A simple complex network restraint control method based on high-order topological potential centrality
By using a simple complex network model based on high-order topological potential centrality, the problems of node selection and control gain incompatibility in traditional methods are solved, achieving efficient and low-energy-consumption complex network constraint synchronization, thus improving control efficiency and robustness.
Patent Information
- Application Number
- CN202611213118.1
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2026-08-11
- Publication Date
- 2026-11-10
AI Technical Summary
Traditional complex network control methods suffer from several drawbacks in high-order networks. These include node selection relying on low-order centrality indices, neglecting node participation in high-order simplex systems, inability of control gain to adapt to dynamic changes, and lack of coordinated control between low-order distributed protocols and high-order compensation mechanisms, resulting in poor control efficiency and energy consumption.
By adopting a simple complex network model based on high-order topological potential centrality, and by constructing node selection criteria, adaptive gain allocation algorithm and hierarchical control architecture, key nodes are selected and control gains are dynamically adjusted using high-order topological information. Low-order consensus protocol and high-order compensation mechanism are constructed to achieve efficient and low-energy-consumption constraint synchronization.
It improves network synchronization convergence speed, reduces control energy consumption, enhances robustness to topology disturbances, and provides more accurate identification and control efficiency for key nodes.
Smart Images

Figure CN123069871A_ABST
Abstract
Description
Technical Field
[0001] This invention relates to the field of control and information technology, and in particular to a simple complex network restraint control method based on the high-order topological potential centrality. Background Technology
[0002] With the development of complex network science, network control technology has been widely applied in various fields such as communication, transportation, biological systems, sociology, and artificial intelligence. Traditional complex networks are usually modeled based on binary interactions between nodes, and restraint control methods regulate the overall network state by controlling a small number of key nodes. However, many complex systems in the real world, such as the coordinated firing of biological neural clusters, the group polarization effect of social networks, and the multi-node coordinated regulation of smart grids, often rely on high-order interactions between multiple nodes in their dynamic evolution. Traditional graph theory models are difficult to characterize such multi-entity cooperation and topological heterogeneity, resulting in limited efficiency of restraint control.
[0003] In recent years, higher-order network theory, by introducing mathematical tools such as the simplex, can naturally describe higher-order interactions between multiple nodes. While existing research has made some progress in higher-order network modeling and synchronization analysis, the restraint control strategies for higher-order networks still have the following shortcomings: node selection often relies on low-order centrality indices, ignoring the node's participation in the higher-order simplex; the selected restraining nodes may not be the core hubs in the higher-order interactions, thus reducing control efficiency; control gains are usually fixed values, unable to adapt to changes in the dynamic coupling characteristics of higher-order networks, making it difficult to balance control energy consumption and synchronization performance; and the lack of a cooperative control architecture integrating low-order distributed protocols and high-order compensation mechanisms may lead to poor control performance in multi-scale dynamic coupling scenarios. Therefore, a higher-order network restraint control method that can fully utilize the structural characteristics of the simplex and achieve efficient, low-energy control is needed. Summary of the Invention
[0004] To address the shortcomings of existing technologies, this invention provides a simple complex network restraint control method based on higher-order topological potential centrality. By constructing a simple complex network model, designing node selection criteria that integrate higher-order topological potential centrality, an adaptive gain allocation algorithm, and a hierarchical control architecture, efficient restraint synchronization of the simple complex network is achieved. By fully utilizing higher-order topological information, the network synchronization convergence speed is improved, and control energy consumption is reduced.
[0005] A simple complex network constraint control method based on high-order topological potential centrality includes the following steps: Step 1: Confirm D The dynamic equations of a 1D simple complex network; Construct a with NSimple complex network with 1 node G =( V , E ),in, V ={ v 1, v 2,…, v N} represents a non-empty finite set of nodes. v N Indicates the first N One node; E ={( v i , v j )} is the edge set, v i and v j They represent the first i The node and the first j 1 node; ( i 1, i 2,…, i d+1 ) represents higher-order interactions between nodes. i d+1 For the first d +1 node; based on this, simplexes are used to represent higher-order interactions between nodes, where the 0-simplex represents a node, the 1-simplex represents an edge, the 2-simplex represents a triangle, and so on. d - simplex d A complete graph consisting of +1 nodes; the simple complex network is undirected, unweighted, and connected; the first... i The dynamic equations for each node are expressed as follows: (1); in, Represents a node i The state vector, x im express x i The m One portion, for x i Regarding time t The first derivative, express m A set of dimensional real vectors, with superscript T Represents the transpose of a vector; It is a continuously differentiable function that describes the dynamics of a node itself; D This represents the maximum dimension of the simplex network, that is, the dimension of the highest-dimensional simplex in the network; yesd- The coupling strength corresponding to the simplex; for Elements of an adjacency tensor of order, if node group constitute a - The value is 1 for a simplex form and 0 otherwise; express - Internal coupling functions between nodal state variables in a simplex; Simple complex network reaches synchronization state x s ( t When ), all node states satisfy , ,in x s ( t Satisfying the equations of an isolated system , This represents the node's own dynamic function. Representing time, the following text will discuss this in the function representation. t Omitted; Step 2: Design a restraint control law based on the centrality of high-order topological potential to enable the simple complex network to achieve the synchronization goal; The constrained control law comprises three stages: selecting constrained control nodes, allocating control gain, and hierarchical control input. First, key nodes in the network are selected as constrained control objects based on a high-order topological potential centrality index. Then, a dual adaptive gain allocation algorithm is employed to dynamically adjust the control gain strength according to synchronization errors. Finally, a hierarchical architecture is constructed, consisting of a low-order consensus protocol, a high-order compensation mechanism, and hybrid control. The constrained control... The dynamic equations of a one-dimensional simplex complex network are expressed as follows: (2) in, For binary variables, at node i If constraints and controls are imposed from above, then ,otherwise ; To apply to the node i The law of restraint and control above; Step 2.1: Select the control node based on the centrality of the higher-order topological potential; High-order topological potential centrality HTC i As a parameter for measuring the multidimensional structural characteristics of nodes in a simple complex network, it integrates both low-order centrality and high-order coupling properties of nodes, and is expressed as follows: (3); in, HTC i Represents a nodei The higher-order topological potential centrality, For normalized nodes i The degree is represented by a node. i First order k i The maximum first degree of all nodes in a simple complex network The ratio, i.e. , ; Represents a node i Participants k - The number of simplex elements, T k For all in the network k -A set of simplexes For indicator functions, if node i If the form is a simplex, then σ is 1; otherwise, it is 0. for k Order coupling weight; As the adjustment factor; calculate the adjustment factor for all nodes. Value, select the first one with the largest value. p Each node acts as a check and balance control node. p The number of pre-set control nodes is set; Step 2.2: Design a dual adaptive gain allocation algorithm; The global gain adjustment dynamically adjusts the overall control gain strength based on the network synchronization error, as shown below: (4); in, As the global gain reference; This is the error threshold; It is the hyperbolic tangent function; This indicates taking the minimum value. Let be the Euclidean norm of the synchronization error; Local gain allocation is based on the proportion of a single node's higher-order topological potential centrality to that of all nodes, and the node gain is allocated differentially, as shown below: (5); in, Indicates the first Control gain of each node N i For nodes i The set of neighbors; and They are nodes i and nodes j The higher-order topological potential centrality; Step 2.3: Construct a hierarchical control architecture; The hierarchical control architecture consists of a low-order control layer, a high-order coordination layer, and a hybrid control layer. The control inputs for each layer are represented as follows: The lower-order control layer employs the following distributed consensus protocol for the 1-simplex, as shown below: (6); in, For low-order control input, a ij For elements of the adjacency matrix, if node i and nodes j If there is an edge connecting them, then ,otherwise ; Higher-level coordination layer The first-order simplex adds higher-order compensation coefficients, as shown below: (7); in, for Step-compensation input; These are higher-order compensation coefficients; For all in the network -A set of simplexes; For indicator functions, if node Belongs to simplex If the value is 1, then the value is 0; otherwise, the value is 0. Represents the simplex The number of nodes included; The hybrid control layer combines the control inputs from the low-order control layer and the high-order coordination layer into a total control input, as shown below: (8); Step 3: Use the maximum absolute error to determine whether the simple complex network is synchronized; When determining whether a simple complex network is synchronized to the desired state, the states of all nodes are compared with the synchronization state. The maximum absolute value of the difference between them is used as the criterion for judgment, that is: (9); in, Indicates the maximum absolute error. Represents a node state vector The One portion, Synchronization state The One portion, The total number of nodes. Let be the dimension of the state vector. This indicates taking the maximum value, symbol This indicates taking the absolute value; by designing a constraint control law that satisfies the above judgment criteria, the simple complex network can achieve the synchronization goal.
[0006] The beneficial effects of adopting the above technical solution are as follows: This invention provides a simplex complex network restraint control method based on high-order topological potential centrality. By fusing the first-order degree of nodes with the high-order simplex participation degree through the high-order topological potential centrality index, it can more accurately identify key control nodes. The dual adaptive gain allocation algorithm can dynamically adjust the control gain strength according to the synchronization error, which is more effective in reducing energy consumption than the fixed gain strategy. The hierarchical control architecture coordinates low-order and high-order dynamics, improving the synchronization convergence speed and robustness to topological disturbances. This invention can provide an effective theoretical tool for the regulation of complex systems such as biological neural networks, smart grids, and social networks. Attached Figure Description
[0007] Figure 1 This is a flowchart of a simple complex network restraint control method according to an embodiment of the present invention; Figure 2 This is a schematic diagram of the structure of a simple complex network in an embodiment of the present invention; Figure 3 This is a time evolution diagram of the node state and maximum error obtained by using traditional fixed-gain tethering control in the implementation of this invention; Wherein, (a) is the node state diagram, and (b) is the maximum error diagram; Figure 4 The time evolution diagram of node state and maximum error obtained by the adaptive hierarchical restraint control law based on high-order topological potential centrality proposed in this invention is shown in the embodiment of this invention. Among them, (a) is the node state diagram and (b) is the maximum error diagram. Detailed Implementation
[0008] The specific implementation methods of this application will be further described in detail below with reference to the accompanying drawings and embodiments.
[0009] Example 1: A simple complex network constraint control method based on high-order topological potential centrality, such as... Figure 1 As shown, it includes the following steps: Step 1: Confirm D The dynamic equations of a 1D simple complex network; Construct a with N Simple complex network with 1 node G =( V , E ),in, V ={ v 1, v 2,…,v N} represents a non-empty finite set of nodes. v N Indicates the first N One node; E ={( v i , v j )} is the edge set, v i and v j They represent the first i The node and the first j 1 node; ( i 1, i 2,…, i d+1 ) represents higher-order interactions between nodes. i d+1 For the first d +1 node; based on this, simplexes are used to represent higher-order interactions between nodes, where the 0-simplex represents a node, the 1-simplex represents an edge, the 2-simplex represents a triangle, and so on. d - simplex d A complete graph consisting of +1 nodes; this invention uses simplex complexes to represent higher-order networks, wherein the simplex complex network is undirected, unweighted, and connected; the first i The dynamic equations for each node are expressed as follows: (1); in, Represents a node i The state vector, x im express x i The m One portion, for x i Regarding time t The first derivative, express m A set of dimensional real vectors, with superscript T Represents the transpose of a vector; It is a continuously differentiable function that describes the dynamics of a node itself; D This represents the maximum dimension of the simplex network, that is, the dimension of the highest-dimensional simplex in the network; yes d- The coupling strength corresponding to the simplex; for Elements of an adjacency tensor of order, if node group constitute a - The value is 1 for a simplex form and 0 otherwise; express - Internal coupling functions between nodal state variables in a simplex; Simple complex network reaches synchronization state x s ( t When ), all node states satisfy , ,in x s ( t Satisfying the equations of an isolated system , This represents the node's own dynamic function. Representing time, the following text will discuss this in the function representation. t Omitted; This embodiment uses, as follows Figure 2 Taking the simple complex network shown as an example, when the coupling strength is taken as , Fixed gain At that time, the network contains 12 0-simulacra (12 nodes), 15 1-simulacra (15 edges), and 4 2-simulacra. and All the simplexes mentioned above constitute a simplex complex network, and the maximum dimension of this simplex complex network is... ; In this embodiment, regarding the node's own dynamics Consider it as a Rössler system; a single Rössler system is described as follows:
[0010] in, , Under these parameters, the system has an unstable equilibrium point. , It is a single system A solution that satisfies ; In this embodiment, the internal coupling function is: and ; Coupling strength is taken , ; Step 2: Design a restraint control law based on the centrality of high-order topological potential to enable the simple complex network to achieve the synchronization goal; The constrained control law comprises three stages: selecting constrained control nodes, allocating control gain, and hierarchical control input. First, key nodes in the network are selected as constrained control objects based on a high-order topological potential centrality index. Then, a dual adaptive gain allocation algorithm is employed to dynamically adjust the control gain strength according to synchronization errors. Finally, a hierarchical architecture is constructed, consisting of a low-order consensus protocol, a high-order compensation mechanism, and hybrid control. The constrained control... The dynamic equations of a one-dimensional simplex complex network are expressed as follows: (2) in, For binary variables, at node i If constraints and controls are imposed from above, then ,otherwise ; To apply to the node i The law of restraint and control above; Step 2.1: Select the control node based on the centrality of the higher-order topological potential; High-order topological potential centrality HTC i As a parameter for measuring the multidimensional structural characteristics of nodes in a simple complex network, it integrates both low-order centrality and high-order coupling properties of nodes, and is expressed as follows: (3); in, HTC i Represents a node i The higher-order topological potential centrality is a composite index that combines the lower-order centrality of nodes with the higher-order simplex participation, and is used to measure the multidimensional structural importance of nodes in simplex complex networks. For normalized nodes i The degree is represented by a node. i First order k i The maximum first degree of all nodes in a simple complex network The ratio, i.e. , ; Represents a node i Participants k - The number of simplex elements, T k For all in the network k -A set of simplexes For indicator functions, if node i If the form is a simplex, then σ is 1; otherwise, it is 0. for k Order coupling weight; As the adjustment factor; calculate the adjustment factor for all nodes. Value, select the first one with the largest value.p Each node acts as a check and balance control node. p The number of pre-set control nodes is set; Step 2.2: Design a dual adaptive gain allocation algorithm; The global gain adjustment dynamically adjusts the overall control gain strength based on the network synchronization error, as shown below: (4); in, As the global gain reference; This is the error threshold; It is the hyperbolic tangent function; This indicates taking the minimum value. Let be the Euclidean norm of the synchronization error; Local gain allocation is based on the proportion of a single node's higher-order topological potential centrality to that of all nodes, and the node gain is allocated differentially, as shown below: (5); in, Indicates the first Control gain of each node N i For nodes i The set of neighbors; and They are nodes i and nodes j The higher-order topological potential centrality; Step 2.3: Construct a hierarchical control architecture; The hierarchical control architecture consists of a low-order control layer, a high-order coordination layer, and a hybrid control layer. The control inputs for each layer are represented as follows: The lower-order control layer employs the following distributed consensus protocol for the 1-simplex, as shown below: (6); in, For low-order control input, a ij For elements of the adjacency matrix, if node i and nodes j If there is an edge connecting them, then ,otherwise ; Higher-level coordination layer The first-order simplex adds higher-order compensation coefficients, as shown below: (7); in, for Step-compensation input; These are higher-order compensation coefficients; For all in the network -A set of simplexes; For indicator functions, if node Belongs to simplex If the value is 1, then the value is 0; otherwise, the value is 0. Represents the simplex The number of nodes included; The hybrid control layer combines the control inputs from the low-order control layer and the high-order coordination layer into a total control input, as shown below: (8); In this embodiment, the values of each parameter are as follows: (The values used in the calculation of the centrality of higher-order topological potential are...) , In adaptive gain allocation, the reference gain is taken. Error threshold In a hierarchical control architecture, higher-order compensation coefficients are used. ; Step 3: Use the maximum absolute error to determine whether the simple complex network is synchronized; When determining whether a simple complex network is synchronized to the desired state, the states of all nodes are compared with the synchronization state. The maximum absolute value of the difference between them is used as the criterion for judgment, that is: (9); in, Indicates the maximum absolute error. Represents a node state vector The One portion, Synchronization state The One portion, The total number of nodes. Let be the dimension of the state vector. This indicates taking the maximum value, symbol This indicates taking the absolute value; by designing a constraint control law that satisfies the above judgment criteria, the simple complex network can achieve the synchronization goal.
[0011] Example 2: In this embodiment, Figure 2 Nodes 6, 7, and 10 in Traditional constraint control is applied. Under the current parameters, the node state and error change curves of the network are as follows: Figure 3 As shown in (a) and (b) in the simulation, the node states of the network eventually stabilize at the equilibrium point. The error curve eventually approaches 0, indicating that the network has achieved synchronization.
[0012] Using a restraint control law based on the centrality of higher-order topological potentials to Figure 2 When control is applied to a network, with the same number of controlled nodes, the curves showing the changes in node states and errors are as follows: Figure 4 As shown in (a) and (b) in the simulation diagram, the network eventually synchronizes to the desired state, and the control speed is significantly better than that of traditional restraint control.
[0013] Based on this understanding, the technical solution of this application, in essence, or the part that contributes to the prior art, or part of the technical solution, can be embodied in the form of a computer program product.
[0014] The various embodiments in this application are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.
[0015] The scope of protection of this application is not limited to the embodiments described above. Obviously, those skilled in the art can make various modifications and variations to this disclosure without departing from the scope and spirit of this disclosure. If such modifications and variations fall within the scope of the methods disclosed herein and their equivalents, then the intent of this disclosure also includes such modifications and variations.
Claims
1. A simple complex network restraint control method based on high-order topological potential centrality, characterized in that, Includes the following steps: Step 1: Confirm D The dynamic equations of a 1D simple complex network; Step 2: Design a restraint control law based on the centrality of high-order topological potential to enable the simple complex network to achieve the synchronization goal; Step 3: Use the maximum absolute error to determine whether the simple complex network is synchronized.
2. The simple complex network restraint control method based on high-order topological potential centrality according to claim 1, characterized in that, Step 1 specifically involves: constructing a system with... N Simple complex network with 1 node G =( V , E ),in, V ={ v 1, v 2,…, v N } represents a non-empty finite set of nodes. v N Indicates the first N One node; E ={( v i , v j )} is the edge set, v i and v j They represent the first i The node and the first j 1 node; ( i 1, i 2,…, i d+1 ) represents higher-order interactions between nodes. i d+1 For the first d +1 node; based on this, simplexes are used to represent higher-order interactions between nodes, where the 0-simplex represents a node, the 1-simplex represents an edge, the 2-simplex represents a triangle, and so on. d - simplex d A complete graph consisting of +1 nodes; the simple complex network is undirected, unweighted, and connected; the first... i The dynamic equations for each node are expressed as follows: (1); in, Represents a node i The state vector, x im express x i The m One portion, for x i Regarding time t The first derivative, express m A set of dimensional real vectors, with superscript T Represents the transpose of a vector; It is a continuously differentiable function that describes the dynamics of a node itself; D This represents the maximum dimension of the simplex network, that is, the dimension of the highest-dimensional simplex in the network; yes d- The coupling strength corresponding to the simplex; for Elements of an adjacency tensor of order, if node group constitute a - The value is 1 for a simplex form and 0 otherwise; express - Internal coupling functions between nodal state variables in a simplex; Simple complex networks achieve synchronization x s ( t When ), all node states satisfy , ,in x s ( t Satisfying the equations of an isolated system , This represents the node's own dynamic function. Representing time, the following text will discuss this in the function representation. t Omitted.
3. The simple complex network restraint control method based on high-order topological potential centrality according to claim 2, characterized in that, Step 2 specifically involves the following: The restraint control law comprises three stages: selecting restraint control nodes, allocating control gain, and hierarchical control input. First, key nodes in the network are selected as restraint control objects based on the high-order topological potential centrality index. Then, a dual adaptive gain allocation algorithm is used to dynamically adjust the control gain strength according to the synchronization error, and a hierarchical architecture consisting of a low-order consensus protocol, a high-order compensation mechanism, and hybrid control is constructed. The restrained nodes... The dynamic equations of a one-dimensional simplex complex network are expressed as follows: (2); in, For binary variables, at node i If constraints and controls are imposed from above, then ,otherwise ; To apply to the node i The law of restraint and control.
4. The simple complex network restraint control method based on high-order topological potential centrality according to claim 3, characterized in that, Step 2 includes the following steps: Step 2.1: Select the control node based on the centrality of the higher-order topological potential; High-order topological potential centrality HTC i As a parameter for measuring the multidimensional structural characteristics of nodes in a simple complex network, it integrates both low-order centrality and high-order coupling properties of nodes, and is expressed as follows: (3); in, HTC i Represents a node i higher-order topological potential centrality For normalized nodes i The degree is represented by a node. i First order k i The maximum first degree of all nodes in a simple complex network The ratio, i.e. , ; Represents a node i Participants k - The number of simplex elements, T k For all in the network k -A set of simplexes For indicator functions, if node i If the form is a simplex, then σ is 1; otherwise, it is 0. for k Order coupling weights; As the adjustment factor; calculate the adjustment factor for all nodes. Value, select the first one with the largest value. p Each node acts as a check and balance control node. p The number of pre-set control nodes is set. Step 2.2: Design a dual adaptive gain allocation algorithm; The global gain adjustment dynamically adjusts the overall control gain strength based on the network synchronization error, as shown below: (4); in, As the global gain reference; This is the error threshold; It is the hyperbolic tangent function; This indicates taking the minimum value. Let be the Euclidean norm of the synchronization error; Local gain allocation is based on the proportion of a single node's higher-order topological potential centrality to that of all nodes, and the node gain is allocated differentially, as shown below: (5); in, Indicates the first Control gain of each node N i For nodes i The set of neighbors; and They are nodes i and nodes j The higher-order topological potential centrality; Step 2.3: Construct a hierarchical control architecture; The hierarchical control architecture consists of a low-order control layer, a high-order coordination layer, and a hybrid control layer. The control inputs for each layer are represented as follows: The lower-order control layer employs the following distributed consensus protocol for the 1-simplex, as shown below: (6); in, For low-order control input, a ij For elements of the adjacency matrix, if node i and nodes j If there is an edge connecting them, then ,otherwise ; Higher-level coordination layer The first-order simplex adds higher-order compensation coefficients, as shown below: (7); in, for Step-compensation input; These are higher-order compensation coefficients; For all in the network -A set of simplexes; For indicator functions, if node Belongs to simplex If the value is 1, then the value is 0; otherwise, the value is 0. Represents the simplex The number of nodes included; The hybrid control layer combines the control inputs from the low-order control layer and the high-order coordination layer into a total control input, as shown below: (8)。 5. The simple complex network restraint control method based on high-order topological potential centrality according to claim 4, characterized in that, Step 3 specifically involves: when determining whether the simple complex network has synchronized to the desired state, comparing the states of all nodes with the synchronization state. The maximum absolute value of the difference between them is used as the criterion for judgment, that is: (9); in, Indicates the maximum absolute error. Represents a node state vector The One portion, Synchronization state The One portion, The total number of nodes. Let be the dimension of the state vector. This indicates taking the maximum value, symbol This indicates taking the absolute value; by designing a constraint control law that satisfies the above judgment criteria, the simple complex network can achieve the synchronization goal.
