CN115145156B - Self-adaptive anti-synchronization method of inertial memristor neural network - Google Patents
Self-adaptive anti-synchronization method of inertial memristor neural network Download PDFInfo
- Publication number
- CN115145156B CN115145156B CN202210911743.9A CN202210911743A CN115145156B CN 115145156 B CN115145156 B CN 115145156B CN 202210911743 A CN202210911743 A CN 202210911743A CN 115145156 B CN115145156 B CN 115145156B
- Authority
- CN
- China
- Prior art keywords
- inertial
- neural network
- memristor
- adaptive
- synchronization
- 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
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 37
- 238000000034 method Methods 0.000 title claims abstract description 19
- 230000004044 response Effects 0.000 claims abstract description 30
- 230000003044 adaptive effect Effects 0.000 claims abstract description 29
- 210000002569 neuron Anatomy 0.000 claims description 6
- 230000004913 activation Effects 0.000 claims description 5
- 230000001360 synchronised effect Effects 0.000 claims 1
- 230000006870 function Effects 0.000 description 10
- 238000004088 simulation Methods 0.000 description 4
- 238000004891 communication Methods 0.000 description 2
- 238000010276 construction Methods 0.000 description 2
- 230000008571 general function Effects 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 230000008569 process Effects 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 210000004556 brain Anatomy 0.000 description 1
- 230000006386 memory function Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G05—CONTROLLING; REGULATING
- G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
- G05B13/042—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- Medical Informatics (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
Abstract
The invention provides a self-adaptive anti-synchronization method of an inertial memristor neural network, which comprises the following steps: step S1: based on the inertial memristor neural network, a driving system and a response system of the inertial memristor neural network with unbounded distribution time lag are established; step S2: establishing an anti-synchronization error system according to a driving system and a response system of the inertial memristor neural network with unbounded distribution time lag established in the step S1; step S3: the adaptive controller is designed such that the drive system and the response system are desynchronized. The invention can realize self-adaptive desynchronization of the inertial memristor neural network without boundary distribution time lag.
Description
Technical Field
The invention relates to the field of information and communication science, in particular to a self-adaptive anti-synchronization method of an inertial memristor neural network.
Background
The resistance of a memristor is determined by the charge flowing through it, so that the amount of charge flowing through it, which is the memory function of the memristor, can be obtained by measuring the resistance of the memristor. Based on the characteristics of the memristor, a memristor neural network which is very suitable for simulating human brain is constructed by replacing the resistor in the traditional neural network circuit. In recent years, memristive neural networks have been highly focused by scientists as their advantages are increasingly apparent.
The desynchronization is an important dynamic behavior in the memristive neural network, and has important application prospects in the aspects of cooperative control, safety communication and the like of artificial intelligence. The desynchronization of memristive neural networks may also be applied in the field of information security, for example: image encryption and associative memory. Therefore, the research on the self-adaptive anti-synchronization method of the inertial memristor neural network with unbounded distribution time lag is a work with positive significance.
Disclosure of Invention
Therefore, the invention aims to provide an adaptive anti-synchronization method of an inertial memristor neural network, which can realize the adaptive anti-synchronization of the inertial memristor neural network with unbounded distribution time lag.
The invention is realized by adopting the following scheme: an adaptive anti-synchronization method of an inertial memristor neural network comprises the following steps:
step S1: based on the inertial memristor neural network, a driving system and a response system of the inertial memristor neural network with unbounded distribution time lag are established;
step S2: establishing an anti-synchronization error system according to a driving system and a response system of the inertial memristor neural network with unbounded distribution time lag established in the step S1;
step S3: the adaptive controller is designed such that the drive system and the response system achieve an adaptive anti-synchronization.
Further, the step S1 specifically includes:
step S11: establishing a state equation of a driving system of an inertial memristive neural network with unbounded distributed time lags:
step S12: establishing a state equation of a response system of an inertial memristive neural network with unbounded distributed time lags:
wherein x is i (t) and y i (t) represents the state variable at time t for the ith neuron, α i ,β i Is constant and satisfies alpha i >0,β i >0,f j (x j (t)) and f j (y j (t)) represents the activation function of the jth neuron, τ j And (t) is time lag, kij (t):is a non-negative delay core real value function related to unbounded distributed time lags, the initial value of the driving system (1) satisfies x i (s)=φ i (s),/>s∈[-∞,0],φ i (s),/>The initial value of the response system (2) satisfies +.>Wherein a is ij (x i (t)),b ij (x i (t)),c ij (x i (t)),a ij (y i (t)),b ij (y i (t)),c ij (y i (t)) represents memristor weights, respectively satisfying:
wherein, gamma i Is a switching threshold and gamma i >0。
Further, the step S2 specifically includes:
step S21: the anti-synchronization error of the driving system and the response system is set as follows:
e i (t)=y i (t)+x i (t) (4)
obtaining an anti-synchronization error system (5):
further, the step S3 specifically includes:
step S31: the constructed adaptive controller expression is:
wherein, gamma i (t) and ζ i (t) is the control gain,ρ i >0,m i the (t) is specifically as follows: />
Bringing the adaptive controller (6) into the anti-synchronization error system (5), two cases of the anti-synchronization error system (5) are obtained:
(1) When |x i (t)|≤γ i ,|y i (t)|≤γ i Or |x i (t)|>γ i ,|y i (t)|≤γ i Obtaining an anti-synchronization error system (7):
(2) When |x i (t)|>γ i ,|y i (t)|>γ i Or |x i (t)|≤γ i ,|y i (t)|>γ i Obtaining an anti-synchronization error system (8):
according to the above cases (1) and (2), an anti-synchronization error system (9) is obtained:
wherein f j (e j (t))=f j (x j (t))+f j (y j (t)),f j (e j (t-τ j (t)))=f j (x j (t-τ j (t)))+f j (y j (t-τ j (t)))。
The invention is based on Lyapunov stability theory, combines the self-adaptive controller, and proves that the driving system and the response system are in anti-synchronization, and specifically comprises the following steps:
the construction of Lyapunov general functions is specifically as follows:
deriving the Lyapunov generalization function, and bringing an error system (9) into the derivative of the Lyapunov generalization function to obtain:
consistent with the hypothesis of the present invention, it has proven to be effective to derive the adaptive anti-synchronization stability theory of the present invention.
Compared with the prior art, the invention has the following beneficial effects:
1. compared with the prior art, under the condition of the same system parameters and controller gains, the method has the advantages that the anti-synchronization process of the driving-responding system is simpler and easier to understand;
2. the self-adaptive controller designed by the invention can effectively enable the driving system and the response system to achieve anti-synchronization, so that the self-adaptive controller has wider application scenes and improves the stability of the system.
Drawings
FIG. 1 is a flow chart of an adaptive anti-synchronization method of an inertial memristive neural network of the present disclosure;
FIG. 2 shows the anti-synchronization error e without controller in embodiment 1 of the present invention 1 (t)、e 2 A curve of (t);
FIG. 3In embodiment 1 of the present invention, there is an adaptive controller-based anti-synchronization error e 1 (t)、e 2 A curve of (t);
FIG. 4 shows the control gain gamma of the adaptive controller according to embodiment 1 of the present invention 1 (t)、γ 2 (t)、ξ 1 (t)、ξ 2 A curve of (t);
FIG. 5 shows the driving system and the response system state x under the control of the adaptive controller in embodiment 1 of the present invention 1 (t) and y 1 An anti-synchronization curve of (t);
FIG. 6 shows the driving system and response system state x under the control of the adaptive controller in embodiment 1 of the present invention 2 (t) and y 2 An anti-synchronization curve of (t).
Detailed Description
To facilitate an understanding of this patent, the patent will be described more fully below with reference to the accompanying drawings. It should be noted that the following detailed description is exemplary and is intended to provide further explanation of the present application. Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs. It is noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to be limiting of example embodiments in accordance with the present application.
As shown in fig. 1, the embodiment provides an adaptive anti-synchronization method of an inertial memristor neural network, including the following steps:
step S1: based on the inertial memristor neural network, a driving system and a response system of the inertial memristor neural network with unbounded distribution time lag are established;
step S2: establishing an anti-synchronization error system according to a driving system and a response system of the inertial memristor neural network with unbounded distribution time lag established in the step S1;
step S3: the adaptive controller is designed such that the drive system and the response system achieve an adaptive anti-synchronization.
In this embodiment, step S1 specifically includes:
step S11: establishing a state equation of a driving system of an inertial memristive neural network with unbounded distributed time lags:
step S12: establishing a state equation of a response system of an inertial memristive neural network with unbounded distributed time lags:
wherein x is i (t) and y i (t) represents the state variable at time t for the ith neuron, α i ,β i Is constant and satisfies alpha i >0,β i >0,f j (x j (t)) and f j (y j (t)) represents the activation function of the jth neuron, τ j (t) is time lag, K ij (t):Is a non-negative delay core real value function with respect to unbounded distributed time lags, the initial value of the drive system (14) satisfies x i (s)=φ i (s),/>The initial value of the response system (15) satisfies +.>And a ij (x i (t)),b ij (x i (t)),c ij (x i (t)),a ij (y i (t)),b ij (y i (t)),c ij (y i (t)) represents memristor weights, respectively satisfying: />
Wherein the method comprises the steps of,γ i Is a switching threshold and gamma i >0。
Further, the step S2 specifically includes:
step S21: the anti-synchronization error of the driving system and the response system is set as follows:
e i (t)=y i (t)+x i (t) (17)
obtaining an anti-synchronization error system (18):
further, the step S3 specifically includes:
step S31: the constructed adaptive controller expression is:
bringing the adaptive controller (19) into the anti-synchronization error system (18), two cases of the anti-synchronization error system (18) are obtained:
(1) When |x i (t)|≤γ i ,|y i (t)|≤γ i Or |x i (t)|>γ i ,|y i (t)|≤γ i Obtaining an anti-synchronization error system (20):
(2) When |x i (t)|>γ i ,|y i (t)|>γ i Or |x i (t)|≤γ i ,|y i (t)|>γ i Obtaining an anti-synchronization error system (21):
according to the above cases (1) and (2), the anti-synchronization error system is obtained as (22):
wherein f j (e j (t))=f j (x j (t))+f j (y j (t)),f j (e j (t-τ j (t)))=f j (x j (t-τ j (t)))+f j (y j (t-τ j (t)))。
For the drive system and the response system, if desynchronization needs to be achieved by the designed adaptive controller, three assumptions first need to be satisfied:
suppose 1: activation function f j (. Cndot.) is Lipohsz continuous, i.e. there is a constant F j > 0, such that
|F j (y j (s))-f j (x j (s))|≤F j |y j (s)-x j (s)| (23)
Suppose 2: time lag τ j (t) (j=1, 2, …, n) satisfies
Wherein τ 1 And τ 2 Is a positive constant.
Suppose 3: storing the articlesAt normal several kappa ij (I, j=1, 2, …, n) such that the following formula holds
The invention is based on Lyapunov stability theory, and combines the self-adaptive controller to prove the anti-synchronization of the driving system and the response system, and the specific contents are as follows:
the construction of Lyapunov general functions is specifically as follows:
deriving the Lyapunov generalization function and bringing an error system (22) into the derivative of the Lyapunov generalization function to obtain:
the method further comprises the following steps:
the above inequality is taken into the derivative equation of Lypunov and, according to assumption 2, the inequality is found to be specifically:
the method further comprises the following steps:
according to the above proving process, the driving system (14) and the response system (15) can reach an anti-synchronization under the influence of the adaptive controller (19).
Example 1
Inertial memristor neural network model driving system (30) with borderless distribution time lag is as follows
Inertial memristor neural network model response system (31) with borderless distribution time lag is as follows
Wherein the activation function is f j (x j (·))=tanh(x j (·)),K ij =e -θ And τ j (t)=0.1e t /(1+e t ) I, j=1, 2. The parameter is chosen to be alpha 1 =α 2 =1,β 1 =β 2 =0.8,κ ij =1,τ 1 =1.25,τ 2 =0.5,á 11 =-2,à 11 =-2.2,á 12 =0.5,à 12 =1,á 21 =6,à 21 =4,á 22 =-2.4,à 22 =3, The adaptive controller parameter is set to +.>γ 1 (t)=γ 2 (t)=ξ 1 (t)=ξ 2 (t) =0. The initial condition is set to x 1 (t)=x 2 (t)=1,/>y 1 (t)=y 2 (t)=-0.8,/>
The following was a simulation experiment based on the specific parameters selected above. Shown in FIG. 2, which shows the simulation results, is the anti-synchronization error e without the addition of a controller 1 (t),e 2 (t) curve, FIG. 3 shows the anti-synchronization error e of the adaptive controller 1 (t),e 2 (t) curve, FIG. 4 of simulation experiment results shows the control gain γ of the adaptive controller 1 (t),γ 2 (t),ξ 1 (t),ξ 2 (t) curves, FIGS. 5 and 6 of simulation experiment results show the driving system and response system states x under the adaptive controller, respectively 1 (t) and y 1 (t)、x 2 (t) and y 2 An anti-synchronization curve of (t).
The foregoing description of the preferred embodiment of the invention is not intended to be limiting, but rather is intended to cover all modifications, equivalents, and alternatives falling within the spirit and principles of the invention.
Claims (3)
1. The self-adaptive anti-synchronization method of the inertial memristor neural network is characterized by comprising the following steps of:
step S1: based on the inertial memristor neural network, a driving system and a response system of the inertial memristor neural network with unbounded distribution time lag are established; the step S1 specifically comprises the following steps:
step S11: establishing a state equation of a driving system of an inertial memristive neural network with unbounded distributed time lags:
step S12: establishing a state equation of a response system of an inertial memristive neural network with unbounded distributed time lags:
wherein x is i (t) and y i (t) represents the state variable at time t for the ith neuron, α i ,β i Is constant and satisfies alpha i >0,β i >0,f j (x j (t)) and f j (y j (t)) represents the activation function of the jth neuron, τ j (t) is time lag, K ij (t):Is a non-negative delay core real value function related to unbounded distribution time lag, and the initial value of a driving system meets x i (s)=φ i (s),/> s∈[-∞,0],φ i (s),/>Initial value of response system satisfies-> s∈[-∞,0],/>Wherein a is ij (x i (t)),b ij (x i (t)),c ij (x i (t)),a ij (y i (t)),b ij (y i (t)),c ij (y i (t)) represents memristor weights, respectively satisfying:
wherein, gamma i Is a switching threshold and gamma i >0;
Step S2: establishing an anti-synchronization error system according to a driving system and a response system of the inertial memristor neural network with unbounded distribution time lag established in the step S1;
step S3: the adaptive controller is designed such that the drive system and the response system are desynchronized.
2. The adaptive anti-synchronization method of an inertial memristor neural network according to claim 1, wherein step S2 specifically comprises:
step S21: the anti-synchronization error of the driving system and the response system is set as follows:
e i (t)=y i (t)+x i (t)
the obtained anti-synchronization error system is as follows:
3. the adaptive anti-synchronization method of an inertial memristor neural network according to claim 2, wherein step S3 specifically comprises:
step S31: the constructed adaptive controller expression is:
the self-adaptive controller is brought into the anti-synchronous error system, and two conditions of the anti-synchronous error system are obtained:
(1) When |x i (t)|≤γ i ,|y i (t)|≤γ i Or |x i (t)|>γ i ,|y i (t)|≤γ i The obtained anti-synchronization error system is as follows:
(2) When |x i (t)|>γ i ,|y i (t)|>γ i Or |x i (t)|≤γ i ,|y i (t)|>γ i Obtain the inverseThe synchronous error system is as follows:
according to the above cases (1) and (2), an anti-synchronization error system is obtained as follows:
wherein f j (e j (t))=f j (x j (t))+f j (y j (t)),f j (e j (t-τ j (t)))=f j (x j (t-τ j (t)))+f j (y j (t-τ j (t)))。
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210911743.9A CN115145156B (en) | 2022-07-28 | 2022-07-28 | Self-adaptive anti-synchronization method of inertial memristor neural network |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202210911743.9A CN115145156B (en) | 2022-07-28 | 2022-07-28 | Self-adaptive anti-synchronization method of inertial memristor neural network |
Publications (2)
Publication Number | Publication Date |
---|---|
CN115145156A CN115145156A (en) | 2022-10-04 |
CN115145156B true CN115145156B (en) | 2023-06-02 |
Family
ID=83414530
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202210911743.9A Active CN115145156B (en) | 2022-07-28 | 2022-07-28 | Self-adaptive anti-synchronization method of inertial memristor neural network |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN115145156B (en) |
Families Citing this family (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN115860096B (en) * | 2022-12-08 | 2023-07-07 | 盐城工学院 | Exponential synchronization control method for mixed time-varying time-lag inertial neural network |
CN115857349B (en) * | 2022-12-08 | 2023-05-30 | 盐城工学院 | Index synchronous control method of memristive neural network |
CN115903511B (en) * | 2022-12-08 | 2023-05-19 | 盐城工学院 | Self-adaptive index synchronous control method for random memristor neural network |
CN116430715B (en) * | 2022-12-08 | 2023-11-03 | 盐城工学院 | Finite time synchronous control method of time-varying time-delay memristor recurrent neural network |
CN115860075B (en) * | 2022-12-08 | 2023-05-30 | 盐城工学院 | Synchronous control method of fractional order memristor neural network |
CN115755621B (en) * | 2022-12-08 | 2023-05-09 | 盐城工学院 | Finite time self-adaptive synchronous control method for memristor recurrent neural network |
CN116400599B (en) * | 2023-04-07 | 2023-10-03 | 盐城工学院 | Fixed time synchronous control method of inertial CG neural network |
CN116847033B (en) * | 2023-07-03 | 2024-02-23 | 盐城工学院 | Image encryption method and system based on inertial memristor neural network desynchronization |
Family Cites Families (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN106301750B (en) * | 2015-05-18 | 2019-01-04 | 江南大学 | A kind of secret communication method based on time lag memristor chaotic neural network |
WO2016187500A1 (en) * | 2015-05-21 | 2016-11-24 | Cory Merkel | Method and apparatus for training memristive learning systems |
CN108762067A (en) * | 2018-04-28 | 2018-11-06 | 南京理工大学 | A kind of the networking Synchronizing Control Devices and acquisition methods of memristor neural network |
CN110348570B (en) * | 2019-05-30 | 2021-01-05 | 中国地质大学(武汉) | Neural network associative memory method based on memristor |
CN110879533B (en) * | 2019-12-13 | 2022-01-04 | 福州大学 | Scheduled time projection synchronization method of delay memristive neural network with unknown disturbance resistance |
-
2022
- 2022-07-28 CN CN202210911743.9A patent/CN115145156B/en active Active
Also Published As
Publication number | Publication date |
---|---|
CN115145156A (en) | 2022-10-04 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN115145156B (en) | Self-adaptive anti-synchronization method of inertial memristor neural network | |
Jia et al. | Global stabilization of fractional-order memristor-based neural networks with time delay | |
Mu et al. | Cooperative differential game-based optimal control and its application to power systems | |
CN115169539B (en) | Secret communication method based on inertial complex value memristor neural network | |
Zhang et al. | Distributed adaptive tracking synchronization for coupled reaction–diffusion neural network | |
CN115563633B (en) | Secret communication method based on inertial memristor competition neural network | |
CN107086916B (en) | Fractional order adaptive sliding mode control-based chaotic system synchronization method | |
Luo et al. | A laguerre neural network-based ADP learning scheme with its application to tracking control in the Internet of Things | |
Qureshi et al. | Push-SAGA: A decentralized stochastic algorithm with variance reduction over directed graphs | |
Zhao et al. | Adaptive fuzzy dynamic surface control with prespecified tracking performance for a class of nonlinear systems | |
CN116847033B (en) | Image encryption method and system based on inertial memristor neural network desynchronization | |
CN115903470A (en) | Hysteresis synchronous control method of inertia complex value memristor neural network | |
CN115755621A (en) | Finite time self-adaptive synchronous control method of memristor recurrent neural network | |
Lu et al. | Network-based fuzzy H∞ controller design for TS fuzzy systems via a new event-triggered communication scheme | |
Liu et al. | Fractional-order echo state network backstepping control of fractional-order nonlinear systems | |
CN115860075A (en) | Synchronous control method of fractional order memristor neural network | |
Song | Robust initialization of a Jordan network with recurrent constrained learning | |
Qian et al. | Distributed dynamic event-triggered control of power buffers in DC microgrids | |
CN116760933A (en) | Image encryption method and system based on neural network with reactive diffusion | |
Ming et al. | Mixed H 2/H∞ control for nonlinear closed-loop Stackelberg games with application to power systems | |
Xia et al. | Adaptive quantized output feedback DSC of uncertain systems with output constraints and unmodeled dynamics based on reduced-order K-filters | |
CN109581873A (en) | The finite time specified performance control algolithm of unknown actuator dead zone switching system | |
CN117335957A (en) | Secret communication method for BAM memristor neural network correction function projection synchronization | |
Kao et al. | Design of an adaptive self-organizing fuzzy neural network controller for uncertain nonlinear chaotic systems | |
Lin et al. | Based on interval type-2 adaptive fuzzy H∞ tracking controller for SISO time-delay nonlinear systems |
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 | ||
TR01 | Transfer of patent right | ||
TR01 | Transfer of patent right |
Effective date of registration: 20240319 Address after: Building A 2081, No. 88 Jianghai West Road, Liangxi District, Wuxi City, Jiangsu Province, 214063 Patentee after: Wuxi Xiangyuan Information Technology Co.,Ltd. Country or region after: Zhong Guo Address before: 224051 middle road of hope Avenue, Yancheng City, Jiangsu Province, No. 1 Patentee before: YANCHENG INSTITUTE OF TECHNOLOGY Country or region before: Zhong Guo |