively obtained into optimal memory span m with Cao method and mutual information method
And delay time T, undetermined coefficient h corresponding in EEG signals formula (2) is determined with the mixing method of leapfrogingi(i), undetermined coefficient h2(i,
J), by optimal memory span m, delay time T, undetermined coefficient hi(i) and h2(i, j) substitutes into formula (2), obtains EEG signals
Coding.
Above-mentioned mixing leapfrogs method step are as follows:
1) frog group is generated
It is random to generate the initial frog group of 500 frogs compositions, the fitness function value F of every frog is determined with formula (3)i(t):
Y in formulaiFor the EEG signals value of input,For model predication value, L is that the length of prediction data is 500, by population
The middle maximum frog of fitness value is denoted as Xg, frog group is individual to be arranged as { X by fitness value descending1,X2,XR, XiIndicate i-th frog.
2) mould is generated because of group
Mould is generated because of group by formula (4) are as follows:
Mk={ Xk+Q(l-1)∈P|1≤l≤R},1≤k≤Q (4)
P is the set of frog group, M in formulakIt is k-th of mould because of the set of group frog, Xk+M(l-1)For the frog in frog group, Q is divided into
Mould because group number be 50.
3) evolution mould because
To MkThe interior frog successively presses formula (5), formula (6), formula (7) and evolves:
D=rc (Xb-Xw)+W (5)
W=[r1ω1,max,r2ω2,max,,rSωS,max]T (6)
S is space dimensionality, r 0.78, c 1.5, r in formulaiIt is 0.5, ωi,max(1≤i≤S) is i-th dimension search space
Maximum perception and movement uncertainty, D is evolution step-length, DTIt is the transposition of D, DmaxIt is that the frog allows to change position maximum
Value, XbAnd XwMould is respectively indicated because having the frog of the frog and worst adaptive value that are preferably adapted to value in group;Mould presses formula because of the frog optimal in group
(8) it updates:
When evolution number is equal to setting the number of iterations 1000 times, mould is completed because evolving.
4) hybrid guided mode is because of group
By the mould after the completion of evolution because the frog in group mixes at random, new frog group P ' is obtained, if hybrid guided mode is because of group number i etc.
When the global mixed iteration number 1000 times or detection input signal brain electricity of setting and the error of predicted value are less than 0.2, output
As a result optimal undetermined coefficient h is obtainedi(i) and h2(i, j), otherwise by 1) to frog group sequence and repetition step 2) and step
3), iteration again.
The EEG signals in the 23rd channel chosen with the mixing method of leapfroging correspond to undetermined coefficient h1(i) and undetermined coefficient h2
(i, j) is shown in Table 1, table 2.
The best undetermined coefficient h of 1 embodiment 1 of table1(i)
h1(1) | h1(1) | h1(1) | h1(1) | h1(1) | h1(1) | h1(1) | h1(1) |
0.3587 | 0.7562 | -0.2718 | 1 | 0.2564 | 0.1522 | -1 | 0.7462 |
The best undetermined coefficient h of 2 embodiment 1 of table2(i, j)
h2(i, j) | I=1 | I=2 | I=3 | I=4 | I=5 | I=6 | I=7 | I=8 |
J=1 | 0.3587 | 0.7562 | -0.2718 | 1 | 0.2564 | 0.1522 | -1 | 0.7462 |
J=2 | 0.3649 | -0.5612 | 0.7512 | -0.1163 | 0.9547 | 0.7624 | -0.4413 | |
J=3 | -0.2489 | 0.8526 | 0.5649 | -0.7436 | 0.8534 | -0.6523 | ||
J=4 | -0.8529 | 0.2356 | 0.3214 | -0.5342 | 0.7243 | |||
J=5 | -0.5124 | 0.2365 | -1.0023 | 1 | ||||
J=6 | -0.5496 | 0.1554 | 0.2542 | |||||
J=7 | 0.3265 | 0.4185 | ||||||
J=8 | -0.3458 |
By table 1, table 2 as it can be seen that the EEG signals undetermined coefficient h in the 23rd channel chosen1(i)、h2(i, j) is data in table
When, the worst error of sample cumulative is 0.0928, has reached error range, by h1(i)、h2(i, j) substitutes into formula (2), obtains chaos
The coding of voice signal.
(4) it decodes
By optimal memory span m, delay time T, the undetermined coefficient h of the EEG signals of extractioni(i) and h2(i, j) generation
Enter formula (2), obtain the prediction model to induction signal, according to the data having after encoding, is conventionally decoded.
Embodiment 2
To choose collected preceding 500 samples in the 23rd channel in MMSPG brain wave acquisition database as EEG signals
For, mixing leapfrogs, and steps are as follows for the EEG signals coding and decoding methods of-Volterra model:
(1) EEG signals of input are pre-processed
The step is same as Example 1.
(2) prediction model is constructed with Volterra modeling method
The step is same as Example 1.
(3) it determines EEG signals prediction model and encodes
The mixing of the step leapfrogs method step are as follows:
1) frog group is generated
The step is same as Example 1.
2) mould is generated because of group
The step is same as Example 1.
3) evolution mould because
To MkThe interior frog successively presses formula (5), formula (6), formula (7) and evolves:
D=rc (Xb-Xw)+W (5)
W=[r1ω1,max,r2ω2,max,,rSωS,max]T (6)
S is space dimensionality, r 0.01, c 1, r in formulaiFor -1, ωi,max(1≤i≤S) be i-th dimension search space most
The big uncertainty perceived and move, D are evolution step-length, DTIt is the transposition of D, DmaxIt is that the frog allows to change position maximum value, Xb
And XwMould is respectively indicated because having the frog of the frog and worst adaptive value that are preferably adapted to value in group;Mould presses formula (8) because of the frog optimal in group
It updates:
When evolution number is equal to setting the number of iterations 500 times, mould is completed because evolving.
Determine that other steps in EEG signals prediction model and coding step (3) are same as Example 1.
The EEG signals in the 23rd channel chosen with the mixing method of leapfroging correspond to undetermined coefficient h1(i) and undetermined coefficient h2
(i, j) is shown in Table 3, table 4.
Best undetermined coefficient h in 3 embodiment 2 of table1(i)
h1(1) | h1(2) | h1(3) | h1(4) | h1(5) | h1(6) | h1(7) | h1(8) |
0.6895 | -0.3537 | 0 | -0.9450 | -2.2977 | 0.7788 | 0 | 0.9151 |
Best undetermined coefficient h in 4 embodiment 2 of table2(i, j)
h2(i, j) | I=1 | I=2 | I=3 | I=4 | I=5 | I=6 | I=7 | I=8 |
J=1 | 0 | 0 | 3.3789 | -10.6625 | 0 | 0 | 0 | 0.0572 |
J=2 | 0.7724 | 0 | 13.9320 | -74.4365 | 5.8409 | 3.99.9 | 0 | |
J=3 | -0.3361 | 0 | -0.8273 | 40.2332 | -0.1804 | -1.0835 | ||
J=4 | 1.7927 | 0 | -0.0065 | 0 | -2.9221 | |||
J=5 | 4.5823 | 0 | 0 | -2.2859 | ||||
J=6 | 0 | 0.0353 | 0 | |||||
J=7 | 0 | 0 | ||||||
J=8 | 9.2989 |
By table 3, table 4 as it can be seen that the preceding 500 samples EEG signals undetermined coefficient h for the 23rd channel acquisition chosen1(i)、h2
When (i, j) is data in table, the worst error of sample cumulative is 0.1128, has reached error range, by h1(i)、h2(i, j) generation
Enter formula (2), obtains the coding of chaos voice signal.
Other steps are same as Example 1.
Embodiment 3
To choose collected preceding 500 samples in the 23rd channel in MMSPG brain wave acquisition database as EEG signals
For, mixing leapfrogs, and steps are as follows for the EEG signals coding and decoding methods of-Volterra model:
(1) EEG signals of input are pre-processed
The step is same as Example 1.
(2) prediction model is constructed with Volterra modeling method
The step is same as Example 1.
(3) it determines EEG signals prediction model and encodes
The mixing of the step leapfrogs method step are as follows:
1) frog group is generated
The step is same as Example 1.
2) mould is generated because of group
The step is same as Example 1.
3) evolution mould because
To MkThe interior frog successively presses formula (5), formula (6), formula (7) and evolves:
D=rc (Xb-Xw)+W(5)
W=[r1ω1,max,r2ω2,max,,rSωS,max]T (6)
S is space dimensionality, r 0.99, c 2, r in formulaiIt is 1, ωi,max(1≤i≤S) be i-th dimension search space most
The big uncertainty perceived and move, D are evolution step-length, DTIt is the transposition of D, DmaxIt is that the frog allows to change position maximum value, Xb
And XwMould is respectively indicated because having the frog of the frog and worst adaptive value that are preferably adapted to value in group;Mould presses formula (8) because of the frog optimal in group
It updates:
When evolution number is equal to setting the number of iterations 2000 times, mould is completed because evolving.
Determine that other steps in EEG signals prediction model and coding step (3) are same as Example 1.
The EEG signals in the 23rd channel chosen with the mixing method of leapfroging correspond to undetermined coefficient h1(i) and undetermined coefficient h2
(i, j) is shown in Table 5, table 6.
Table 5 applies the best undetermined coefficient h of example 31(i)
h1(1) | h1(1) | h1(1) | h1(1) | h1(1) | h1(1) | h1(1) | h1(1) |
0.9860 | 0.9670 | 0.9535 | 0.9418 | 0.9299 | 0.9159 | 0.9027 | 0.8892 |
The best undetermined coefficient h of 6 embodiment 3 of table2(i, j)
h2(i, j) | I=1 | I=2 | I=3 | I=4 | I=5 | I=6 | I=7 | I=8 |
J=1 | 0.8146 | 0.8573 | 0.8379 | 0.8619 | 0.8721 | 0.8972 | 0.9034 | 0.9191 |
J=2 | 0.9330 | 0.9487 | 0.9634 | 0.9775 | 0.9973 | 0.9775 | 0.9973 | |
J=3 | 1.0191 | 0.8521 | 0.8661 | 0.8852 | 0.8915 | 0.9121 | ||
J=4 | 0.9261 | 0.9431 | 0.9559 | 0.9699 | 0.9892 | |||
J=5 | 0.8500 | 0.8648 | 0.8834 | 0.8944 | ||||
J=6 | 0.9401 | 0.9524 | 0.9663 | |||||
J=7 | 0.8496 | 0.8638 | ||||||
J=8 | 0.8810 |
By table 5, table 6 as it can be seen that the preceding 500 samples EEG signals undetermined coefficient h for the 23rd channel acquisition chosen1(i)、h2
When (i, j) is data in table, the worst error of sample cumulative is 0.0741, has reached error range, by h1(i)、h2(i, j) generation
Enter formula (2), obtains the coding of chaos voice signal.
Other steps are same as Example 1.
Embodiment 4
In above Examples 1 to 3, different channels is chosen in MMSPG brain wave acquisition database and acquires 10000
Within sample as EEG signals, mixing leapfrog-Volterra model EEG signals coding and decoding methods and embodiment 1~
3 is identical.
Claims (2)
- The EEG signals coding and decoding methods of-Volterra model 1. a kind of mixing leapfrogs, it is characterised in that by following steps Composition:(1) EEG signals of input are pre-processedIn the EEG signals of input, the uniform frame of waveform is found as analysis frame, carries out preemphasis, adding window, framing pretreatment;Above-mentioned adding window is carried out using formula (1) window function:N is limited positive integer in formula;(2) prediction model is constructed with Volterra modeling methodBy the information of step (1) analysis frame, EEG signals prediction model is established by formula (2):In formula u (n-i τ) be input analysis frame signal, m be Chaotic time series forecasting model memory span be it is limited just Integer, h1(i) and h2(i, j) is undetermined coefficient, and u (n-i τ) is the n-th-i τ sample of correspondence analysis frame, and n-i τ is step (1) The sample serial number of middle analysis frame, u (n-j τ) are the n-th-j τ sample of corresponding analysis frame, and n-j τ is analysis frame in step (1) Sample serial number, τ be delay time be limited positive integer, j, n be limited positive integer;(3) it determines EEG signals prediction model and encodesThe EEG signals of analysis frame in step (1) are respectively obtained into optimal memory span m with Cao method and mutual information method and are prolonged Slow time τ, undetermined coefficient h corresponding in EEG signals formula (2) is determined with the mixing method of leapfrogingi(i), undetermined coefficient h2(i, j), By optimal memory span m, delay time T, undetermined coefficient hi(i) and h2(i, j) substitutes into formula (2), obtains the volume of EEG signals Code;(4) it decodesBy optimal memory span m, delay time T, the undetermined coefficient h of the EEG signals of extractioni(i) and h2(i, j) substitutes into formula (2), the prediction model to induction signal is obtained, according to the data having after encoding, is conventionally decoded.
- 2. it is according to claim 1 based on mixing leapfrog method Volterra modeling EEG signals coding and decoding methods, The method step it is characterized in that the mixing in the step (3) leapfrogs are as follows:1) frog group is generatedIt is random to generate the initial frog group of R frog composition, the fitness function value F of every frog is determined with formula (3)i(t):Y in formulaiFor the EEG signals value of input,For model predication value, L is that the length of prediction data is limited positive integer, will The maximum frog of fitness value is denoted as X in populationg, frog group individual is by the arrangement of fitness value descending:{X1,X2,XR, XiIndicate i-th frog;2) mould is generated because of groupMould is generated because of group by formula (4) are as follows:Mk={ Xk+Q(l-1)∈P|1≤l≤R},1≤k≤Q (4)P is the set of frog group, M in formulakIt is k-th of mould because of the set of group frog, Xk+M(l-1)For the frog in frog group, the mould that Q is divided into Because group number is limited positive integer;3) evolution mould becauseTo MkThe interior frog successively presses formula (5), formula (6), formula (7) and evolves:D=rc (Xb-Xw)+W(5)W=[r1ω1,max,r2ω2,max, ,rSωS,max]T (6)S is space dimensionality, random number of the r between (0,1), a constant of the c between [1,2], r in formulai(1≤i≤S) is Random number between [- 1,1], ωi,max(1≤i≤S) is the uncertainty of the maximum perception and movement of i-th dimension search space, D For evolution step-length, DTIt is the transposition of D, DmaxIt is that the frog allows to change position maximum value, XbAnd XwMould is respectively indicated because having in group It is preferably adapted to the frog of value and the frog of worst adaptive value;Mould is because the frog optimal in group is by formula (8) update:When evolution number is equal to setting the number of iterations 500~2000 times, mould is completed because evolving;4) hybrid guided mode is because of groupBy the mould after the completion of evolution because the frog in group mixes at random, new frog group P ' is obtained, if hybrid guided mode is set because group number i is equal to When fixed global mixed iteration number 1000 times or detection input signal brain electricity and predicted value obtain error less than 0.1, result is exported Obtain optimal undetermined coefficient hi(i) and h2(i, j) is otherwise pressed and step 2) and step 3) 1) is sorted and repeated to frog group, then Secondary iteration.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811612253.9A CN109793514A (en) | 2018-12-27 | 2018-12-27 | Mix the EEG signals coding and decoding methods for-Volterra model that leapfrogs |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201811612253.9A CN109793514A (en) | 2018-12-27 | 2018-12-27 | Mix the EEG signals coding and decoding methods for-Volterra model that leapfrogs |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109793514A true CN109793514A (en) | 2019-05-24 |
Family
ID=66557820
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201811612253.9A Pending CN109793514A (en) | 2018-12-27 | 2018-12-27 | Mix the EEG signals coding and decoding methods for-Volterra model that leapfrogs |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109793514A (en) |
Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108417220A (en) * | 2018-02-11 | 2018-08-17 | 陕西师范大学 | Voice signal coding and decoding methods based on agent model Volterra modelings |
-
2018
- 2018-12-27 CN CN201811612253.9A patent/CN109793514A/en active Pending
Patent Citations (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN108417220A (en) * | 2018-02-11 | 2018-08-17 | 陕西师范大学 | Voice signal coding and decoding methods based on agent model Volterra modelings |
Non-Patent Citations (2)
Title |
---|
THAI-HOANG HUYNH: "A modified shuffled frog leaping algorithm for optimal tuning of multivariable PID controllers", 《2008 IEEE INTERNATIONAL CONFERENCE ON INDUSTRIAL TECHNOLOGY》 * |
张玉梅等: "语音信号序列的Volterra预测模型", 《物理学报》 * |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN111259761B (en) | Electroencephalogram emotion recognition method and device based on movable attention neural network | |
Auffinger et al. | Parisi formula for the ground state energy in the mixed p-spin model | |
RU98113925A (en) | METHOD AND DEVICE OF SCALABLE CODING-DECODING OF STEREOPHONIC AUDIO SIGNAL (OPTIONS) | |
CN103778919A (en) | Speech coding method based on compressed sensing and sparse representation | |
RU2016105613A (en) | AUDIO CODER, AUDIO DECODER AND RELATED METHODS USING TWO-CHANNEL PROCESSING IN INFRASTRUCTURE OF INTELLECTUAL FILLING OF SIGNAL INTERVALS | |
CN104901795B (en) | Physical layer key extraction method based on channel characteristics | |
CN110473557B (en) | Speech signal coding and decoding method based on depth self-encoder | |
CN112883690A (en) | Translation model | |
CN115296862B (en) | Network data safety transmission method based on data coding | |
CN111597971B (en) | Urban rail transit short-term arrival passenger flow prediction method | |
Wang et al. | Modified algorithms for fast construction of optimal Latin-hypercube design | |
CN108417220B (en) | Voice signal coding and decoding methods based on agent model Volterra modeling | |
CN107579816A (en) | Password dictionary generation method based on recurrent neural network | |
CN111312228A (en) | End-to-end-based voice navigation method applied to electric power enterprise customer service | |
CN109525253A (en) | Coding method of convolution code based on deep learning and integrated approach | |
CN109525254A (en) | Convolutional code soft-decision decoding method based on deep learning | |
Fogel et al. | Universal batch learning with log-loss | |
CN105118512A (en) | General steganalysis method facing AAC digital audio | |
CN116562365A (en) | Deep quantization federal learning method for privacy protection | |
CN115064216A (en) | Protein coding method based on position sequence matrix | |
Mehmetoglu et al. | Deterministic annealing-based optimization for zero-delay source-channel coding in networks | |
CN114219027A (en) | Lightweight time series prediction method based on discrete wavelet transform | |
CN109793514A (en) | Mix the EEG signals coding and decoding methods for-Volterra model that leapfrogs | |
CN106782577A (en) | A kind of voice signal coding and decoding methods based on Chaotic time series forecasting model | |
CN113095377A (en) | Dangerous driving scene data random generation method and 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 | ||
RJ01 | Rejection of invention patent application after publication | ||
RJ01 | Rejection of invention patent application after publication |
Application publication date: 20190524 |