Personal perception network structure optimization method based on improved snake vulture algorithm

Abstract

The invention discloses a presence / absence perception network structure optimization method based on an improved snake vulture algorithm, and the method comprises the steps: building a multi-target optimization model according to a heterogeneous presence / absence collaborative perception network structure based on the degree of nodes and the average value of the reciprocal of the shortest path between node pairs, initializing the parameters of the improved snake vulture algorithm, and carrying out the optimization of the multi-target optimization model. An upper triangular coding method is adopted to form individuals in the population, and Tent chaotic mapping is introduced. Solving the multi-objective optimization model by using an improved snake vulture algorithm, and updating the position and fitness value of the individual through two escape strategies of the snake vulture; gaussian and Cauchy variation disturbance is introduced to help the algorithm jump out of local optimum and search towards a globally optimal solution; a greedy strategy is introduced, and the algorithm is guided to keep tracking of an optimal solution in the search process by comparing fitness values. According to the invention, the optimal heterogeneous presence / absence cooperative sensing network structure under different tasks can be obtained, and the overall efficiency of the presence / absence system is improved.

Description

Technical Field

[0001] The invention belongs to the field of collaborative perception network with and without personnel, and in particular relates to a method for optimizing the structure of a presence / absence perception network based on an improved secretary vulture algorithm. Background Art

[0002] With the rapid development of unmanned systems, the future joint system will develop towards "unmanned systems". When unmanned systems perform collaborative perception tasks, the network structure supporting the unmanned systems plays an important role, especially in complex environments, where tasks are dynamic and changeable, and each unit is heterogeneous, which puts higher requirements on the adaptability and timeliness of the network structure. However, the existing network structure of unmanned systems has poor robustness and timeliness for target recognition tasks, and excessive energy loss and poor connectivity for target tracking tasks, which cannot meet the task requirements, and the network structure cannot adapt to changes in the environment. To this end, the network structure of unmanned systems needs to have flexible adjustment capabilities, which can be adjusted in real time according to changes in tasks, so that the network structure of unmanned systems can move with the tasks, thereby ensuring the collaborative effectiveness of unmanned systems. At present, in terms of the optimization of the network structure of unmanned collaborative perception, domestic and foreign research mainly focuses on model construction and algorithm design. There are still some shortcomings in the relevant technology: in terms of modeling, node heterogeneity is not taken into account, and multiple perception scenarios are not considered, resulting in the architecture being unable to adapt to environmental changes; in terms of model solving, the convergence and efficiency of the algorithm still need to be improved; in terms of comprehensive performance analysis of network structure, the selected indicators are too single, and such comprehensive performance indicators of network structure are difficult to guarantee. Summary of the invention

[0003] In view of the problems existing in the prior art, the present invention provides a method for optimizing the structure of a network with and without personnel perception based on an improved snake vulture algorithm. The method can establish an optimization model for the structure of a heterogeneous collaborative perception network with and without personnel perception based on the characteristics of an actual environment with the optimization goals of improving the robustness of the architecture, the timeliness of information, and reducing energy loss. The method can also meet the needs of different tasks by adjusting the weights of the objective functions. At the same time, the improved snake vulture algorithm is used for solving the problem, so as to obtain the optimal heterogeneous collaborative perception network structure with and without personnel perception under different tasks, so as to improve the overall performance of the system with and without personnel perception.

[0004] In order to solve the above technical problems, the present invention provides the following technical solutions: a method for optimizing the network structure of presence or absence perception based on an improved secretary vulture algorithm, comprising the following steps: S1. According to the heterogeneous collaborative perception network structure with and without human, a multi-objective optimization model is established based on the node degree and the average value of the inverse of the shortest path between node pairs. Specifically, objective functions are constructed to maximize robustness, maximize information timeliness, and minimize energy loss. Then, the weight coefficient of the objective function is adjusted according to the task requirements, and the constraint conditions of the objective function are added according to the characteristics of the heterogeneous collaborative perception network structure with and without human. S2. Initialize the parameters in the improved secretary vulture algorithm; for the adjacency matrix of the heterogeneous collaborative sensing network structure with and without humans, use the upper triangle coding method to form individuals in the population; in the population initialization stage, introduce the Tent chaotic mapping to increase the diversity of the population; S3, using the improved secretary vulture algorithm to solve the multi-objective optimization model, including the following steps: S3.1, start iteration, and update the individual positions of secretary bird through its three hunting strategies; S3.2, during the iteration process, the positions of the individuals are updated and their fitness values ​​are updated through the two escape strategies of the secretary bird; S3.3, introduce Gaussian and Cauchy mutation disturbances, update the fitness value, help the algorithm jump out of the local optimum, and continue to search for the global optimal solution; S3.4, based on the fitness value in S3.3, introduce a greedy strategy to guide the algorithm to keep tracking the optimal solution during the search process by comparing the fitness values; S3.5. Determine whether the maximum number of iterations is met. If so, the heterogeneous collaborative perception network structure with or without personnel corresponding to the individuals in the population after iteration is output as the most optimal collaborative perception network structure with or without personnel. Otherwise, execute step S3.1.

[0005] Furthermore, in the aforementioned step S1, based on the degree of the node k i , the objective function to maximize the robustness R is constructed as follows: , , In the formula, I i For Node i The importance of k i For the i is the degree of a node, and N is the total number of nodes.

[0006] Furthermore, in the aforementioned step S1, based on the average value T of the inverse of the shortest path length between node pairs, an objective function for maximizing information timeliness is constructed as follows: , In the formula, d ij For Node i and nodes j The shortest path length between .

[0007] Furthermore, in the aforementioned step S1, based on the distance between nodes, an objective function with the minimum energy loss as the goal is constructed as follows: , Where, d is the distance between two communication nodes; E elec Represents the energy required to receive and send a unit of data; E fs represents the energy dissipation coefficient of the free space mode circuit; E mp Represents the energy consumption coefficient of the circuit in multipath fading mode.

[0008] Furthermore, in the aforementioned step S1, the constraints of the objective function are as follows: Constraints on edges in heterogeneous collaborative sensing networks: , , In the formula, W C-R represents the number of edges between the command node and the combat soldiers, W R-R represents the number of edges between combat soldiers, W R-UGV represents the edge between combat soldiers and unmanned vehicles, W R-UAV Represents the edge between combat soldiers and drones; The constraints on the degree of nodes in the heterogeneous unmanned collaborative sensing network structure are: , , In the formula, k person Indicates the degree of commander and combat soldiers. k non-person represents the speed of drones and unmanned vehicles, k const Indicates node saturation; Constraints on network density in heterogeneous collaborative sensing network structures: , In the formula, D represents the network density, D MAX Indicates the maximum network density threshold; Constraints on energy loss in heterogeneous collaborative sensing network structures: , In the formula, E i Representation Node i energy loss; E max Representation Nodei Maximum energy capacity; Adjacency matrix in heterogeneous collaborative sensing network structure A The value constraints are: , In the formula, the adjacency matrix of heterogeneous collaborative sensing network structure is A Medium Element a ij , a value of 1 indicates that the nodes are connected, and a value of 0 indicates that the nodes are not connected.

[0009] Furthermore, the parameters for initializing the improved secretary vulture algorithm in the aforementioned step S2 include: population size P , number of iterations G , chaos coefficient t , positional parameters μ , scale parameter γ, degrees of freedom v ,parameter r .

[0010] Furthermore, the three hunting strategies of the secretary bird in the aforementioned step S3.1 are divided into three time intervals, namely: , , , corresponding to the three stages of predation by secretary birds: finding prey, consuming prey and attacking prey; When the number of iterations When , the formula is as follows: , In the formula, t represents the current number of iterations, T represents the maximum number of iterations, Indicates i The new state of the secretary bird in the stage of searching for prey, and is a random candidate solution for the iteration of the prey search phase; R1 represents the dimension randomly generated in the interval [0,1] An array of is the dimension of the solution space; represents the value of its j-th dimension, Represents the fitness value of its objective function; When the number of iterations When , the formula is as follows: , In the formula, RB represents the dimension randomly generated from the standard normal distribution. Array ，x best Indicates the current optimal value; When the number of iterations When , the formula is as follows: , In the formula, RL is the flight step length in Levy flight, t is the current number of iterations, and T is the maximum number of iterations. Indicates i The new state of the secretary bird in the stage of searching for prey, Represents the fitness value of its objective function.

[0011] Furthermore, in the aforementioned step S3.2, the positions of the individuals are updated by using the two escape strategies of the secretary bird. The two escape strategy formulas of the secretary bird are as follows: , Among them, R2 means randomly generating dimensions from a normal distribution Array of x random represents the random candidate solution of the current iteration, and K represents a random selection of an integer 1 or 2; The fitness function in step S3.2 is as follows: , In the formula, fit is the objective function, Fit is the fitness value, Ave_d ij is the number of path hops between nodes, R is the system robustness, T is the information timeliness, W ij For Node i With Node j The energy loss between w 1. w 2. w 3. And w 4 represent the weight coefficients of the path hop count, system robustness, information timeliness and energy loss between nodes, respectively. is the penalty for violating the constraint.

[0012] Furthermore, in the aforementioned step S3.3, Gaussian and Cauchy variation disturbances are introduced, and the formula is as follows: , In the formula, is the updated solution, is subject to the mean of 0 and the variance of The normal distribution of r i Obey positional parameters , scale parameter , degrees of freedom The Cauchy distribution of and are the weight coefficients of Gaussian distribution and Cauchy distribution respectively, t and T are the current number of iterations and the maximum number of iterations respectively.

[0013] Furthermore, the aforementioned method for optimizing the structure of an unmanned sensing network based on the improved snake vulture algorithm also includes an analysis of the performance of the most optimal structure of the unmanned collaborative sensing network structure, including: robustness, timeliness, connectivity, average path length, average betweenness centrality, clustering coefficient and association closeness, as well as robustness, timeliness and average path length under node attacks.

[0014] Compared with the prior art, the beneficial technical effects of the present invention using the above technical solution are as follows: (1) The model is constructed by taking into account system robustness, information timeliness, and energy loss, thereby ensuring the comprehensive performance of the heterogeneous collaborative sensing network structure with or without human intervention; (2) According to different task requirements, adjust the weight of the objective function to obtain the optimal network structure under different tasks, meet the requirements of perception tasks, and improve the adaptability and timeliness of the network structure; (3) Constrain different nodes and their edges in the model to reflect node heterogeneity and effectively simulate the characteristics of the actual environment; (4) By improving the initial population of the secretary vulture algorithm and introducing Gaussian and Cauchy mutation perturbations and greedy strategies, the algorithm has better convergence and search efficiency when solving the structural optimization problem of heterogeneous collaborative sensing networks with and without humans. (5) The robustness, timeliness, connectivity, average path length, average betweenness centrality, clustering coefficient and correlation closeness indicators in complex network theory are used to compare and analyze the heterogeneous collaborative sensing network structure with and without humans before and after optimization, which shows the effectiveness of the optimized network structure; (6) The heterogeneous collaborative perception network structure optimization method based on the improved snake vulture algorithm designed by the present invention is concise and efficient, and can quickly obtain the optimal heterogeneous collaborative perception network structure. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] Figure 1 It is a flowchart of a method for optimizing a heterogeneous collaborative sensing network structure with or without humans based on an improved secretary vulture algorithm provided according to an embodiment of the present invention.

[0016] Figure 2 It is a schematic diagram of the upper triangle coding rule provided according to an embodiment of the present invention.

[0017] Figure 3It is a comparison chart of the fitness convergence of the improved secretary vulture algorithm (LSBOA) and the basic secretary vulture algorithm (SBOA), the dung beetle optimization algorithm (DBO), the grasshopper optimization algorithm (GOA), and the improved dung beetle optimization algorithm (TWDBO) under the target recognition task provided by the embodiment of the present invention.

[0018] Figure 4 It is an optimal collaborative perception network structure diagram with or without personnel obtained by improving the secretary vulture algorithm under the target recognition task provided by the embodiment of the present invention.

[0019] Figure 5 It is a comparison chart of the timeliness of the target recognition task provided by an embodiment of the present invention, before and after optimization, with / without the collaborative perception network under random attacks on nodes.

[0020] Figure 6 It is a comparison chart of the robustness of the collaborative perception network with / without optimization under random attacks on nodes in the target recognition task provided by an embodiment of the present invention.

[0021] Figure 7 It is a comparison chart of average path lengths of a collaborative sensing network with / without human under random attacks on nodes before and after optimization in a target recognition task provided by an embodiment of the present invention.

[0022] Figure 8 It is a comparison chart of the fitness convergence of the improved secretary vulture algorithm (LSBOA) and the basic secretary vulture algorithm (SBOA), the dung beetle optimization algorithm (DBO), the grasshopper optimization algorithm (GOA), and the improved dung beetle optimization algorithm (TWDBO) under the target tracking task provided by an embodiment of the present invention.

[0023] Fig. 9 It is an optimal collaborative perception network structure diagram with or without human intervention obtained by improving the secretary vulture algorithm under the target tracking task provided by an embodiment of the present invention.

[0024] Fig.10 It is a comparison chart of the timeliness of the target tracking task provided by an embodiment of the present invention before and after optimization with and without an unmanned collaborative perception network under random attacks on nodes.

[0025] Fig.11 It is a comparison chart of the robustness of the unmanned collaborative perception network under random attacks on nodes before and after optimization in the target tracking task provided by an embodiment of the present invention.

[0026] Fig.12 It is a comparison chart of average path lengths before and after optimization under random attacks on nodes in an unmanned collaborative perception network in a target tracking task provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0027] In order to better understand the technical content of the present invention, specific embodiments are given and described as follows in conjunction with the accompanying drawings.

[0028] Various aspects of the invention are described herein with reference to the accompanying drawings, in which many illustrative embodiments are shown. The embodiments of the invention are not limited to those described in the accompanying drawings. It should be understood that the invention is implemented by any of the various concepts and embodiments described above, as well as the concepts and embodiments described in detail below, because the concepts and embodiments disclosed in the invention are not limited to any implementation. In addition, some aspects disclosed in the invention may be used alone or in any appropriate combination with other aspects disclosed in the invention.

[0029] refer to Figure 1 The present invention provides a method for optimizing a network structure based on an improved snake vulture algorithm, comprising the following steps: S1. According to the heterogeneous collaborative perception network structure with and without human, a multi-objective optimization model is established based on the node degree and the average value of the inverse of the shortest path between node pairs. Specifically, objective functions are constructed to maximize robustness, maximize information timeliness, and minimize energy loss. Then, the weight coefficient of the objective function is adjusted according to the task requirements, and the constraint conditions of the objective function are added according to the characteristics of the heterogeneous collaborative perception network structure with and without human. S2. Initialize the parameters in the improved secretary vulture algorithm; for the adjacency matrix of the heterogeneous collaborative sensing network structure with and without humans, use the upper triangle coding method to form individuals in the population; in the population initialization stage, introduce the Tent chaotic mapping to increase the diversity of the population; S3, using the improved secretary vulture algorithm to solve the multi-objective optimization model, including the following steps: S3.1, start iteration, and update the individual positions of secretary birds through their three hunting strategies; S3.2, during the iteration process, the positions of the individuals are updated and their fitness values ​​are updated through the two escape strategies of the secretary bird; S3.3, introduce Gaussian and Cauchy mutation disturbances, update the fitness value, help the algorithm jump out of the local optimum, and continue to search for the global optimal solution; S3.4, based on the fitness value in S3.3, introduce a greedy strategy to guide the algorithm to keep tracking the optimal solution during the search process by comparing the fitness values; S3.5. Determine whether the maximum number of iterations is met. If so, the heterogeneous collaborative perception network structure with or without personnel corresponding to the individuals in the population after iteration is output as the most optimal collaborative perception network structure with or without personnel. Otherwise, execute step S3.1.

[0030] As a preferred embodiment of the present invention, in step S1, based on the degree of the node, an objective function with the goal of maximizing robustness is constructed as follows: Based on node degree k i , the objective function to maximize the robustness R is constructed as follows: , , In the formula, I i For Node i The importance of k i For the i is the degree of a node, and N is the total number of nodes.

[0031] As a preferred embodiment of the present invention, in step S1, based on the average value T of the inverse of the shortest path length between node pairs, an objective function with the goal of maximizing information timeliness is constructed as follows: , In the formula, d ij For Node i and nodes j The shortest path length between .

[0032] As a preferred embodiment of the present invention, in step S1, based on the distance between nodes, an objective function with minimum energy loss as the goal is constructed as follows: , Where, d is the distance between two communication nodes; E elec Represents the energy required to receive and send a unit of data; E fs represents the energy dissipation coefficient of the free space mode circuit; E mp Represents the energy consumption coefficient of the circuit in multipath fading mode.

[0033] As a preferred embodiment of the present invention, in step S1, the constraint conditions of the objective function are as follows: Constraints on edges in heterogeneous collaborative sensing networks: , , In the formula, W C-R represents the number of edges between the command node and the combat soldiers, W R-R represents the number of edges between combat soldiers, W R-UGV represents the edge between combat soldiers and unmanned vehicles, WR-UAV Represents the edge between combat soldiers and drones; The constraints on the degree of nodes in the heterogeneous unmanned collaborative sensing network structure are: , , In the formula, k person Indicates the degree of commander and combat soldiers. k non-person represents the speed of drones and unmanned vehicles, k const Indicates node saturation; Constraints on network density in heterogeneous collaborative sensing network structures: , In the formula, D represents the network density, D MAX Indicates the maximum network density threshold; Constraints on energy loss in heterogeneous collaborative sensing network structures: , In the formula, E i Representation Node i energy loss; E max Representation Node i Maximum energy capacity; Adjacency matrix in heterogeneous collaborative sensing network structure A The value constraints are: , In the formula, the adjacency matrix of heterogeneous collaborative sensing network structure is A Medium Element a ij , a value of 1 indicates that the nodes are connected, and a value of 0 indicates that the nodes are not connected.

[0034] As a preferred embodiment of the present invention, the parameters in the improved secretary vulture algorithm initialized in step S2 include: population size P=100 , number of iterations G=500 , chaos coefficient t=1.1 , positional parameters μ=0 , scale parameter γ = 1, degrees of freedom v=1 ,parameter r= 0.5 .

[0035] refer to Figure 2, upper triangle coding method, specifically: the adjacency matrix A of the heterogeneous collaborative sensing network structure with or without humans is encoded in binary, with 1 and 0 representing whether each node in the network structure is connected or not, for the mth row in the adjacency matrix A of the heterogeneous collaborative sensing network structure with or without humans, delete the first m elements in the mth row, and arrange the remaining elements into a row vector to form the secretary bird in the improved secretary bird algorithm.

[0036] As a preferred embodiment of the present invention, the three hunting strategies of secretary birds in step S3.1 are divided into three time intervals, namely: , , , corresponding to the three stages of predation by secretary birds: finding prey, consuming prey and attacking prey; When the number of iterations When , the formula is as follows: , In the formula, t represents the current number of iterations, T represents the maximum number of iterations, Indicates i The new state of the secretary bird in the stage of searching for prey, and is a random candidate solution for the iteration of the prey search phase; R1 represents the dimension randomly generated in the interval [0,1] An array of is the dimension of the solution space; represents the value of its j-th dimension, Represents the fitness value of its objective function; When the number of iterations When , the formula is as follows: , In the formula, RB represents the dimension randomly generated from the standard normal distribution. Array ，x best Indicates the current optimal value; When the number of iterations When , the formula is as follows: , In the formula, RL is the flight step length in Levy flight, t is the current number of iterations, and T is the maximum number of iterations. Indicates i The new state of the secretary bird in the stage of searching for prey, Represents the fitness value of its objective function.

[0037] As a preferred embodiment of the present invention, in step S3.2, the positions of the individuals are updated by using two escape strategies of the secretary bird. The two escape strategy formulas of the secretary bird are as follows: , Among them, R2 means randomly generating dimensions from a normal distribution Array of x random represents the random candidate solution of the current iteration, and K represents a random selection of an integer 1 or 2; The fitness function in step S3.2 is as follows: , In the formula, fit is the objective function, Fit is the fitness value, Ave_d ij is the number of path hops between nodes, R is the system robustness, T is the information timeliness, W ij For Node i With Node j The energy loss between w 1. w 2. w 3. And w 4 represent the weight coefficients of the path hop count, system robustness, information timeliness and energy loss between nodes, respectively. is the penalty for violating the constraint.

[0038] As a preferred embodiment of the present invention, Gaussian and Cauchy variation disturbances are introduced in the aforementioned step S3.3, and the formula is as follows: , In the formula, is the updated solution, is subject to the mean of 0 and the variance of The normal distribution of r i Obey positional parameters , scale parameter , degrees of freedom The Cauchy distribution of and are the weight coefficients of Gaussian distribution and Cauchy distribution respectively, t and T are the current number of iterations and the maximum number of iterations respectively.

[0039] As a preferred embodiment of the method for optimizing the structure of an unmanned perception network based on the improved snake vulture algorithm of the present invention, it also includes an analysis of the performance of the most excellent structure of the unmanned collaborative perception network structure, including: robustness, timeliness, connectivity, average path length, average betweenness centrality, clustering coefficient and correlation closeness, as well as robustness, timeliness and average path length under node attacks.

[0040] Table 1: Comparison of the optimal architecture in terms of robustness, timeliness, connectivity, and average path length under the target recognition task. Table 2: Comparison of the optimal architecture in terms of average betweenness centrality, global clustering coefficient, and association closeness under the target recognition task.

[0041] Table 1 Architecture Type Centralized architecture 1.1659 0.4684 0.0439 2.4000 DBO Optimized Architecture 3.3842 0.5737 0.1683 2.1895 SBOA Optimized Architecture 4.0503 0.6737 0.2120 1.6632 LSBOA Optimized Architecture 6.2583 0.7235 0.3486 1.2341 Table 2 Architecture Type Average betweenness centrality Global clustering coefficient Closeness of association Centralized architecture 0 0.006 0.2312 DBO Optimized Architecture 3.3000 0.1326 0.3679 SBOA Optimized Architecture 6.3500 0.2956 0.5876 LSBOA Optimized Architecture 7.2500 0.3567 0.6979 Figure 4 The optimal network structure diagram of cooperative perception with and without human beings obtained by the improved secretary vulture algorithm under the target recognition task provided by the embodiment of the present invention is shown. It can be seen from the figure that under the target recognition task, the network structure after LSBOA optimization is more evenly distributed. Figure 5 The comparison chart of the timeliness of the target recognition task provided by the embodiment of the present invention before and after optimization with / without the coordinated sensing network under random attacks on nodes is shown. It can be seen from the figure that in addition to the distributed network structure under ideal conditions, the optimal network structure considering heterogeneity has better timeliness when facing the target recognition task. Figure 6 This is a comparison chart of the robustness of the coordinated sensing network with and without human intervention under random attacks on nodes before and after optimization in the target recognition task provided by an embodiment of the present invention. It can be seen from the figure that in addition to the distributed network structure under ideal conditions, the optimal network structure considering heterogeneity has better robustness when facing the target recognition task. Figure 7 This is a comparison chart of the average path length of the coordinated sensing network with and without human intervention under random attacks on nodes before and after optimization under the target recognition task provided by an embodiment of the present invention. It can be seen from the figure that in addition to the distributed network structure under ideal conditions, the optimal network structure considering heterogeneity has a lower average path length when facing the target recognition task. Figure 8 This is a comparison chart of the fitness convergence of the improved secretary vulture algorithm (LSBOA) and the basic secretary vulture algorithm (SBOA), the dung beetle optimization algorithm (DBO), the grasshopper optimization algorithm (GOA), and the improved dung beetle optimization algorithm (TWDBO) under the target tracking task provided by the embodiment of the present invention. It can be seen from the figure that when facing the target tracking task, the LSBOA algorithm has better convergence and a lower fitness value. Fig. 9 This is the optimal network structure diagram of cooperative sensing with or without human beings obtained by improving the secretary vulture algorithm under the target tracking task provided by the embodiment of the present invention. As can be seen from the figure, under the target tracking task, the network structure after LSBOA optimization is more evenly distributed.

[0042] Fig.10This is a comparison chart of the timeliness of the target tracking task provided by the embodiment of the present invention, before and after optimization, with and without the coordinated sensing network under random attacks on nodes. It can be seen from the figure that in addition to the distributed network structure under ideal conditions, the optimal network structure considering heterogeneity has better timeliness when facing the target tracking task. Fig.11 This is a comparison chart of the robustness of the unmanned collaborative perception network under random attacks on nodes before and after optimization in the target tracking task provided by the embodiment of the present invention. It can be seen from the figure that in addition to the distributed network structure under ideal conditions, the optimal network structure considering heterogeneity has better robustness when facing the target tracking task. Fig.12 This is a comparison chart of the average path lengths of the unmanned collaborative sensing network under random attacks on nodes before and after optimization under the target tracking task provided by the embodiment of the present invention. It can be seen from the figure that in addition to the distributed network structure under ideal conditions, the optimal network structure considering heterogeneity has a lower average path length when facing the target tracking task.

[0043] In summary, the present invention selects system robustness, information timeliness and energy loss as optimization objectives, adjusts the weight of the objective function according to different task requirements, and proposes an optimization method for an unmanned perception network structure based on an improved secretary vulture algorithm. The upper triangle coding method is used to form a secretary vulture, and the Tent chaotic mapping is introduced in the population initialization of the algorithm. During the iteration process, Gaussian and Cauchy mutation perturbations are introduced. Through the greedy strategy, the fitness value is updated to accelerate the convergence of the algorithm, and finally the optimal heterogeneous unmanned collaborative perception network structure is obtained, and the network structures before and after optimization are compared and analyzed. This command and control system network structure optimization method based on the improved secretary vulture algorithm is concise and efficient, and can quickly obtain a heterogeneous unmanned collaborative perception network structure with the highest robustness and timeliness and the lowest energy loss under limited network density and degree constraints.

[0044] Although the present invention has been described above with preferred embodiments, it is not intended to limit the present invention. A person skilled in the art of the present invention may make various modifications and improvements without departing from the spirit and scope of the present invention. Therefore, the scope of protection of the present invention shall be determined by the claims.

Claims

1. A method for optimizing network structure based on improved secretary vulture algorithm, characterized in that: The following steps are involved: S1. According to the heterogeneous collaborative perception network structure with and without human, a multi-objective optimization model is established based on the node degree and the average value of the inverse of the shortest path between node pairs. Specifically, objective functions are constructed to maximize robustness, maximize information timeliness, and minimize energy loss. Then, the weight coefficient of the objective function is adjusted according to the task requirements, and the constraint conditions of the objective function are added according to the characteristics of the heterogeneous collaborative perception network structure with and without human. S2. Initialize the parameters in the improved secretary vulture algorithm; for the adjacency matrix of the heterogeneous collaborative sensing network structure with or without humans, use the upper triangle coding method to form individuals in the population; In the population initialization stage, the Tent chaotic mapping is introduced to increase the diversity of the population; S3, using the improved secretary vulture algorithm to solve the multi-objective optimization model, including the following steps: S3.1, start iteration, and update the individual positions of secretary bird through its three hunting strategies; S3.2, during the iteration process, the positions of the individuals are updated and their fitness values ​​are updated through the two escape strategies of the secretary bird; S3.3, introduce Gaussian and Cauchy mutation disturbances, update the fitness value, help the algorithm jump out of the local optimum, and continue to search for the global optimal solution; S3.4, based on the fitness value in S3.3, introduce a greedy strategy to guide the algorithm to keep tracking the optimal solution during the search process by comparing the fitness values; S3.

5. Determine whether the maximum number of iterations is met. If so, the heterogeneous collaborative perception network structure with or without personnel corresponding to the individuals in the iterated population is output as the most optimal collaborative perception network structure with or without personnel. Otherwise, execute step S3.

1.

2. According to claim 1, a method for optimizing network structure based on improved secretary vulture algorithm for presence or absence perception is characterized in that: In step S1, based on the degree of the node k i , the objective function to maximize the robustness R is constructed as follows: , , In the formula, I i For Node i The importance of k i For the i is the degree of a node, and N is the total number of nodes.

3. The method for optimizing network structure based on improved secretary vulture algorithm for presence or absence perception according to claim 1 is characterized in that: In step S1, based on the average value T of the inverse of the shortest path length between node pairs, an objective function with the goal of maximizing information timeliness is constructed as follows: ， In the formula, d ij For Node i and nodes j The shortest path length between .

4. The method for optimizing network structure based on improved secretary vulture algorithm for presence or absence perception according to claim 1 is characterized in that: In step S1, based on the distance between nodes, the objective function with the minimum energy loss as the goal is constructed as follows: ， Where, d is the distance between two communication nodes; E elec Represents the energy required to receive and send a unit of data; E fs represents the energy dissipation coefficient of the free space mode circuit; E mp Represents the energy consumption coefficient of the circuit in multipath fading mode.

5. The method for optimizing network structure based on improved secretary vulture algorithm for presence or absence perception according to claim 1 is characterized in that: In step S1, the constraints of the objective function are as follows: Constraints on edges in heterogeneous collaborative sensing networks: ， ， In the formula, W C-R represents the number of edges between the command node and the combat soldiers, W R-R represents the number of edges between combat soldiers, W R-UGV represents the edge between combat soldiers and unmanned vehicles, W R-UAV Represents the edge between combat soldiers and drones; The constraints on the degree of nodes in the heterogeneous unmanned collaborative sensing network structure are: , , In the formula, k person Indicates the degree of commander and combat soldiers. k non-person represents the speed of drones and unmanned vehicles, k const Indicates node saturation; Constraints on network density in heterogeneous collaborative sensing network structures: , In the formula, D represents the network density, D MAX Indicates the maximum network density threshold; Constraints on energy loss in heterogeneous collaborative sensing network structures: ， In the formula, E i Representation Node i energy loss; E max Representation Node i Maximum energy capacity; Adjacency matrix in heterogeneous collaborative sensing network structure A The value constraints are: ， In the formula, the adjacency matrix of heterogeneous collaborative sensing network structure is A Medium Element a ij , a value of 1 indicates that the nodes are connected, and a value of 0 indicates that the nodes are not connected.

6. The method for optimizing network structure based on improved secretary vulture algorithm for presence or absence perception according to claim 1, characterized in that: The parameters in the improved secretary vulture algorithm initialized in step S2 include: population size P , number of iterations G , chaos coefficient t , positional parameters μ , scale parameter γ, degrees of freedom v ,parameter r .

7. The method for optimizing network structure based on improved secretary vulture algorithm for presence or absence perception according to claim 1 is characterized in that: The three hunting strategies of the secretary bird in step S3.1 are divided into three time intervals, namely: , , , corresponding to the three stages of predation by secretary birds: finding prey, consuming prey and attacking prey; When the number of iterations When , the formula is as follows: ， In the formula, t represents the current number of iterations, T represents the maximum number of iterations, Indicates i The new state of the secretary bird in the stage of searching for prey, and is a random candidate solution for the iteration of the prey search phase; R1 represents the dimension randomly generated in the interval [0,1] An array of is the dimension of the solution space; represents the value of its j-th dimension, Represents the fitness value of its objective function; When the number of iterations When , the formula is as follows: ， In the formula, RB represents the dimension randomly generated from the standard normal distribution. Array ，x best Indicates the current optimal value; When the number of iterations When , the formula is as follows: ， In the formula, RL is the flight step length in Levy flight, t is the current number of iterations, and T is the maximum number of iterations. Indicates i The new state of the secretary bird in the stage of searching for prey, Represents the fitness value of its objective function.

8. The method for optimizing network structure based on improved secretary vulture algorithm for presence or absence perception according to claim 1 is characterized in that: In step S3.2, the positions of the individuals are updated by using the two escape strategies of the secretary bird. The two escape strategy formulas of the secretary bird are as follows: , Among them, R2 means randomly generating dimensions from a normal distribution Array of x random represents the random candidate solution of the current iteration, and K represents a random selection of an integer 1 or 2; The fitness function in step S3.2 is as follows: ， In the formula, fit is the objective function, Fit is the fitness value, Ave_d ij is the number of path hops between nodes, R is the system robustness, T is the information timeliness, W ij For Node i With Node j The energy loss between w 1. w 2. w 3. And w 4 represent the weight coefficients of the path hop count, system robustness, information timeliness and energy loss between nodes, respectively. is the penalty for violating the constraint.

9. The method for optimizing network structure based on improved secretary vulture algorithm for presence or absence perception according to claim 1, characterized in that: In step S3.3, Gaussian and Cauchy variation disturbances are introduced, and the formula is as follows: ， In the formula, is the updated solution, is subject to the mean of 0 and the variance of The normal distribution of r i Obey positional parameters , scale parameter , degrees of freedom The Cauchy distribution of and are the weight coefficients of Gaussian distribution and Cauchy distribution respectively, t and T are the current number of iterations and the maximum number of iterations respectively.

10. The method for optimizing network structure based on improved secretary vulture algorithm for presence or absence perception according to claim 1, characterized in that: It also includes an analysis of the performance of the most optimal unmanned collaborative sensing network structure, including robustness, timeliness, connectivity, average path length, average betweenness centrality, clustering coefficient, and association closeness, as well as robustness, timeliness, and average path length under node attacks.

Images