Multi-layer network distributed data sampling synchronization method with high-dimensional node characteristics

Abstract

The invention relates to a multi-layer network distributed data sampling synchronization method with high-dimensional node characteristics. The method comprises the following steps: S1, establishing a multi-layer mathematical model of unconstrained interlayer and intra-layer parameters; s2, based on a multi-layer mathematical model of unconstrained interlayer and intra-layer parameters, regarding a target state as a virtual node embedded network topology, and designing a data sampling controller containing virtual nodes; s3, defining a synchronization error, and establishing a multi-layer network output progressive synchronization problem with a high-dimensional node feature; and S4, based on the data sampling controller, solving the problem of non-progressive output synchronization of the multi-layer network with the high-dimensional node characteristics, establishing an output progressive synchronization criterion suitable for the multi-layer network with the high-dimensional node characteristics, and realizing carrier synchronization in the multi-layer network. Therefore, the modulation signal sent by any node can be synchronously demodulated at other nodes in the global situation.

Description

Technical Field

[0001] The present invention relates to the field of production network technology, and particularly to a multi-layer network distributed data sampling synchronization method with high-dimensional node features. Background Art

[0002] In industrial application scenarios, manufacturing systems are usually composed of geographically dispersed factories or workshops, forming a highly distributed architecture. These systems achieve the efficient execution of multiple industrial manufacturing processes and product assembly tasks through autonomous coordination and cooperation mechanisms. To effectively characterize the structural features and internal interaction relationships of industrial manufacturing systems, a single-layer complex network model has been proposed as an ideal tool and widely used to describe the topological characteristics and dynamic behaviors of industrial processes. However, industrial manufacturing systems are usually composed of multiple levels and heterogeneous components, including production lines, equipment, warehouses, and supply chains, etc. These components have significant differences in behavior patterns and functional characteristics and cannot be modeled by simple superposition. For example, when completing product manufacturing and assembly tasks, multiple subsystems such as communication networks, control networks, and power grids need to cooperate closely to efficiently achieve production goals. These networks form a complex multi-layer network system through deep integration. Therefore, the single-layer network model is difficult to meet the modeling requirements of actual manufacturing systems, and the multi-layer network theory has emerged as the times require.

[0003] With the rapid development of intelligent manufacturing technology, especially the gradual realization of high-end intelligent manufacturing production processes, traditional network layering methods are facing increasingly severe challenges in dealing with complex industrial scenarios. Existing layering methods usually assume a strict one-to-one correspondence relationship between physical domain and information domain nodes, that is, each physical object (such as a process, equipment, etc.) is mapped to a unique information node (such as a sensor, data acquisition device, etc.). However, in the actual industrial environment, this strict correspondence relationship often does not hold, and there are generally complex many-to-many or many-to-one mapping relationships between physical domain and information domain nodes. In addition, there may be significant differences in the number of nodes between different levels (such as the number of physical layer devices is much larger than the number of information layer data acquisition points), which further exacerbates the complexity of modeling. Therefore, constructing a node mapping mechanism suitable for the industrial manufacturing environment to support non-one-to-one correspondence relationships and adapt to the dynamic inconsistencies in the number of nodes at each layer has become a key scientific problem that needs to be solved urgently. Summary of the Invention

[0004] To solve the above problems, the purpose of the present invention is to provide a multi-layer network distributed data sampling synchronization method with high-dimensional node features, which effectively solves the synchronization control problem of multi-dimensional coupling in intelligent manufacturing systems.

[0005] To achieve the above purpose, the present invention adopts the following technical solutions: A distributed data sampling synchronization method for a multi-layer network with high-dimensional node features, comprising the following steps: S1: Establish a multi-layer mathematical model without constraining inter-layer and intra-layer parameters; S2: Based on the multi-layer mathematical model without constraining inter-layer and intra-layer parameters, regard the target state as a virtual node embedded in the network topology, and design a data sampling controller containing virtual nodes; S3: Define the synchronization error and establish the problem of asymptotic output synchronization of a multi-layer network with high-dimensional node features; S4: Based on the data sampling controller, solve the non-asymptotic output synchronization problem of a multi-layer network with high-dimensional node features, establish an asymptotic output synchronization criterion applicable to a multi-layer network with high-dimensional node features, and achieve carrier synchronization in the multi-layer network, so that the modulation signals sent by any node can be synchronously demodulated at other nodes within the global scope.

[0006] Further, establish a multi-layer mathematical model without constraining inter-layer and intra-layer parameters, specifically as follows: Consider an undirected weighted network composed of layers, each layer has identical node systems, and the dynamics of the th layer: (1) where q represents adjacent nodes, represents the intra-layer connection matrix, represents the inter-layer connection matrix, represents the th layer, is defined as the first derivative of , is defined as the state of the th node in the th layer, is defined as the state of the th node in the th layer, is the control gain, is the state matrix, is the input state matrix, represents the internal coupling matrix, represents the external coupling matrix, represents the output coefficient matrix, represents the th node in the th layer's input vector, represents the th node in the th layer's output vector, represents dimensional Euclidean space, denote dimensional Euclidean space denote dimensional Euclidean space denote dimensional Euclidean space denote dimensional Euclidean space; Describe the state vector of the entire multi - layer network, where \(T\) represents the transpose of the vector. is the state vector of layer Define the th node state of layer is the vector describing the output of the entire multi - layer network system, where is the output variable of layer denote the th output vector of the node in layer is the vector describing the input of the entire multi - layer network system, where is the th denote the th input vector of the node in layer The production target is a target trajectory, the output of the following autonomous system: (2) where denote the state matrix of the entire leader 0, denote the output matrix of the entire leader 0, and denote the state and output of leader 0; Regard the target system (2) as the virtual vertex of the network (1), label it as the th node, let the intermediate variable , and obtain the augmented coupling graph definition of the multi - layer network (2) as , the superscript denote layer The vertex set is , denote the th directed edge from node to node th The connection between a node and the th neighbor node ; otherwise . Obviously, , represents the connection between the th node in the th layer and the th neighbor node; represents the connection weight; and , the adjacency matrix is: ; where represents the 0 vector, represents the original network connection matrix, the vector , represents the original network Laplacian matrix, represents the connection weight of the edge . The Laplacian matrix of Figure is defined as and is expressed as: , where represents the element in the p-th row and p-th column of the matrix, represents the diagonal matrix, For is a positive definite matrix.

[0007] Furthermore, the design of the data sampling controller is as follows: The reference generator for the th layer node is (3) where represents the state of the reference generator of the th layer node , and . represents the output of the reference generator of the th layer node , and represents the output of the th layer, represents the output. represents the distributed control law of the reference generator of the th layer node , and represents the control law of the th layer, represents the overall network control law; Due to communication limitations, the sampled-data control method is adopted, and the zero-order hold maintains the value within the sampling interval at the sampling instant satisfies , where the subscript s represents the sampling node time; the sampling interval is , and the distributed controller is designed as is aperiodic sampling, and the sampling upper bound is , that is , represents the sampling interval, represents identically equal to. The data-sampled distributed controller is designed as: (4) where represents the sampling interval, represents the control gain matrix, represents the number of nodes, represents the number of layers, represents the state of the th node in layer represents the state of the th node in layer

[0008] Based on the sampled-data distributed control law in (4), the dynamics of the reference generator in (3) are rewritten as: (5).

[0009] Furthermore, the synchronization error is defined, and the problem of asymptotic synchronization of the output of a multi-layer network with high-dimensional node characteristics is established, as follows: Define the error vector : (6) where represents the number of nodes, represents the number of layers, represents the error vector, and represents the error vector of layer Define the error vector of the th node in layer ; It can be calculated that: ; (7) where represents the state of leader 0, and ;

[0010] where, represents the The th node between layers is connected to the th node between layers, indicating that the th node between layers is connected to the th node between layers; therefore, the th node between layers is connected to the th node between layers; thus, (8) where represents the p-dimensional identity matrix, represents the direct sum of the off-diagonal elements between layers, represents the diagonal elements between layers, represents the number of layers, represents the loop multiplication, Furthermore, calculate: (9) where represents -dimensional identity matrix, represents the Laplacian matrix within the layer, represents the direct sum of matrices, represents the addition of two matrices, represents the loop multiplication, represents a unit vector with the th component being 1 and the rest being 0.

[0011] Furthermore, S4 is as follows: Parameterize the of data sampling distributed fixed-time control as , where is the solution of an algebraic Riccati inequality, and select the scalar as a positive scalar; if there exist positive definite matrices and such that:

[0012] holds, h represents the upper bound of aperiodic sampling, represents an intermediate variable, then and are stable, where represents an intermediate variable.

[0013] Proof: Calculate (10) where represents the transfer function, represents the identity matrix; Therefore, the closed-loop system is represented by the feedback interconnection of and , where represents an operator; Define the transfer function as: (11) represents the input, represents the output, represents the input signal of the system, and the time-delay integral operator is: (12) The arrow represents the transformation or mapping from to ; represents the integration of the output from time to time ; because:

[0014] I represents the identity matrix, represents taking the 2-norm, and select the Lyapunov function as: , Therefore, the derivative of is: where , .

[0015] Therefore, we get: (14) where represents an intermediate variable: ; If , we can obtain ; by the Schur lemma, if and only if the following inequality holds: , Multiply the above matrix by the diagonal block matrix to get: , Let , 。

[0016] Therefore, if there exists a positive definite matrix and such that , then holds. According to Lyapunov stability theorem, the stability of the closed-loop system is guaranteed.

[0017] The present invention has the following beneficial effects: 1. Based on the multi-layer network output progressive synchronization method of distributed data sampling control, the present invention establishes a new multi-layer network model without constrained inter-layer and intra-layer parameters, enabling nodes in all layers to have inconsistent characteristics and structures to meet the diverse modeling requirements of complex industrial applications; 2. The designed target trajectory of the present invention is the output of an autonomous linear time-invariant system, which is used as a virtual node to enhance the original network, eliminating the topological connectivity constraint, reducing the conservatism of the synchronization condition, and designing a new multi-layer network (MCNS) sampled-data control strategy for industrial applications, overcoming the assumption that traditional physical domain nodes and information domain nodes must correspond one-to-one. For nodes with different time scales in MCNS, the controller performs distributed calculations based on local information. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] Figure 1 is the principle block diagram of the control method of the present invention; Figure 2 is the topological graph of the network in an embodiment of the present invention; Figure 3 is the synchronous demodulation architecture diagram in an embodiment of the present invention; Figure 4 is the signal modulation and demodulation framework diagram based on the carrier synchronization network in an embodiment of the present invention; Figure 5 is the state diagram of the leader and the follower in an embodiment of the present invention; Figure 6 is the signal modulation diagram in an embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0019] The following further describes the present invention in detail with reference to the drawings and specific embodiments: Referring to Figure 1 , in this embodiment, a multi-layer network distributed data sampling synchronization method with high-dimensional node features is provided, including the following steps: S1: Establish a multi-layer mathematical model without constrained inter-layer and intra-layer parameters; S2: Based on the multi-layer mathematical model without constrained inter-layer and intra-layer parameters, regard the target state as a virtual node and embed it into the network topology, and design a data sampling controller containing the virtual node; S3: Define the synchronization error and establish the problem of asymptotic synchronization of the output of a multi-layer network with high-dimensional node features; S4: Based on the data sampling controller, solve the problem of non-asymptotic output synchronization of a multi-layer network with high-dimensional node features, establish an asymptotic output synchronization criterion applicable to a multi-layer network with high-dimensional node features, and achieve carrier synchronization in the multi-layer network, so that the modulation signals sent by any node can be synchronously demodulated at other nodes within the global scope.

[0020] In this embodiment, a multi-layer mathematical model without constraints on inter-layer and intra-layer parameters is established as follows: Consider an undirected weighted network consisting of layers, with identical node systems in each layer. The dynamics of the th layer: (1) where q represents adjacent nodes, represents the intra-layer connection matrix, represents the inter-layer connection matrix, represents the th layer, is defined as the first derivative of, is defined as the state of the th node in the th layer, is defined as the state of the th node in the th layer, is the control gain, is the state matrix, is the input state matrix, represents the internal coupling matrix, represents the external coupling matrix, represents the output coefficient matrix, represents the th layer node input vector, represents the th layer node output vector, represents dimensional Euclidean space, represents dimensional Euclidean space, represents dimensional Euclidean space, represents dimensional Euclidean space, represents dimensional Euclidean space; The state vector describing the entire multi - layer network (T represents the transpose of the vector), is the state vector of layer which is defined as the state of the -th node in layer ; is the vector describing the output of the entire multi - layer network system, where is the output variable of layer which represents the output vector of the -th node in layer ; is the vector describing the input of the entire multi - layer network system, where is the input vector of layer which is the control input of the -th node in layer and represents the input vector of the -th node in layer The production target is a target trajectory, which is the output of the following autonomous system: (2) where represents the state matrix of the entire leader 0, represents the output matrix of the entire leader 0, and represent the state and output of leader 0; Regarding the target system (2) as a virtual vertex of the network (1), marked as the -th node, let the intermediate variable , and the augmented coupling graph of the multi - layer network (2) is defined as , the superscript represents layer , the vertex set is , and the directed edge set is , represents the directed edge from node to node if and only if then there is a connection between the -th node and the -th neighbor node in layer , otherwise . Obviously, . , represents the connection between the -th node and the -th node in layer Connect between neighbor nodes; Indicates the connection weight; ( Indicates the th layer, the th node and the th neighbor node connection) and , the adjacency matrix is: ; Where Indicates the 0 vector, Indicates the original network connection matrix, the vector , Indicates the original network Laplacian matrix, Indicates the edge connection weight. The Laplacian matrix of Figure is defined as Expressed as: , Where Indicates the element in the p-th row and p-th column of the matrix, Indicates the diagonal matrix, For is a positive definite matrix.

[0021] In this embodiment, the design of the data sampling controller is as follows: The designed th layer node reference generator is (3) Where Indicates the th layer node reference generator state, and . Indicates the th layer node reference generator output, and Indicates the th layer output, Indicates the output. Indicates the th layer node reference generator's distributed control law, and Indicates the th layer control law, Indicates the entire network control law; Due to communication limitations, a sampled-data control method is adopted. The zero-order hold maintains the value within the sampling interval, and the sampling instant satisfies , where the subscript s represents the sampling node time; the sampling interval is , and the distributed controller is designed as is aperiodic sampling, and the sampling upper bound is , that is , represents the sampling interval, represents identically equal to. The data sampling distributed controller is designed as: (4) where represents the sampling interval, represents the control gain matrix, represents the number of nodes, represents the number of layers, represents the state of the -th node in the layer, represents the state of the

[0022] Based on the sampling data distributed control law in (4), the dynamics of the reference generator in (3) is rewritten as: (5).

[0023] In this embodiment, the synchronization error is defined, and the problem of output asymptotic synchronization of a multi - layer network with high - dimensional node features is established as follows: Define the error vector : (6) where represents the number of nodes, represents the number of layers, represents the error vector, and represents the error vector of the -th layer, and define the error vector of the -th node in the ; (7) where represents the state of the leader 0, and ;

[0024] where, represents the -th node between the -th layer and the Connected to a node Indicating the The th node between layers is connected to the The th node between layers. Therefore, (8) Among them, Indicates a p-dimensional identity matrix, Indicates the direct sum of off-diagonal elements between layers, Indicates the diagonal elements between layers, Indicates the number of layers, Indicates the circle multiplication, Furthermore, calculate: (9) Among them Indicates A -dimensional identity matrix, Indicates the Laplacian matrix within a layer, Indicates the direct sum of matrices, Indicates the addition of two matrices, Indicates a th unit vector with the

[0025] In this embodiment, S4 is as follows: Parameterize the data sampling distributed fixed-time control as , where is the solution of an algebraic Riccati inequality, and select the scalar as a positive scalar; if there exist positive definite matrices and such that:

[0026] holds, h represents the upper bound of aperiodic sampling, Indicates an intermediate variable, then and The link of is stable, where

[0027] Proof: Calculate (10) Among them Indicates the transfer function, Indicates the identity matrix; Therefore, the closed-loop system consists of and represented by the feedback interconnection, represent the operator; define the transfer function as: (11) represent the input, represent the output, represent the input signal of the system, and the time-delay integral operator is: (12) The arrow represents the conversion or mapping from to ; represents the integral of the output from time to time . Because:

[0028] I represents the identity matrix, represents taking the 2-norm, and select the Lyapunov function as: , Therefore, the derivative of (13) where , .

[0029] Therefore, we get: (14) where represents the intermediate variable ; If , we can get ; by the Schur lemma, if and only if the following inequality holds: , Multiply the above matrix by the diagonal block matrix to get: , Let , .

[0030] Therefore, if there exist positive definite matrices and such that , then holds, and according to Lyapunov stability theorem, the stability of the closed-loop system is guaranteed. Embodiment:

[0031] To introduce the present invention in detail, a specific example is given below to illustrate the proposed multi-layer network synchronization control method with high-dimensional node features.

[0032] In the intelligent production process, the detection and transmission of the state of mechanical equipment are important and indispensable links in the production monitoring system. Since analog signals usually encounter noise interference during transmission, which may lead to signal distortion, in order to ensure the reliable transmission of physical layer signals, the analog signal sender usually modulates the signal before transmission, and the receiver demodulates it.

[0033] During the demodulation process, there are usually three techniques: synchronous demodulation, envelope detection, and phase-sensitive detection. Since synchronous demodulation has a higher signal-to-noise ratio (SNR), better anti-noise ability, and better spectral efficiency, it can more accurately reproduce the original signal and is more suitable for signal transmission in the case of spectrum hole changes. This makes synchronous demodulation widely used in signal transmission in intelligent industrial sites and 5G communications.

[0034] A multi-layer network mathematical model is established for the sensor network, field debugging network, and central control network in the production manufacturing network. Aiming at the problem of data (signal) transmission and the difficulty of carrier synchronization in the modulation and demodulation process, the proposed carrier synchronization scheme can quickly construct a globally synchronized carrier signal, providing a high-precision data transmission method for the production manufacturing process, and then improving the collaborative ability of multi-level devices in the intelligent manufacturing system. Signal conditioning and conversion are indispensable links in the production manufacturing process, and signal modulation and demodulation are important processes in signal transmission. During the modulation and demodulation process, synchronous demodulation can better retain amplitude and phase information. However, due to the characteristics of difficult generation and easy deviation of synchronous signals, it is difficult to be widely used in industrial production processes. To solve this problem, the present invention aims to design a synchronous carrier generation device applicable to any signal receiver (provided that the receiver is stable). The device consists of three parts, including a carrier generation node (synchronous state), a carrier sampler (RG), and a carrier receiver (RE). Among them, the carrier generation node is usually physically implemented as a signal generation transpose, and the reference design scheme is as Figure 2 shown.

[0035] RG is a carrier sampler, which is physically implemented as a signal sampling and holding circuit and a signal generator of the same specification. Since the signal transmission process in the industrial field is vulnerable to various interferences, the signal reception method of the above-designed data sampling method has been shown to have robust stability (small gain theorem) in the theoretical proof section. The carrier sampling method can better handle the carrier reception under various interferences in the industrial field.

[0036] RE is a carrier reception transpose, which is physically implemented as a signal conditioning circuit arranged in front of the signal receivers at each node, used to further read the carrier signal from the carrier sampler RG and input it into the signal receiver in the local segment. Such reception allows anti-interference devices and anomaly monitoring devices to be added during this process to ensure the correct construction of the synchronous carrier network.

[0037] The signal receiver in this structure can be of any model, as long as it remains stable, which means that using this patent to generate the global carrier does not require replacing the original signal receiver. On the other hand, the network topology of the RG layer is the same as that of the original node network, which means that the solution in this invention does not require laying new communication links on site and can be completed through the original communication links.

[0038] Through the carrier synchronization network designed in this structure, global signal modulation and synchronous demodulation can be achieved, that is, the signal can be sent from any node in the network, modulated using the carrier signal of that node, and then synchronously demodulated using the signals of other arbitrary nodes, so that the original signal with no error in amplitude and phase can be received at any node.

[0039] Carrier synchronization in the multi-layer network is achieved through a global leader, enabling the modulated signals sent by any node to be synchronously demodulated at other nodes within the global scope, as Figures 3 - 4 shown.

[0040] The following is through the leader (15) As a carrier generator, the carrier signal expression is The carrier replicator is a multi-layer network node in this example. To ensure compliance with the Shannon sampling theorem, the sampling frequency in this example is taken as h = 1000Hz. Global carrier synchronization can be achieved through the controller proposed in this paper, which can be verified in the numerical simulation example. In this example, the message signal is considered as , where and represent the message signal frequencies and . Add the noise signal , represents the original message signal, which is then used as the modulated signal. Randomly select a node modulate the carrier signal and then demodulate it through the carrier signals of each node globally. The specific simulation is as Figures 5 - 6 shown.

[0041] Those skilled in the art should understand that the embodiments of the present invention can be provided as a method, a system, or a computer program product. Therefore, the present invention can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present invention can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) that contain computer-usable program code.

[0042] The present invention is described with reference to the flowcharts and / or block diagrams of methods, apparatuses (systems), and computer program products according to the embodiments of the present invention. It should be understood that each flow and / or block in the flowchart and / or block diagram can be implemented by computer program instructions, and the combination of the flows and / or blocks in the flowchart and / or block diagram can also be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, such that the instructions executed by the processor of the computer or other programmable data processing devices generate means for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0043] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, such that the instructions stored in the computer-readable memory generate a manufactured article including instruction means that implement the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0044] These computer program instructions can also be loaded onto a computer or other programmable data processing device, such that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide steps for implementing the functions specified in Figure 1 one process or multiple processes and / or blocks Figure 1 one block or multiple blocks.

[0045] The above are only the preferred embodiments of the present invention, and are not intended to limit the present invention in other forms. Any person skilled in the art may use the technical content disclosed above to make changes or modifications into equivalent embodiments with equivalent changes. However, any simple modifications, equivalent changes and modifications made to the above embodiments based on the technical essence of the present invention without departing from the technical solution content of the present invention still fall within the protection scope of the technical solution of the present invention.

Claims

1. A distributed data sampling synchronization method for a multi-layer network with high-dimensional node features, characterized in that It includes the following steps: S1: Establish a multi-layer mathematical model with unconstrained inter-layer and intra-layer parameters; S2: Based on the multi-layer mathematical model with unconstrained inter-layer and intra-layer parameters, regard the target state as a virtual node and embed it into the network topology, and design a data sampling controller containing virtual nodes; S3: Define the synchronization error and establish the problem of multi-layer network output asymptotic synchronization with high-dimensional node features; S4: Based on the data sampling controller, solve the problem of non-asymptotic output synchronization of the multi-layer network with high-dimensional node features, establish an output asymptotic synchronization criterion applicable to the multi-layer network with high-dimensional node features, and achieve carrier synchronization in the multi-layer network, so that the modulation signals sent by any node can be synchronously demodulated at other nodes within the whole network.

2. The multi-layer network distributed data sampling synchronization method with high-dimensional node features according to claim 1, wherein The establishment of the multi-layer mathematical model with unconstrained inter-layer and intra-layer parameters is specifically as follows: Consider an undirected weighted network consisting of layers, each layer having identical node systems. The dynamics of the th layer: （1） where q represents adjacent nodes, represents the in-layer connection matrix, represents the inter-layer connection matrix, represents the th layer, defines the first derivative, defines the state of the th node in the th layer, defines the state of the th node in the th layer, is the control gain, is the state matrix, is the input state matrix, represents the internal coupling matrix, represents the external coupling matrix, represents the output coefficient matrix, represents the th layer of nodes input vector, represents the th layer of nodes output vector, represents dimensional Euclidean space, represents dimensional Euclidean space, represents dimensional Euclidean space, represents dimensional Euclidean space, represents dimensional Euclidean space; A state vector describing the entire multi-layer network, where T represents the transpose of the vector, is the state vector of the layer defined as the state of the th node in the is a vector describing the output of the entire multi-layer network system, where is the output variable of the layer indicating the nodes of the layer output vector; is a vector describing the input of the entire multi-layer network system, where is the control input of the input vector of the nodes in the th layer, representing the nodes in the th layer; The production target is a target trajectory, which is the output of the following autonomous system: （2） Among them represents the state matrix of the entire leader 0, represents the output matrix of the entire leader 0, and represent the state and output of leader 0; Regarding the target system (2) as a virtual vertex of the network (1), it is labeled as the th node. Let the intermediate variable be . The augmented coupling graph of the multi-layer network (2) is defined as . The superscript represents the th layer. The vertex set is , and the directed edge set is . represents the directed edge from node to node . If and only if , then there is a connection between the th node in the th layer and its th neighbor node. Otherwise . Among them, , represents the connection between the th node in the th layer and its th neighbor node; represents the connection weight; and . The adjacency matrix is: ； Among them represents the 0 vector, represents the original network connection matrix, vector , represents the original network Laplacian matrix, represents the edge 's connection weight; The Laplacian matrix of graph is defined as expressed as: ， Among them represents the element in the p-th row and p-th column of the matrix, represents a diagonal matrix, For is a positive definite matrix.

3. The multi-layer network distributed data sampling synchronization method with high-dimensional node features according to claim 2, wherein The design of the data sampling controller is specifically as follows: The layer node reference generator is （3） wherein represents the reference generator state of the layer node , and ; represents the reference generator output of the layer node , and represents the layer output, represents the output; represents the distributed control law of the reference generator of the layer node , and represents the layer control law, represents the overall network control law; Due to communication limitations, the sampled-data control method is adopted. The zero-order hold maintains a fixed constant within the sampling interval, and the sampling instant satisfies , where the subscript s represents the sampling node time; the sampling interval is , and the distributed controller is designed as is non-periodic sampling, and the sampling upper bound is , that is . represents the sampling interval, represents identically equal to; the data sampling distributed controller is designed to be: （4） Among them represents the sampling interval represents the control gain matrix represents the number of nodes represents the number of layers represents the state of the th node in the represents the state of the th node in the layer; Based on the sampled-data distributed control law in (4), rewrite the dynamics of the reference generator in (3) as: （5）。 4. The multi-layer network distributed data sampling synchronization method with high-dimensional node features according to claim 3, characterized in that The definition of the synchronization error and the establishment of the problem of multi-layer network output asymptotic synchronization with high-dimensional node features are specifically as follows: Define the error vector : （6） Among them represents the number of nodes, represents the number of layers, represents the error vector, and represents the error vector of the layer, defines the error vector of the th node in the It can be calculated that: ；（7） wherein represents the state of leader 0, and ； Among them, represents the nth node between layers connected to the nth node between layers ; represents the nth node between layers connected to the nth node between layers ; therefore, （8） Among them, represents a p-dimensional identity matrix, represents the direct sum of the off-diagonal elements between layers, represents the diagonal elements between layers, represents the number of layers, represents the cycle multiplication, And calculate: （9） Among them denotes the -dimensional identity matrix, denotes the intra-layer Laplacian matrix, denotes the direct sum of matrices, denotes the addition of two matrices, denotes the cycle multiplication, denotes a unit vector whose -th component is 1 and the rest are 0.

5. The multi-layer network distributed data sampling synchronization method with high-dimensional node features according to claim 1, characterized in that S4 is specifically as follows: Parameterize the data sampling distributed fixed-time control of as , where is the solution of an algebraic Riccati inequality, and select the scalar as a positive scalar; if there exist positive definite matrices and such that: ； If it holds, then and the link is stable, where h represents the upper bound of aperiodic sampling, represents an intermediate variable, represents an intermediate variable.

Images