Hybrid time-delay inertial memristive neural network reachable set estimation method and device
By constructing a preset inequality group and an adaptive controller, the reachable set of the hybrid time-delay inertia-memristor neural network is determined, which solves the problem of complex network dynamic behavior and improves the estimation accuracy and system stability.
Patent Information
- Application Number
- CN202510731908.8
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-06-03
- Publication Date
- 2025-09-26
AI Technical Summary
Existing technologies find it difficult to effectively solve the reachable set estimation problem of hybrid time-delay inertial memristor neural networks, especially in the presence of discrete time-varying delays and distributed time-varying delays, which leads to complex network dynamic behavior and affects system stability and reliability.
By constructing a preset group of inequalities and solving the target reachable set constant, the adaptive controller and Lyapunov-Krasovsky functional are combined to verify whether the state of the hybrid time-delay inertial memristor neural network is within the target reachable set. The adaptive controller is constructed using the preset first and second control gains to ensure that the network state is within the reachable set.
The accuracy of the reachable set of the hybrid time-delay inertial memristor neural network is improved, the problems of increased system dimension and increased computational complexity due to order reduction processing are avoided, and the stability and reliability of the system are improved.
Smart Images
Figure CN120706472A_ABST
Abstract
Description
Technical Field
[0001] The present application relates to the field of artificial intelligence, and in particular to a method and device for estimating the reachable set of a hybrid time-delay inertial memristor neural network, an electronic device, and a computer-readable storage medium. Background Art
[0002] Memristor neural networks, a neural network model that incorporates the unique physical properties of memristors, represent a new generation of computing architectures. They not only overcome the bottlenecks of the von Neumann architecture but also, with their powerful nonlinear computing capabilities and dynamic adaptability, become key to intelligent hardware acceleration. Memristor neural networks have demonstrated significant advantages in fields such as neuromorphic computing, image recognition, and natural language processing, and play a vital role in applications such as autonomous driving, robotic control, and medical diagnosis. However, as memristor neural networks continue to gain traction in applications, their nonlinear dynamic behavior and strong coupling have posed unprecedented challenges, particularly in ensuring system stability and reliability. To better apply memristor neural networks in practical engineering, studying their dynamic behavior and the problem of reaching set estimation has become a critical issue that needs to be addressed.
[0003] The model of a memristive neural network (IMNN) that incorporates an inertial term can be formulated as a second-order differential equation with neuronal states. Compared to first-order memristive neural networks, memristive neural networks with inertial terms have significant advantages, such as large storage capacity and strong fault tolerance. Therefore, research on the dynamic characteristics of inertial memristive neural networks (IMNNs) has attracted widespread attention. In practical applications, IMNNs are often affected by both discrete and distributed time-varying delays. This mixed delay complicates the network's dynamic behavior. Discrete time-varying delay typically manifests as a time lag between system states at different time points, while distributed delay refers to the time delay caused by various factors when signals are transmitted from one neuron to another in a neural network. Distributed delay considers the distribution of time delays over a certain time interval and is more consistent with the complex characteristics of signal transmission in real biological neural networks. The combined effect of these two delays can significantly alter network stability. Therefore, studying the reachable set estimation problem for IMNNs with mixed delays is crucial for understanding their dynamic characteristics and designing effective control systems. Summary of the Invention
[0004] In view of this, it is necessary to provide a method and device for estimating the reachable set of a hybrid time-delay inertial memristor neural network, an electronic device and a computer-readable storage medium to determine the reachable set of the hybrid time-delay inertial memristor neural network.
[0005] In order to achieve the above-mentioned purpose of determining the reachable set of a hybrid time-delay inertial memristor neural network, the present application provides a reachable set estimation method for a hybrid time-delay inertial memristor neural network, including: obtaining a target inequality group based on a preset inequality group and a target hybrid time-delay inertial memristor neural network, the preset inequalities including discrete time-varying delay parameters and distributed time-varying delay parameters of the hybrid time-delay inertial memristor neural network; solving a target reachable set constant that satisfies the target inequality group, and constructing a target reachable set based on the target reachable set constant; wherein, under a preset adaptive controller, the hybrid time-delay inertial memristor neural network is defined within the target reachable set.
[0006] In an optional embodiment, the preset adaptive controller includes a preset first control gain and a preset second control gain, and the hybrid time-delay inertia memristor neural network reachable set estimation method further includes: constructing a Lyapunov-Krasovsky functional according to the target reachable set constant, the preset first control gain, and the preset second control gain, wherein the Lyapunov-Krasovsky functional includes a discrete time-varying delay parameter and a distributed time-varying delay parameter of the hybrid time-delay inertial memristor neural network; Verify whether the state of the hybrid time-delay inertia memristor neural network is within the target reachable set under the preset adaptive controller according to the Lyapunov-Krasovsky functional.
[0007] In an optional embodiment, verifying, according to the Lyapunov-Krasovsky functional, whether the hybrid time-delay inertia memristor neural network is defined within the target reachable set under the preset adaptive controller includes: calculating derivatives of the Lyapunov-Krasovsky functional transformation; converting the derivative form of the Lyapunov-Krasovsky functional into a target inequality according to the adaptive controller, wherein one side of the target inequality satisfies a preset format; Determining whether the Lyapunov-Krasovsky functional is less than a preset constant threshold value based on the preset inequality group; If the Lyapunov-Krasovsky functional is less than the preset constant threshold, it is determined that the hybrid time-delay inertia memristor neural network is defined within the target reachable set.
[0008] In an optional embodiment, the Lyapunov-Krasovsky functional includes: ; in is the reachable set constant, The hybrid time-delay inertia memristor neural network neurons in The state of the moment, is the discrete time-varying delay of the hybrid time-delay inertial memristor neural network, is the distributed time-varying delay of the hybrid time-delay inertial memristor neural network, is the time-varying lag parameter of the distribution, and , , , , , , , , , , , is a constant, is the first control gain, is the second control gain, and is a positive constant, is the activation function.
[0009] In an optional embodiment, the preset inequality group includes: , , ; ; in is the reachable set constant, is a constant, , is the self-feedback coefficient of the hybrid time-delay inertial memristor neural network, and is the discrete time-varying lag parameter, and is the time-varying lag parameter of the distribution, and , , , , is the discrete time-varying delay of the hybrid time-delay inertial memristor neural network, is the distributed time-varying delay of the hybrid time-delay inertial memristor neural network, , , , , , , , , , , is a constant, is the first control gain, is the second control gain, and Is a positive number.
[0010] In an optional embodiment, before solving the target reachable set constant that satisfies the target inequality group, the hybrid time-delay inertia memristor neural network reachable set estimation method further includes: Determining whether an activation function of the hybrid time-delay inertia-memristor neural network is bounded and differentiable; If the activation function of the hybrid time-delay inertia-memristor neural network is bounded and differentiable, the step of solving a target reachable set constant that satisfies the target inequality group is performed.
[0011] In a second aspect, an embodiment of the present application provides a hybrid time-delay inertia memristor neural network reachable set estimation device, comprising: A solution module, the solution module being configured to obtain a target inequality group based on a preset inequality group and a target hybrid time-delay inertial memristor neural network, and solve a target reachable set constant that satisfies the target inequality group, wherein the preset inequality includes a discrete time-varying delay parameter and a distributed time-varying delay parameter of the hybrid time-delay inertial memristor neural network; A reachable set construction module, wherein the reachable set construction module is used to construct a target reachable set according to the target reachable set constant; a controller building module, the controller building module being configured to build a target adaptive controller and a preset adaptive controller according to the target first control gain and the target second control gain; Under a preset adaptive controller, the hybrid time-delay inertia memristor neural network is defined within the target reachable set.
[0012] In an optional embodiment, the preset adaptive controller includes a preset first control gain and a preset second control gain, and the hybrid time-delay inertia memristor neural network reachable set estimation device further includes: A functional construction module, wherein the functional construction module is used to construct a Lyapunov-Krasovsky functional according to the target reachable set constant, the preset first control gain, and the preset second control gain, wherein the Lyapunov-Krasovsky functional includes a discrete time-varying delay parameter and a distributed time-varying delay parameter of a hybrid delay-inertial-memristor neural network; A verification module is used to verify whether the state of the hybrid time-delay inertia memristor neural network is defined within the target reachable set under the preset adaptive controller based on the Lyapunov-Krasovsky functional.
[0013] In a third aspect, an embodiment of the present application further provides an electronic device comprising a memory and a processor, wherein the memory is used to store a program; and the processor is coupled to the memory and is used to execute the program stored in the memory to implement the aforementioned hybrid time-delay inertial memristor neural network reachable set estimation method.
[0014] In a fourth aspect, an embodiment of the present application further provides a computer-readable storage medium for storing computer-readable programs or instructions, which, when executed by a processor, can implement the aforementioned hybrid time-delay inertial memristor neural network reachable set estimation method.
[0015] Beneficial effects of this application: In the present application, by solving the target reachable set constant that satisfies the target inequality group, since the preset inequality group includes the discrete time-varying delay parameters and distributed time-varying delay parameters of the hybrid time-delay inertial memristor neural network, that is, the solved target reachable set constant is affected by both the discrete time-varying delay characteristics and the distributed time-varying delay characteristics, the target reachable set subsequently constructed according to the target reachable set constant is related to the discrete time-varying delay characteristics and the distributed time-varying delay characteristics of the hybrid time-delay inertial memristor neural network, thereby improving the accuracy of the determined reachable set of the hybrid time-delay inertial memristor neural network; in addition, for the second-order inertial memristor neural network, there is no need to reduce its order when determining its reachable set in the present application, so there will be no problem of increased system dimension and increased computational complexity due to order reduction. BRIEF DESCRIPTION OF THE DRAWINGS
[0016] FIG1 is a flow chart of an embodiment of a method for estimating reachable sets of a hybrid time-delay inertial memristor neural network provided by the present application; Figure 2 A method flow chart of another embodiment of the reachable set estimation method for a hybrid time-delay inertial memristor neural network provided by the present application; Figure 3 A method flow chart of another embodiment of the reachable set estimation method of the hybrid time-delay inertial memristor neural network provided by the present application; Figure 4 This is a structural diagram of an embodiment of a hybrid time-delay inertial memristor neural network reachable set estimation device provided by the present application; Figure 5 This is a structural diagram of another embodiment of the hybrid time-delay inertial memristor neural network reachable set estimation device provided by the present application; Figure 6 This is a schematic structural diagram of an embodiment of an electronic device provided in this application. DETAILED DESCRIPTION
[0017] The preferred embodiments of the present application are described in detail below in conjunction with the accompanying drawings, wherein the accompanying drawings constitute a part of the present application and are used together with the embodiments of the present application to illustrate the principles of the present application, and are not used to limit the scope of the present application.
[0018] It should be noted that, in the following embodiments of the present application, , , , , , For ,satisfy , ,but express convex combination of . represents the set of all continuous functions in a Banach space. Represents the identity matrix.
[0019] Furthermore, in the embodiment of the present application, the following Lemma 1 is also pre-set: Assume is a non-negative function that satisfies If it exists , Make ,but .
[0020] Please refer to Figure 1 A specific embodiment of the present application discloses a reachable set estimation method for a hybrid time-delay inertial memristor neural network, comprising the following steps: Step S101: Based on a preset inequality group and a target hybrid time-delay inertia memristor neural network, a target inequality group is obtained.
[0021] Step S102: solving a target reachable set constant that satisfies a target inequality group, where the preset inequalities include a discrete time-varying delay parameter and a distributed time-varying delay parameter of the hybrid delay-inertial-memristor neural network.
[0022] Step S103: Construct a target reachable set according to the target reachable set constant.
[0023] In the present application, by solving the target reachable set constant that satisfies the target inequality group, since the preset inequality group includes the discrete time-varying delay parameters and distributed time-varying delay parameters of the hybrid time-delay inertial memristor neural network, that is, the solved target reachable set constant is affected by both the discrete time-varying delay characteristics and the distributed time-varying delay characteristics, the target reachable set subsequently constructed according to the target reachable set constant is related to the discrete time-varying delay characteristics and the distributed time-varying delay characteristics of the hybrid time-delay inertial memristor neural network, thereby improving the accuracy of the determined reachable set of the hybrid time-delay inertial memristor neural network; in addition, for the second-order inertial memristor neural network, there is no need to reduce its order when determining its reachable set in the present application, so there will be no problem of increased system dimension and increased computational complexity due to order reduction.
[0024] In step S101, the preset inequality group is a pre-set inequality group including multiple variables. In step S101, the specific relevant parameters of the target hybrid time-delay inertia memristor neural network can be substituted into the preset inequality group to obtain a target inequality group corresponding to the target hybrid time-delay inertia memristor neural network. Specifically, the preset inequality group includes multiple parameters related to the hybrid time-delay inertia memristor neural network and the adaptive controller. To better illustrate the preset inequality group, the hybrid time-delay inertia memristor neural network and the adaptive controller are preferentially described in this embodiment.
[0025] For a hybrid time-delay inertial memristor neural network, in one embodiment of the present application, a hybrid time-delay inertial memristor neural network including discrete time-varying delay and distributed time-varying delay can be specifically expressed by the following formula (1): (1) in Indicates the neurons in The state of the moment, , is the self-feedback coefficient, is the neuron activation function. and represent discrete time-varying lag and distributed time-varying lag respectively, and satisfy , , , , . represents a bounded external disturbance, satisfy , is a constant greater than zero. Represents a controller. , and is the memristor connection weight, To switch the variable, the volt-ampere characteristic of the memristor changes and the following conditions are met:
[0026] in , , , , , , , Is a constant, switching threshold .
[0027] Due to the switching characteristics of the memristor connection weight, the model equation represented by equation (1) is a differential equation with discontinuity on the right side. , , , , , , , , so that the solutions of the above formula (1) are all solutions in the sense of Filippov.
[0028] For the initial conditions , , .in , Hybrid time-delay inertia memristor neural network. Applying differential inclusion theory and set-valued mapping theory, the above formula (1) can be written as the following formula (2): (2) in
[0029] On this basis, according to the theory of measurable selection, there exists a measurable function , , the above formula (2) can be further expressed as the following formula (3): (3) The purpose of determining the reachable set is to determine a set as small as possible so that the state of the hybrid time-delay inertia-memristor neural network expressed in equation (1) above is confined within this set. Therefore, the reachable set of the hybrid time-delay inertia-memristor neural network is defined as follows: .
[0030] Set the reachable set constant , in the above formula The polygon set definition can be expressed as: .
[0031] For the adaptive controller, in the embodiment of the present application, a preset adaptive controller represented by the following equations (4) and (5) may be pre-set: (4) (5) in, is the first control gain, is the second control gain, and Is a positive number.
[0032] Based on the above formulas 1-5, the preset inequality group provided in this embodiment may specifically include: , , ; ; in, , is the self-feedback coefficient of the hybrid time-delay inertial memristor neural network, is a constant.
[0033] Based on the above preset inequality group, the target inequality group can be obtained by substituting the relevant data of the target hybrid time-delay inertia memristor neural network into the above preset inequality group. , , By solving the target inequality, the reachable set constant, the first control gain and the second control gain that satisfy the target inequality group can be obtained as the target first control gain and the target second control gain.
[0034] In step S103, the target reachable set is constructed according to the target reachable set constant. Specifically, the target reachable set constant obtained by the solution is substituted into the formula: The target reachable set is obtained.
[0035] For the preset adaptive controller and the target reachable set, under the preset adaptive controller, the hybrid time-delay inertia memristor neural network is defined within the target reachable set.
[0036] For further information, please refer to Figure 2 Before step S102 of the present application, the following steps may also be included: Step S104: Determine whether the activation function of the hybrid time-delay inertia-memristor neural network is bounded and differentiable. If so, execute step S102; if not, end the process.
[0037] Among them, if the activation function is bounded and differentiable, then there exists a positive constant , So that for any satisfy , and satisfies , In step S104, it can be determined based on this characteristic whether the activation function of the hybrid time-delay inertia memristor neural network is bounded and differentiable.
[0038] For further information, please refer to Figure 3 In some embodiments of the present application, in addition to the steps of determining the reachable set and the adaptive controller in steps S101 to S103, the following verification steps may also be included, specifically including: Step S105: constructing a Lyapunov-Krasovsky functional according to the target reachable set constant, the target first control gain, and the target second control gain.
[0039] In this step, the Lyapunov-Krasovskii functional can be specifically expressed as the following formula (6), including: (6) in, is the reachable set constant, The hybrid time-delay inertia memristor neural network neurons in The state of the moment, is the discrete time-varying lag parameter, is the distributed time-varying lag parameter, and , , , , , , , , , , , is a constant, is the first control gain, is the second control gain, and is a positive constant, is the activation function.
[0040] Step S107: Verify whether the hybrid time-delay inertia-memristor neural network is within the target reachable set under the preset adaptive controller according to the Lyapunov-Krasovsky functional.
[0041] In this step, for the Lyapunov-Krasovsky functional shown in Equation (6) above, combined with Equation (1), the Lyapunov-Krasovsky functional is converted into the derivative form shown in Equation (7) below, which is specifically expressed as: (7) in, , combined with the above formula (7), we can get the following inequality (8), (8) The above formula (8) is combined with adaptive control The following formula (9) can be obtained (9) The above formula (9) can be further sorted out to get the following formula (10): (10) Based on the above inequality (10), combined with the preset inequality group, we can know that: , Further combining the above Theorem 1, we can get , where 1 is the preset constant threshold, and whether the Lyapunov-Krasovsky functional is less than the preset constant threshold is determined to determine whether , if satisfied, it is determined that the hybrid time-delay inertia memristor neural network is within the target reachable set.
[0042] Please refer to Figure 4 , an embodiment of the present application also provides a hybrid time-delay inertial memristor neural network reachable set estimation device, including: a solution module 100, the solution module 100 is used to obtain a target inequality group based on a preset inequality group and a target hybrid time-delay inertial memristor neural network, and solve a target reachable set constant that satisfies the target inequality group, the preset inequalities include discrete time-varying delay parameters and distributed time-varying delay parameters of the hybrid time-delay inertial memristor neural network; a reachable set construction module 200, the reachable set construction module 200 is used to construct a target reachable set according to the target reachable set constant; under a preset adaptive controller, the hybrid time-delay inertial memristor neural network is defined within the target reachable set.
[0043] For further information, please refer to Figure 5In some embodiments of the present application, the hybrid time-delay inertial memristor neural network reachable set estimation device may further include a functional construction module 300, the preset adaptive controller includes a preset first control gain and a preset second control gain, the functional construction module 300 is used to construct a Lyapunov-Krasovsky functional according to the target reachable set constant, the preset first control gain and the preset second control gain, the Lyapunov-Krasovsky functional includes a discrete time-varying delay parameter and a distributed time-varying delay parameter of the hybrid time-delay inertial memristor neural network; a verification module 400, the verification module 400 is used to verify whether the hybrid time-delay inertial memristor neural network is defined within the target reachable set under the preset adaptive controller according to the Lyapunov-Krasovsky functional.
[0044] Please refer to Figure 6 The present invention also provides an electronic device. The electronic device includes a processor 701 and a memory 702. Figure 6 Only some of the components of the electronic device are shown, but it should be understood that it is not required to implement all of the shown components, and more or fewer components may be implemented instead.
[0045] In some embodiments, the processor 701 may be a central processing unit (CPU), a microprocessor, or other data processing chip, configured to execute program codes or process data stored in the memory 702 , such as the magnetic resonance image optimization method of the present invention.
[0046] In some embodiments, processor 701 may be a single server or a server group. The server group may be centralized or distributed. In some embodiments, processor 701 may be local or remote. In some embodiments, processor 701 may be implemented on a cloud platform. In one embodiment, the cloud platform may include a private cloud, a public cloud, a hybrid cloud, a community cloud, a distributed cloud, an on-premises cloud, a multi-cloud, or any combination thereof.
[0047] In some embodiments, the memory 702 may be an internal storage unit of the electronic device, such as a hard disk or memory of the electronic device. In other embodiments, the memory 702 may also be an external storage device of the electronic device, such as a plug-in hard disk, a smart media card (SMC), a secure digital (SD) card, a flash card, etc. equipped on the electronic device.
[0048] Furthermore, the memory 702 may include both an internal storage unit of the electronic device and an external storage device. The memory 702 is used to store application software installed in the electronic device and various data.
[0049] Furthermore, the embodiments of the present invention do not specifically limit the types of electronic devices mentioned, and the electronic devices may be portable electronic devices such as mobile phones, tablet computers, personal digital assistants (PDAs), wearable devices, and laptop computers. Exemplary embodiments of portable electronic devices include but are not limited to portable electronic devices equipped with iOS, Android, Microsoft, or other operating systems. The above-mentioned portable electronic devices may also be other portable electronic devices, such as laptop computers with touch-sensitive surfaces (e.g., touch panels). It should also be understood that in some other embodiments of the present invention, the electronic device may not be a portable electronic device, but a desktop computer with a touch-sensitive surface (e.g., a touch panel).
[0050] Accordingly, an embodiment of the present application also provides a computer-readable storage medium, which is used to store computer-readable programs or instructions. When the program or instructions are executed by a processor, the steps or functions in the magnetic resonance image optimization method provided by the above-mentioned method embodiments can be implemented.
[0051] Those skilled in the art will appreciate that all or part of the process flow of the above-described method embodiment can be implemented by instructing related hardware (such as a processor, controller, etc.) through a computer program, and the computer program can be stored in a computer-readable storage medium. The computer-readable storage medium may be a magnetic disk, an optical disk, a read-only memory, or a random access memory.
Claims
1. A reachable set estimation method for a hybrid time-delay inertial memristor neural network, characterized in that: include: Based on a preset inequality group and a target hybrid time-delay inertial memristor neural network, a target inequality group is obtained, wherein the preset inequality includes a discrete time-varying delay parameter and a distributed time-varying delay parameter of the hybrid time-delay inertial memristor neural network; Solving a target reachable set constant that satisfies the target inequality group, and constructing a target reachable set according to the target reachable set constant; Wherein, under the preset adaptive controller, the hybrid time-delay inertia memristor neural network is defined within the target reachable set.
2. The reachable set estimation method of a hybrid time-delay inertial memristor neural network according to claim 1, characterized in that: The preset adaptive controller includes a preset first control gain and a preset second control gain, and the hybrid time-delay inertia memristor neural network reachable set estimation method further includes: constructing a Lyapunov-Krasovsky functional according to the target reachable set constant, the preset first control gain, and the preset second control gain, wherein the Lyapunov-Krasovsky functional includes a discrete time-varying delay parameter and a distributed time-varying delay parameter of the hybrid time-delay inertial memristor neural network; Verify whether the state of the hybrid time-delay inertia memristor neural network is within the target reachable set under the preset adaptive controller according to the Lyapunov-Krasovsky functional.
3. The reachable set estimation method of a hybrid time-delay inertial memristor neural network according to claim 2, characterized in that: Verifying, based on the Lyapunov-Krasovsky functional, whether the hybrid time-delay inertia-memristor neural network is within the target reachable set under the preset adaptive controller includes: calculating derivatives of the Lyapunov-Krasovsky functional transformation; converting the derivative form of the Lyapunov-Krasovsky functional into a target inequality according to the adaptive controller, wherein one side of the target inequality satisfies a preset format; Determining whether the Lyapunov-Krasovsky functional is less than a preset constant threshold value based on the preset inequality group; If the Lyapunov-Krasovsky functional is less than the preset constant threshold, it is determined that the hybrid time-delay inertia memristor neural network is within the target reachable set.
4. The reachable set estimation method of a hybrid time-delay inertial memristor neural network according to claim 2, characterized in that: The Lyapunov-Krasovsky functional includes: ； in is the reachable set constant, The hybrid time-delay inertia memristor neural network neurons in The state of the moment, is the discrete time-varying delay of the hybrid time-delay inertial memristor neural network, is the distributed time-varying delay of the hybrid time-delay inertial memristor neural network, is the time-varying lag parameter of the distribution, and , , , , , , , , , , , is a constant, is the first control gain, is the second control gain, and is a positive constant, is the activation function.
5. The reachable set estimation method of a hybrid time-delay inertial memristor neural network according to claim 1, characterized in that: The preset inequality group includes: ， ， ； ； in is the reachable set constant, is a constant, , is the self-feedback coefficient of the hybrid time-delay inertial memristor neural network, and is the discrete time-varying lag parameter, and is the time-varying lag parameter of the distribution, and , , , , is the discrete time-varying delay of the hybrid time-delay inertial memristor neural network, is the distributed time-varying delay of the hybrid time-delay inertial memristor neural network, , , , , , , , , , , is a constant, is the first control gain, is the second control gain, and Is a positive number.
6. The reachable set estimation method of a hybrid time-delay inertial memristor neural network according to claim 1, characterized in that: Before solving the target reachable set constant that satisfies the target inequality group, the hybrid time-delay inertia memristor neural network reachable set estimation method further includes: Determining whether an activation function of the hybrid time-delay inertia-memristor neural network is bounded and differentiable; If the activation function of the hybrid time-delay inertia-memristor neural network is bounded and differentiable, the step of solving a target reachable set constant that satisfies the target inequality group is performed.
7. A hybrid time-delay inertial memristor neural network reachable set estimation device, characterized in that: include: A solution module, the solution module being configured to obtain a target inequality group based on a preset inequality group and a target hybrid time-delay inertial memristor neural network, and solve a target reachable set constant that satisfies the target inequality group, wherein the preset inequality includes a discrete time-varying delay parameter and a distributed time-varying delay parameter of the hybrid time-delay inertial memristor neural network; A reachable set construction module, wherein the reachable set construction module is used to construct a target reachable set according to the target reachable set constant; a controller building module, the controller building module being configured to build a target adaptive controller and a preset adaptive controller according to the target first control gain and the target second control gain; Under a preset adaptive controller, the hybrid time-delay inertia memristor neural network is defined within the target reachable set.
8. The hybrid time-delay inertia-memristor neural network reachable set estimation device according to claim 7, characterized in that: The preset adaptive controller includes a preset first control gain and a preset second control gain, and the hybrid time-delay inertia memristor neural network reachable set estimation device also includes: A functional construction module, wherein the functional construction module is used to construct a Lyapunov-Krasovsky functional according to the target reachable set constant, the preset first control gain, and the preset second control gain, wherein the Lyapunov-Krasovsky functional includes a discrete time-varying delay parameter and a distributed time-varying delay parameter of a hybrid delay-inertial-memristor neural network; A verification module is used to verify whether the state of the hybrid time-delay inertia memristor neural network is defined within the target reachable set under the preset adaptive controller based on the Lyapunov-Krasovsky functional.
9. An electronic device, characterized in that: comprising a memory and a processor, wherein, The memory is used to store programs; The processor is coupled to the memory and is used to execute the program stored in the memory to implement the hybrid time-delay inertia memristor neural network reachable set estimation method as described in any one of claims 1 to 6.
10. A computer-readable storage medium, characterized in that Used to store computer-readable programs or instructions, which, when executed by a processor, can implement the hybrid time-delay inertia memristor neural network reachable set estimation method as described in any one of claims 1 to 6.