CN-122001790-A - Packet beehive control method based on connectivity estimation

CN122001790ACN 122001790 ACN122001790 ACN 122001790ACN-122001790-A

Abstract

The invention discloses a grouping beehive congestion control method based on connectivity estimation, which comprises the steps of constructing a cluster comprising a plurality of self-bodies and dividing the cluster into a plurality of sub-clusters, constructing a global network and a local network based on a communication structure among the self-bodies, respectively utilizing two PI average consistency estimators to estimate global connectivity and local connectivity of the sub-clusters to which the self-bodies belong on line by the self-bodies, constructing a connectivity maintenance potential function by combining two connectivity estimation values and calculating gradients of the connectivity maintenance potential function, simultaneously taking algebraic connectivity of the global network and algebraic connectivity of the local network into acquisition of neighbor information of sub-clusters, and calculating a control law based on the neighbor information of the sub-clusters to drive the self-bodies to realize beehive congestion in the sub-clusters and maintain necessary connectivity among different sub-clusters. The invention can effectively realize the distributed type beehive congestion control of multiple groups and ensure the dual constraint of global and local network connectivity.

Inventors

LIU ZHONGCHANG
LIU JUNHAN
LI LONGTAO
YUE WEI
LI LILI

Assignees

大连海事大学

Dates

Publication Date: 20260508
Application Date: 20260320

Claims (8)

1. A method for controlling packet congestion based on connectivity estimation, comprising: s1, constructing a group of clusters comprising a plurality of self-bodies; s2, dividing the clusters to obtain a plurality of sub-clusters; S3, constructing a global network of the cluster and a local network of the subset group according to a communication structure between the self-body and the self-body; S4, each main body estimates the global network through a first PI average consistency estimator to obtain a global connectivity estimation value of the global network, and simultaneously, each main body estimates the local network of the sub-cluster to which the main body belongs through a second PI average consistency estimator to obtain a connectivity estimation value of the local network of the sub-cluster to which the main body belongs; S5, calculating a connectivity maintenance potential function according to the global connectivity estimation value of the global network and the connectivity estimation value of the local network of the sub-group to which the autonomous body belongs, so as to obtain the gradient of the connectivity maintenance potential function, and acquiring the information of all the autonomous bodies except the same sub-group in the autonomous communication range based on algebraic connectivity of the global network and the local network; and S6, calculating a control law based on the information of all self-bodies except the same sub-group in the self-body communication range so as to realize the grouping beep congestion control.
2. The method for controlling packet congestion based on connectivity estimation according to claim 1, wherein the self-body's kinetic equation is: (1) in the formula, Is self-body Is a position of (2); Is self-body Is a speed of (2); Is self-body Control input of (a); Is self-body Is a transient speed of (2); Is self-body Is a transient acceleration of (a); The communication strength of the communication structure between the autonomous bodies is represented by weight, and the expression is: (2) in the formula, Is the weight of the communication intensity between the autonomy; Is self-body And an autonomous body A distance therebetween; Is the maximum communication distance, and ; Is a design parameter.
3. The method for controlling packet congestion based on connectivity estimation according to claim 2, wherein the specific step of estimating the global network and the local network of the sub-cluster to which the autonomous body belongs by the respective bodies through the PI average consistency estimator is as follows: S41, defining a first PI average consistency estimator/a second PI average consistency estimator as follows: (3) in the formula, Input for respective subject An estimate of the average value of (2); Is self-body A neighbor set within a communication range; Estimating a hierarchy for the subset group; Is self-body Is a local input of (a); is an internal variable of the estimator; Gain for normal number; Is that The time derivative of (a), i.e. the rate of change of the average estimate; Is that I.e. the rate of change of the internal variables of the estimator; S42, setting the state of the first PI average consistency estimator as And sets the state of the second PI average consistency estimator to Wherein, the method comprises the steps of, For the estimate output by the first PI mean consistency estimator, An internal variable that is a first PI mean consistency estimator; for the estimate output by the second PI mean consistency estimator, An internal variable that is a second PI mean consistency estimator; S43, will be from the host Is set to By substituting the first PI average consistency estimator, an autonomous body is obtained Wherein, Is self-body A feature vector component value at the current time; S44, will be from the host Is set to Obtaining an autonomous body by substituting the second PI average consistency estimator Wherein, Is self-body The square value of the eigenvector component at the current time; S45, according to the autonomous body Global mean estimation and autonomous of (a) Is combined with the global square mean estimation of (1) Received neighbor autonomous information, for an autonomous Is updated, the expression is: (4) in the formula, Is self-body The updated internal state; Estimating a global mean value; Estimating a global square average value; Is self-body Is an internal state of (2); is a neighbor self-body Internal states in the information; Is a normal number gain, and And Greater than And Less than 、、 ; S46, according to the autonomous body And (3) calculating a general connectivity estimation value by the updated internal state and global square average value estimation, wherein the expression is as follows: (5) in the formula, The general connectivity estimation value; Is the number of sub-clusters; Is a global network hierarchy; is a local network hierarchy; When (when) Obtaining the global connectivity estimation value ; When (when) Obtaining connectivity estimation value of local network of the sub-group to which the autonomous body belongs 。
4. A method of packet congestion control based on connectivity estimation according to claim 3, wherein the connectivity maintenance potential function is expressed as: (6) in the formula, Maintaining a potential function for connectivity; Is constant and is smaller than a preset threshold value; Is the exclusion range; limiting the range for the packet; Gain for normal number; the estimated value is the global connectivity; connectivity estimates for the local networks of the sub-group to which the autonomous belongs.
5. The method for packet congestion control based on connectivity estimation according to claim 4, wherein the expression of the control law is: (7) in the formula, Is a control law; Is a beehive control item; separating items for the alien group from the main body; Keep and keep collision avoidance items for connectivity.
6. The method of claim 5, wherein the congestion control term is defined as: (8) in the formula, To be at autonomous An autonomous set within communication range and belonging to the same sub-cluster as it, wherein, Is the first A sub-cluster; Is the relative distance between two autonomous bodies; Representative function Wherein, the gradient of the (c) is that, Is a non-negative potential energy function of limited magnitude with respect to the parameter z; Is self-body Is a function of the speed of the machine.
7. The method of claim 6, wherein the disparate group autonomous discrete items are defined as: (9) in the formula, To be at autonomous Communication range A set of all self-principals within, except for the same subset group; Is a vector of m dimensions; Is self-body Is a position of (c).
8. The method for packet congestion control based on connectivity estimation according to claim 7, wherein the connectivity maintenance and collision avoidance term is defined as: (10) in the formula, The gradient of the potential function is maintained for connectivity.

Description

Packet beehive control method based on connectivity estimation Technical Field The invention relates to the technical field of bee congestion control of a multi-intelligent system cluster, in particular to a packet bee congestion control method based on connectivity estimation. Background The bee congestion control aims at enabling a large-scale multi-autonomous system to emerge orderly collective behaviors through simple local interaction among individuals, and the concept is inspired by phenomena of bird migration, fish swarm tour, insect aggregation and the like in the nature at the earliest. Although each individual in the buzzing clusters has limited capabilities in terms of perception, movement, calculation and the like, the buzzing behaviors such as large-scale cluster aggregation, mutual collision prevention, consistent speed and the like can be realized through a large amount of information exchange between individual members and neighbor members. With the complexity of application scenarios, the large-scale buzzing behavior of a single group is gradually difficult to cope with complex scenarios comprising a plurality of tasks with coupling relations, for example, a plurality of unmanned system clusters are used for covering and searching different areas, and the searched information needs to be shared among all sub-clusters by keeping a desired preset formation. Therefore, aiming at an autonomous system formed by a plurality of sub-clusters, a grouping beeping control algorithm is designed to complete the control targets of the same sub-cluster aggregation and orderly separation of different sub-clusters, and the method has wider application value. In a multi-host cluster with limited communication range, the existing Olfati-Saber algorithm model is the basis for realizing single cluster congestion control, and when the multi-cluster grouping congestion control is realized based on the Olfati-Saber algorithm, the existing method mainly has two ideas, namely, one is to assume that all self-hosts in each sub-cluster exactly know the target position of the cluster, so that the congestion aggregation of each sub-cluster is realized by tracking the global information. However, there will be great difficulty and uncertainty in sharing global information in real time in large-scale clusters under limited communication range constraints. Another idea is to assume that all the self-bodies of each sub-cluster have connectivity in the initial state, then use a control policy to keep the self-bodies of the initial connection within a communication range (i.e. maintain each existing communication link) and then use an aggregation control algorithm to implement aggregation of the sub-clusters, but this method of maintaining each communication link does not allow the self-bodies of different sub-clusters to exceed the communication range, which makes it difficult for different sub-clusters to implement the desired formation structure. Disclosure of Invention The invention provides a grouping beehive congestion control method based on connectivity estimation, which aims to solve the problems that the real-time sharing of global information in a large-scale cluster is difficult and uncertain, and the expected formation structure of different subsets is difficult to realize. In order to achieve the above object, the technical scheme of the present invention is as follows: a method of packet congestion control based on connectivity estimation, comprising: s1, constructing a group of clusters comprising a plurality of self-bodies; s2, dividing the clusters to obtain a plurality of sub-clusters; S3, constructing a global network of the cluster and a local network of the subset group according to a communication structure between the self-body and the self-body; S4, each main body estimates the global network through a first proportional-integral (PI) average consistency estimator to obtain a global connectivity estimation value of the global network; meanwhile, each main body estimates the local network of the sub-cluster to which the main body belongs through a second PI average consistency estimator to obtain a connectivity estimation value of the local network of the sub-cluster to which the main body belongs; S5, calculating a connectivity maintenance potential function according to the global connectivity estimation value of the global network and the connectivity estimation value of the local network of the sub-group to which the autonomous body belongs, so as to obtain the gradient of the connectivity maintenance potential function, and acquiring the information of all the autonomous bodies except the same sub-group in the autonomous communication range based on algebraic connectivity of the global network and the local network; and S6, calculating a control law based on the information of all self-bodies except the same sub-group in the self-body communication range so as to realize the grouping beep congestio