CN114363340A - Unmanned aerial vehicle cluster failure control method and system and storage medium - Google Patents
Unmanned aerial vehicle cluster failure control method and system and storage medium Download PDFInfo
- Publication number
- CN114363340A CN114363340A CN202210033514.1A CN202210033514A CN114363340A CN 114363340 A CN114363340 A CN 114363340A CN 202210033514 A CN202210033514 A CN 202210033514A CN 114363340 A CN114363340 A CN 114363340A
- Authority
- CN
- China
- Prior art keywords
- node
- state
- communication
- failure
- nodes
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 44
- 238000004891 communication Methods 0.000 claims abstract description 110
- 230000006855 networking Effects 0.000 claims abstract description 14
- 230000008569 process Effects 0.000 claims abstract description 10
- 230000002093 peripheral effect Effects 0.000 claims abstract description 4
- 238000012544 monitoring process Methods 0.000 claims description 46
- 238000009826 distribution Methods 0.000 claims description 29
- 230000007246 mechanism Effects 0.000 claims description 15
- 238000002360 preparation method Methods 0.000 claims 1
- 230000000694 effects Effects 0.000 abstract description 14
- 230000002411 adverse Effects 0.000 abstract description 5
- 230000009286 beneficial effect Effects 0.000 abstract description 2
- 230000009467 reduction Effects 0.000 abstract description 2
- 239000013260 porous coordination network Substances 0.000 description 32
- 230000000875 corresponding effect Effects 0.000 description 27
- 238000004422 calculation algorithm Methods 0.000 description 11
- 238000012423 maintenance Methods 0.000 description 10
- 238000004364 calculation method Methods 0.000 description 7
- 238000012546 transfer Methods 0.000 description 6
- 230000005540 biological transmission Effects 0.000 description 5
- 238000011156 evaluation Methods 0.000 description 4
- 230000007704 transition Effects 0.000 description 4
- 238000004458 analytical method Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 3
- 230000002596 correlated effect Effects 0.000 description 3
- 230000001186 cumulative effect Effects 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 230000008859 change Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 238000011084 recovery Methods 0.000 description 2
- 238000010521 absorption reaction Methods 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000010835 comparative analysis Methods 0.000 description 1
- 230000001276 controlling effect Effects 0.000 description 1
- MKXZASYAUGDDCJ-NJAFHUGGSA-N dextromethorphan Chemical compound C([C@@H]12)CCC[C@]11CCN(C)[C@H]2CC2=CC=C(OC)C=C21 MKXZASYAUGDDCJ-NJAFHUGGSA-N 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 230000014759 maintenance of location Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 238000004088 simulation Methods 0.000 description 1
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1004—Server selection for load balancing
- H04L67/1008—Server selection for load balancing based on parameters of servers, e.g. available memory or workload
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/10—Protocols in which an application is distributed across nodes in the network
- H04L67/1001—Protocols in which an application is distributed across nodes in the network for accessing one among a plurality of replicated servers
- H04L67/1034—Reaction to server failures by a load balancer
Landscapes
- Engineering & Computer Science (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Computer Hardware Design (AREA)
- General Engineering & Computer Science (AREA)
- Data Exchanges In Wide-Area Networks (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
The invention relates to a method, a system and a storage medium for controlling cluster failure of an unmanned aerial vehicle, belonging to the technical field of unmanned aerial vehicle cluster failure control; the control method comprises the following steps: in the unmanned aerial vehicle networking process, each newly added node of the unmanned aerial vehicle is in communication connection with a node which can enable the average communication bearing pressure to be minimum in the existing unmanned aerial vehicle networking; classifying the node states of the unmanned aerial vehicle based on the communication bearing pressure indexes, wherein when the node fails, the nodes with strong residual bearing capacity preferentially carry out redundant bearing pressure when the bearing node fails; when the node fails, the peripheral nodes with large residual communication bearing capacity in the failure area are obtained to replace the failed nodes. The cascade failure control method based on the minimum load principle can reduce the adverse influence range caused by failure. In addition, the reduction of failure scale is beneficial to the network to maintain stability and the task execution efficiency, and the service pressure of unmanned aerial vehicle cluster maintainers can be reduced on the basis of ensuring the efficiency and the effect.
Description
Technical Field
The invention relates to a method and a system for controlling cluster failure of an unmanned aerial vehicle and a storage medium, and belongs to the technical field of unmanned aerial vehicle cluster failure control.
Background
An unmanned aerial vehicle cluster is a network which is formed by networking a plurality of unmanned aerial vehicles and keeps intelligent cooperation through mutual communication so as to realize a certain specific task. Cascading failures are common, adverse chain reactions are caused by various risks in the operation of an unmanned aerial vehicle cluster, and the scale of the failures can affect the overall stability of a cluster network and the task execution efficiency. Therefore, by exploring a more advanced cluster networking technology and exploring an unmanned aerial vehicle cluster failure control mode by taking cascade failure as a carrier, the method plays a vital role in improving the operation efficiency and effect of the unmanned aerial vehicle cluster.
Disclosure of Invention
In a first aspect, the application provides a control method for unmanned aerial vehicle cluster failure, which is used for analyzing topological characteristics of nodes and connecting edges to optimize a failure propagation path, and finally reducing failure scale, and can improve stability of a cluster network to a certain extent and maintain cluster task execution efficiency.
In the unmanned aerial vehicle networking process, each unmanned aerial vehicle is defined as a node, and the unmanned aerial vehicles are in communication connection with each other to be regarded as a connecting edge;
each newly added node of the unmanned aerial vehicle is in communication connection with a node which can enable the average communication bearing pressure to be minimum in the existing unmanned aerial vehicle networking;
classifying the node states of the unmanned aerial vehicle based on the communication bearing pressure indexes, wherein when the node fails, the nodes with strong residual bearing capacity preferentially carry out redundant bearing pressure when the bearing node fails;
when the node fails, the peripheral nodes with large residual communication bearing capacity in the failure area are obtained to replace the failed nodes.
In some disclosures, the average communication bearing pressure comprises a communication bearing pressure PCN of a node and a communication bearing pressure PCE of a connecting edge; the communication bearing pressure PCN of the node is as follows:
in the formula (1), PCNlRefers to the average communication bearing pressure, N, of any node l in the network1Representing the number of all other nodes directly connected to the node l; n is a radical of2Representing the number of all other nodes indirectly connected to node l; n refers to the total number of nodes in the unmanned aerial vehicle cluster network (N)1+N2+1 ═ n); deg (i) denotes the value of the corresponding node i, PCNlThe communication pressure that node l can bear;
representing the number of other nodes (including the node j) passing through the shortest path from the node j to the node l, wherein the node j is not directly connected with the node l;
the definition of the communication bearing pressure of the connecting edge is shown as a formula (2);
PCE in equation (2)ijRefers to the bearing pressure, lambda, of a communication connection edge formed by any node i and node j in the network0,λ1,λ2Respectively representing the adjustment coefficients, deg (n), in three different casesi) Then it is node niThe value of (a) is determined,
measure node niAnd node njThe difference of status degree of corresponding connecting edges in all connecting edges of two parties, | PCNi-PCNjI then measure node niAnd node njThe magnitude of the difference between the communication bearer pressures.
In some disclosures, the node states of the drone are classified as: high-order state H, medium-order state M and primary state L, QmaxRepresenting the maximum load capacity of the node,
and
respectively represent the node loads for the non-monitoring task and the monitoring task, then
The node state is at a high level when
The node state is in the intermediate state when
The node state is in the primary state.
In some disclosures, performing node state switching based on a minimum load principle includes: and unidirectional switching is performed in sequence from the directions of the primary state L, the intermediate state M and the advanced state H.
In some disclosures, the high-level state initially fails: finding out a node with the minimum PCN value around the initial failure node, and converting the node from the intermediate state M to the high state H; judging whether the residual bearing capacity of the node can completely absorb the communication pressure released after the original advanced state H node fails, if so, failing and not transmitting to the periphery, and ending the failure; otherwise, the communication pressure is continuously transmitted to the surroundings until the residual bearing capacity of a certain node can completely absorb the released communication pressure; after the failure is finished, the PCN values of all the nodes are redistributed, and the network reaches a stable operation state again;
initial failure in the intermediate state: when a primary state L node directly connected with an initially failed intermediate state M node exists, converting the state of the primary state L node corresponding to the minimum PCN value into an intermediate state M; otherwise, the state switching does not exist, the intermediate state M node containing the minimum PCN value around the initial failure intermediate state M node becomes a substitute node until the residual bearing capacity of a certain node can completely absorb the released communication pressure; after the failure is finished, the PCN values of all the nodes are redistributed, and the network reaches a stable operation state again;
initial failure of primary state: when the primary state L node initially fails, the intermediate state M or the high-level state H node adjacent to the primary state L node directly absorbs the smaller communication bearing pressure originally born by the primary state L node.
In some disclosures, the method comprises cluster network generation and is used for carrying out communication bearing pressure distribution of all nodes and monitoring range distribution of each node;
the cluster network is generated by acquiring a communication bearing pressure index and a minimum load principle.
In some disclosures, the cluster network generation mechanism includes:
initializing communication bearing pressure distribution of all nodes and monitoring range distribution of each node;
the method comprises the steps of performing initial connection, wherein newly-added nodes prefer to be connected with nodes with small communication bearing pressure in existing members, and after connection, the communication bearing pressure distribution of all nodes and the monitoring range distribution of each node are recalculated;
and networking the next node, and establishing connection between the newly added node and the updated node with low communication bearing pressure in the existing members until the monitoring network is generated.
In some disclosures, the stability of the communication network before and after failure and the loss of the monitoring area of the network before and after failure are obtained, and the stability of the monitoring network and the reliability of executing tasks are judged.
In a second aspect, the present application provides a storage medium, configured to analyze topological features of nodes and connecting edges to optimize a failure propagation path, and finally reduce a failure scale, so as to improve stability of a cluster network to a certain extent, and maintain efficiency of executing a cluster task.
A storage medium having a program stored therein, the program, when executed, implementing the control method of the first aspect.
In a third aspect, the application provides a control system for unmanned aerial vehicle cluster failure, which is used for analyzing topological characteristics of nodes and connecting edges to optimize a failure propagation path, and finally reducing failure scale, so that the stability of a cluster network can be improved to a certain extent, and the execution efficiency of cluster tasks is maintained.
The control system for the unmanned aerial vehicle cluster failure comprises a storage medium, a program is stored in the storage medium, and the control method of the first aspect is realized when the program is executed.
Has the advantages that:
the cascade failure control method based on the minimum load principle can effectively reduce the adverse influence range caused by failure and reduce economic loss. In addition, the reduction of failure scale is beneficial to network maintenance stability and task execution efficiency, and the service pressure of unmanned aerial vehicle cluster maintainers can be reduced on the basis of ensuring efficiency and effect, so that light-weight operation and maintenance are realized.
The attached drawings of the specification:
FIG. 1 is a diagram of the implementation path and main ideas of the present patent;
FIG. 2 is a schematic diagram of an example of a key generation process of a cluster network of unmanned planes consisting of 10 unmanned planes;
FIG. 3 is a comparison graph of network structure changes before and after a cluster network fails based on four failure situations in a minimum load mode;
FIG. 4 is a comparison graph of network structure changes before and after cluster network failure based on four failure situations in a random mode;
FIG. 5 is a variation of PCE distribution in four failure scenarios in two failure modes;
FIG. 6 is PCN distributions before and after failure of nodes in the high level H, medium level M and primary level L in the minimum load mode;
FIG. 7 is PCN distribution before and after failure of nodes in high-level H, medium-level M and primary-level L in a random mode;
FIG. 8 shows PCE distribution under two failure mechanisms and their comparison before and after failure when a node in the same state fails;
FIG. 9 is a graph comparing the difference in the effect of four failure situations in two failure modes on the overall stability maintenance degree of the network;
FIG. 10 is a graph comparing the effect of different radii and different numbers of L-shaped nodes on the monitored area;
fig. 11 is a schematic view of a monitoring area of a cluster network.
Detailed Description
The technical solutions in the embodiments of the present disclosure will be clearly and completely described below with reference to the drawings in the embodiments of the present disclosure, and it is obvious that the described embodiments are only a part of the embodiments of the present disclosure, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments disclosed herein without making any creative effort, shall fall within the protection scope of the present disclosure.
The present invention is illustrated by the following five steps. Firstly, defining the communication bearing pressure indexes of nodes and connecting edges from the aspects of topology and complex network characteristics; classifying the node states in the network operation based on the communication bearing pressure index, and giving out the rule of switching the node states when failure occurs; thirdly, defining a cluster network generation mechanism based on a communication bearing pressure index and a minimum load principle; fourthly, aiming at topological characteristics and different failure types of the cluster, failure control methods under different conditions are designed by taking the minimum load as a principle; fifthly, establishing an evaluation index capable of checking the efficiency of the algorithm, and comprehensively evaluating the algorithm by combining a comparison group.
Specifically, the step is to define the communication bearing pressure indexes of the nodes and the connecting edges in the network. The core principle of the network generation mechanism based on the minimum load is to ensure that in the networking process of the unmanned aerial vehicle, a newly added node every time is selected to establish communication connection with a node which can enable the average communication bearing pressure to be minimum in the existing network, and further ensure that the overall potential risk resisting capability of the network is maximized. Two key indexes of the communication bearing pressure PCN of the node and the communication bearing pressure PCE of the connecting edge are involved. And the communication pressure borne by the nodes in the network is calculated in the way shown in formula (1).
In the formula (1), PCNlThe calculation idea is that the communication bearing pressure of any node l in the network is weighted and averaged and added by all corresponding node members in a node set directly connected with the node l and a node set indirectly connected with the node l according to illumination distribution (degree distribution of a certain node in the network is the number of edges directly connected with the node). N is a radical of1Representing all other nodes directly connected to node lAnd (6) counting the number of points. N is a radical of2Representing the number of all other nodes indirectly connected to node l. N refers to the total number of nodes in the unmanned aerial vehicle cluster network (N)1+N2+1 ═ n). deg (i) denotes the value of the corresponding node i, PCNlRefers to the communication pressure that node l can withstand.
And representing the number of other nodes (including the node j) passing through the shortest path from the node j to the node l when the node j is not directly connected with the node l.
On the other hand, the communication bearing pressure of the connected edges is influenced by the communication bearing pressure of the corresponding nodes on both sides, and the definition is shown in formula (2).
PCE in equation (2)ijRefers to the bearing pressure, lambda, of a communication connection edge formed by any node i and node j in the network0,λ1,λ2Respectively representing the adjustment coefficients, deg (n), in three different casesi) Then it is node niThe value of (a) is determined,
measure node niAnd node njThe difference of status degree of corresponding connecting edges in all connecting edges of two parties, | PCNi-PCNjI then measure node niAnd node njThe magnitude of the difference between the communication bearer pressures. Essentially, the size of the PCE value reflects the ease and pressure of communication between any two nodes. The larger the PCE value is, the more frequent and the more stressed communication between the corresponding nodes is indicated, otherwise, the potential communication capability of the connection edge is not fully utilized.
As can be seen from the formula (2), the communication bearing pressure of the connecting edge is affected by the communication bearing pressure of the corresponding node, and when the values of the corresponding nodes (the values of the corresponding nodes are unified and default to the values of the network nodes, which are equal in value to the number of all connecting edges directly establishing communication connection with the corresponding node) are equal, the PCE value is positively correlated with the absolute value of the PCN difference value of the corresponding node; when PCNs of the corresponding nodes are equal, the PCE value is positively correlated with the absolute value of the degree difference value of the corresponding node; when the degree value and the PCN value of the corresponding node are not equal, the PCE value is positively correlated with the absolute value of the degree difference value of the corresponding node and the absolute value of the PCN difference value; and when the value of the corresponding node and the PCN are equal, the PCE value can be equal to the minimum value of PCEs corresponding to other nodes.
And the key objective of the second step is to classify the node states based on the communication bearing pressure indexes and to provide a state switching rule based on a minimum load principle. Specifically, the reason for classifying the states of the nodes of the unmanned aerial vehicle is to distinguish different functions of different types of nodes in the cluster network, fully exert the potential capability of each node according to the actual operation current situation, and finally ensure that the cluster network maintains high-level stability. The classification of the nodes is based on the topological environment in which the nodes are located in the network and the potential capacity of the nodes. Specifically, each node may be in three states, a high-level state (H), a medium-level state (M), and a low-level state (L). Considering that the cluster drone architecture is homogeneous, i.e. the maximum load capacity of all drones is the same, the defining criteria for the three states can be determined by the load ratio for the non-monitoring tasks and the relative size of 1, if QmaxWhich represents the maximum load capacity of the load,
and
(wherein
Constant) and represents the load for the non-monitoring task and the monitoring task, respectively, then when
The node state is at a high level when
The node state is in the intermediate state when
The node state is in the primary state. Equation (3) gives a method for calculating the node status in detail from the perspective of the communication topology, considering that the communication topology may affect the (non-) monitoring task load distribution of the network node members. In addition, the characteristics and differences of the three states are shown in table 1 below.
TABLE 1 State characteristics of the Classification nodes
As shown in table 1, since the load capacity of the node is constant, the high-level state has a minimum remaining communication load-bearing capacity because it performs a plurality of functions including control, relay, and task, and since there are many functions, the load on the monitoring task is small, and thus the monitoring radius of a single node is small. In contrast, the primary state has relatively large monitoring radius and residual pressure bearing capacity due to single function. And each index of the intermediate state node is between the high state and the low state.
A concrete method for judging the node state is given by formula (3), wherein state (n)k) Representing a node n in a networkkThe state of the device. α, β, γ represent the state adjustment coefficients of the high order state, the middle order state and the primary order state, deg (n)k) Representative node nkN, representing the total number of nodes in the network,
is the number of permutation combinations representing the remaining N-1 nodes from the networkAll shortest paths that can be formed between any two points.
Then represents all shortest paths through node nlThe total number of (c). It can be seen that the state of a certain node is determined by two aspects, namely the degree distribution of the node and the number of times that the node is traversed by the shortest path between any two remaining nodes. The two measures the corresponding characteristics of the node under different states from the local and overall angles respectively.
On the other hand, the state of each node can be switched according to the actual running state, so that the functional stability of the network facing an uncertain external environment is ensured. In particular, the two principles of state switching are to fully utilize the remaining communication carrying capacity and to consider the conversion cost, respectively. The specific handover priorities are shown in table 2 below.
TABLE 2 transition priority between different states
As can be seen from table 2, the direction of state switching is unidirectional, i.e. only switching from the low-level state to the high-level state, because the remaining bearing capacity of the low-level state node is always stronger than that of the high-level state node, and such switching helps to fully exploit the potential capacity of the node, thereby maintaining the continuity and stability of the network. In addition, because certain switching cost exists in the cross-level mode among the nodes in different states, when the nodes in different states initially fail, corresponding state switching priorities are different.
And step three, defining a cluster network generation mechanism based on the communication bearing pressure index and the minimum complexity principle. Considering that networking and failure processes are essentially connection and disconnection of connection edges between nodes, the effect of adopting the failure control method based on the minimum load principle is better when the network generated based on the minimum load principle fails, and a specific communication network generation algorithm is as follows.
The algorithm shows several key processes of networking generation, and mainly has three stages. Stage 1: and initializing the communication bearing pressure distribution of all the nodes and the monitoring range distribution of each node. And (2) stage: for the first connection, the newly added node prefers to connect a node with low communication bearing pressure in the existing members (the node has high residual communication processing capacity and high risk absorption capacity in case of cascade failure), as shown in fig. 1. After connection, the communication bearing pressure distribution of all the nodes and the monitoring range distribution of each node are recalculated. And (3) stage: and carrying out networking of the next node, establishing connection between the newly added node and the updated node with low communication bearing pressure in the existing members, and so on until the monitoring network is generated.
And the fourth step aims at designing the failure control method under different conditions by taking the minimum load as the principle according to the topological characteristics of the cluster and different failure types. The basic principle of the failure transfer process based on the minimum load in the transfer direction and path is to fully utilize peripheral nodes with large residual communication carrying capacity in a failure area to replace failed nodes, so that the probability of failure continuing to transfer to the vicinity is reduced, the number of failure reaching the existing nodes is reduced, and the purpose of maintaining the stability of the network is finally achieved.
When the nodes in the three states are initially failed respectively, the failure transmission path and the failure characteristics of the nodes related to failure are analyzed according to the specific failure transmission rules and the differences of the failure transmission rules under different conditions. In addition, the advantage points of the minimum load failure transfer mode in three situations are explored through the setting of key indexes (such as the stability degree of network nodes and connecting edges, the loss condition of the monitoring area of the cluster network) and the comparative analysis of a control group (random mode).
Initial failure of the advanced state: initial failure of a node in the high-level state may involve the transition of a neighboring node from the medium-level state M to the high-level state H on a least-loaded basis, with the specific delivery mechanism pseudocode as shown above. It can be seen that there are three key stages after the failure of the high-level H node. One is to find the node with the minimum PCN value around the initial failure node and convert the node from the middle-level state M to the high-level state H. And secondly, judging whether the residual bearing capacity of the node can completely absorb the communication pressure released after the original advanced state H node fails, if so, failing and not transmitting to the periphery, and ending the failure. Otherwise, the transmission to the surroundings is continued until the remaining bearing capacity of a certain node can completely absorb the released communication pressure used. And thirdly, after the failure is finished, the PCN values of all the nodes are redistributed, and the network reaches a stable operation state again.
initial
Node n with fail # in advanced stateiFail to work
S is 1; # records the cumulative number of failed nodes this time
do
find
# found
# node j having the smallest PCN value among nodes adjacent to the nodex
compare
and
until
# comparison node
# new burdened communication bearer pressure due to state transition and the original remaining communication bearer pressure of the node until the former is smaller than the latter
end do
Residual communication bearing pressure of # adjacent subsequent failed node
dab＝0(a,b∈Fi) (ii) a network updated; # cancels the remaining communication connections between all failed nodes,
update to new steady state after # network failure
Initial failure in the intermediate state: compared with the initial failure situation of the high-level state node, the initial aging of the intermediate-level state node needs to consider whether the state transition process of the corresponding node exists, and the specific failure transmission mechanism is as shown in the above figure. It can be seen that, compared with the situation that the high-level node H initially fails, when the medium-level node M initially fails, whether state switching exists needs to be considered. When a primary state L node directly connected with an initially failed intermediate state M node exists, converting the state of the primary state L node corresponding to the minimum PCN value into an intermediate state M; otherwise, the state switching does not exist, the intermediate-level M node containing the minimum PCN value around the initial failure intermediate-level M node becomes a substitute node, and the failure situations of other key links and the high-level H node are consistent.
Initial failure of primary state: unlike the more complex decision logic when the high-level state H and medium-level state M nodes initially fail, the transfer mechanism when the primary state node initially ages is as shown in the above figure. It can be seen that when the initial-state L node initially fails, the intermediate-state M or high-state H node adjacent to the initial-state L node can directly absorb the smaller communication bearing pressure originally borne by the intermediate-state M or high-state H node.
initial
Node n with fail # in the initial stateiFail to work
S is 1; cumulative number of # recording failed node
if
When the then # exists and fails node niAdjacent high-level node njAt a time there is
# and initial failure node niSum of parts
if
When the then # exists and fails node niAdjacent intermediate node njAt a time there is
# and initial failure node niSum of parts
PCNk ,wholeupdated, k is not equal to i and j; network updated; update the traffic bearer pressure of the network node members,
# simultaneous network reaching New Steady State
And fifthly, establishing an evaluation index capable of checking the efficiency of the algorithm, and comprehensively evaluating the algorithm by combining a comparison group. The stability maintenance degree reflects the degree that each node and connecting edge can be recovered to the state before failure after failure, and the larger the recovery degree is, the more stable the corresponding node and connecting edge are. The degree of stability of the communication network before and after failure can be measured by equations (4), (5) and (6). And the loss condition of the network monitoring area before and after the failure can be calculated by the formulas (7), (8) and (9).
Nj＝count(nj),state(nj)＝L (8)
In the formulas (4) to (9), η is a connecting edge e formed by all nodes i and j in the networkijThe recovery condition of the continuous communication bearing pressure after the network fails for a period of time and operates again reflects the change degree of the communication bearing pressure after the network fails, wherein the larger the eta is, the smaller the change degree is, and the higher the stability degree of the continuous edge is.
And
respectively refer to the failed front and rear connecting edges eijThe communication of (2) carries the pressure value, and
representing a failed front-to-back edge eijIs equal in value to the simple arithmetic mean of the actual variance of the PCE values of all the successive edges. In the aspect of task completion characterization of cluster network, rjRefers to the monitoring radius of any node j, R represents the monitoring radius of the cluster network, NjRepresenting the number of nodes in the primary state in the network, S (r)j,Nj) This indicates that when the monitored radius of a single node is rjThe number of nodes in the initial state is NjIt corresponds to the area of the area that the network can monitor.
The two indexes of the stability maintenance degree and the monitoring area loss condition can respectively reflect the buffer effect of the failure control method on the adverse effect of relieving the failure after the cascade failure occurs from two aspects of the stability of the monitoring network and the reliability of executing the task. Fig. 11 shows the monitoring mode of the drone cluster according to the node status classification and the corresponding functional characteristics. It can be seen that the nodes in the primary state have the least functions but the largest monitoring areas, the nodes in the advanced state are just opposite, and the monitoring areas of all the nodes are infinitely tangent and finally form the integral monitoring area of the cluster network.
In order to verify the technical effect of the above scheme, the present application is illustrated by the following embodiments:
the method comprises the following steps: in this example, a group of 10 monitoring clusters formed by the unmanned aerial vehicles is used as a research object to research the generation process, failure mechanisms and characteristics thereof under different situations. For the communication bearing pressure indexes of the defined nodes and the connecting edges, calculation of corresponding basic indexes and analysis and evaluation of key performance indexes can be carried out in the following links according to a formula (1) and a formula (2).
Step two: classifying the node states in network operation, and formulating rules for switching the node states after failure; according to the three principles of state switching provided by the formula (3), and by combining the basic index calculation modes in the formula (1) and the formula (2), the research basis of algorithm efficiency and effectiveness analysis can be provided for network generation and failure mechanism analysis in the subsequent steps.
Step three: based on the first step and the second step, a cluster network generation mechanism (as shown in fig. 2) may be defined based on the communication bearer pressure indicator and the minimum load principle, and the initial node and the connection edge communication bearer pressure in the cluster network are calculated at the same time as shown in table 1.
Table 1 PCN of initial node in minimum load mode and initial state and PCE distribution table thereof
Step four: and designing a failure control method under different situations by using the minimum load as a principle. After the cluster monitoring network is generated, four situations are respectively researched based on the three switching states of the cascade failure control method, namely high-level state H node initial failure, intermediate-level state M node initial failure, primary-level state L node initial failure (adjacent to the high-level state H node) and primary-level state L node initial failure (adjacent to the intermediate-level state M node). The following failure control method based on the minimum load mode respectively simulates the cluster network monitoring efficiency and the comprehensive performance under the four situations. The key indicators before and after failure in the four cases are shown in tables 2-4 below.
TABLE 2 PCN distribution Table in minimum load mode
Table 3 high-level H and medium-level M failure key index calculation table under minimum load mode
Table 4 primary state L failure key index calculation table under minimum load mode
Step five: and establishing an evaluation index capable of checking the efficiency of the algorithm, and comprehensively evaluating the algorithm by combining a comparison group. In order to verify the effectiveness of the algorithm in the minimum load mode, a control method in a random mode is adopted for a comparison group, and the effectiveness of the algorithm of the cluster monitoring network is verified by combining two key indexes (the stability maintenance degree of a connecting edge and the loss condition of the cluster monitoring area). Next, the stability maintenance degree of the continuous edge and the area loss of the cluster monitoring under the conditions of the minimum load mode and the random mode are calculated respectively, and summarized as shown in the following tables 5 to 7.
TABLE 5 PCN distribution Table in random mode
TABLE 6 Steady Retention level distribution for four failure scenarios in minimum load mode
TABLE 7 Stable maintenance level distribution for four failure scenarios in random mode
The stability degrees of all the connected edges in the cluster network are integrated, so that the stability maintaining degree and the monitoring area loss rate corresponding to the whole network under four different failure conditions can be obtained, and the result is shown in the following table 8 after statistics.
TABLE 8 network stability and monitored area loss rate under different failure conditions
As can be seen from table 8, the network stability after the nodes in different states fail varies in the minimum load mode and the random mode. When the primary state L node adjacent to the advanced state H node fails, the network stability maintaining degree is high, which indicates that the influence degree of the type failure on the network function is small, and the network stability maintaining degree after the advanced state H node fails is minimum, which indicates that the influence of the type failure on the normal operation of the network function is maximum. From the difference of the two failure transfer mechanisms, no matter which type of failure occurs, the minimum load mode can improve the network stability maintenance degree to a certain extent, reduce the adverse effect degree of the failure on the network function, and further ensure the stable operation before and after the network failure. When the primary state L node adjacent to the advanced state H node fails, the effect of the minimum load mode is the same as that of the random mode. When the high-level H node fails, the effect of the minimum load mode is significantly better than that of the random mode.
In general, from the perspective of network node state switching, a failure propagation mechanism based on a minimum load mode is proposed by taking the minimum bearing pressure of a network node and a connecting edge as a core principle, and the following conclusion is obtained through numerical example simulation: (1) the influence degrees of different types of node failures are ranked as follows: intermediate state M > advanced state H > primary state L. (2) The minimum load mode has a promoting effect on releasing the communication pressure of the external connection side of the network, and has a suppressing effect on releasing the communication pressure of the internal connection side. And (3) when the high-level state H node fails, the effect of the minimum load mode is most remarkable. The minimum load mode can reduce the extent of monitoring area loss compared to the random mode. The conclusion can provide reference and basis for optimizing the generation mechanism of the unmanned aerial vehicle cluster monitoring network, emphasizing operation and maintenance of the unmanned aerial vehicle cluster and measuring the cluster network elasticity.
In the description herein, references to the description of "one embodiment," "an example," "a specific example," etc., mean that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the disclosure. In this specification, the schematic representations of the terms used above do not necessarily refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples.
The foregoing illustrates and describes the general principles, principal features, and advantages of the present disclosure. It will be understood by those skilled in the art that the present disclosure is not limited to the embodiments described above, which are presented solely for purposes of illustrating the principles of the disclosure, and that various changes and modifications may be made to the disclosure without departing from the spirit and scope of the disclosure, which is intended to be covered by the claims.
Claims (10)
1. In the unmanned aerial vehicle networking process, each unmanned aerial vehicle is defined as a node, and the unmanned aerial vehicles are in communication connection with each other to be regarded as a connecting edge; it is characterized in that the preparation method is characterized in that,
each newly added node of the unmanned aerial vehicle is in communication connection with a node which can enable the average communication bearing pressure to be minimum in the existing unmanned aerial vehicle networking;
classifying the node states of the unmanned aerial vehicle based on the communication bearing pressure indexes, wherein when the node fails, the nodes with strong residual bearing capacity preferentially carry out redundant bearing pressure when the bearing node fails;
when the node fails, the peripheral nodes with large residual communication bearing capacity in the failure area are obtained to replace the failed nodes.
2. The control method according to claim 1, characterized in that the average communication bearer pressure comprises a communication bearer pressure PCN of the node and a communication bearer pressure PCE of the connecting edge; the communication bearing pressure PCN of the node is as follows:
in the formula (1), PCNlRefers to the average communication bearing pressure, N, of any node l in the network1Representing the number of all other nodes directly connected to the node l; n is a radical of2Representing the number of all other nodes indirectly connected to node l; n refers to the total number of nodes in the unmanned aerial vehicle cluster network (N)1+N2+1 ═ n); deg (i) denotes the value of the corresponding node i, PCNlThe communication pressure that node l can bear;
representing the number of other nodes (including the node j) passing through the shortest path from the node j to the node l, wherein the node j is not directly connected with the node l;
the definition of the communication bearing pressure of the connecting edge is shown as a formula (2);
PCE in equation (2)ijRefers to the bearing pressure, lambda, of a communication connection edge formed by any node i and node j in the network0,λ1,λ2Respectively representing the adjustment coefficients, deg (n), in three different casesi) Then it is node niThe value of (a) is determined,
measure node niAnd node njThe difference of status degree of corresponding connecting edges in all connecting edges of two parties, | PCNi-PCNjI then measure node niAnd node njThe magnitude of the difference between the communication bearer pressures.
3. The control method according to claim 1, wherein the node states of the unmanned aerial vehicle are classified into: high-order state H, medium-order state M and primary state L, QmaxRepresenting the maximum load capacity of the node,
and
respectively represent the node loads for the non-monitoring task and the monitoring task, then
The node state is at a high level when
The node state is in the intermediate state when
The node state is in the primary state.
4. The control method of claim 3, wherein performing node state switching based on a minimum load principle comprises: and unidirectional switching is performed in sequence from the directions of the primary state L, the intermediate state M and the advanced state H.
5. The control method according to claim 4, characterized in that the high-level state is initially disabled: finding out a node with the minimum PCN value around the initial failure node, and converting the node from the intermediate state M to the high state H; judging whether the residual bearing capacity of the node can completely absorb the communication pressure released after the original advanced state H node fails, if so, failing and not transmitting to the periphery, and ending the failure; otherwise, the communication pressure is continuously transmitted to the surroundings until the residual bearing capacity of a certain node can completely absorb the released communication pressure; after the failure is finished, the PCN values of all the nodes are redistributed, and the network reaches a stable operation state again;
initial failure in the intermediate state: when a primary state L node directly connected with an initially failed intermediate state M node exists, converting the state of the primary state L node corresponding to the minimum PCN value into an intermediate state M; otherwise, the state switching does not exist, the intermediate state M node containing the minimum PCN value around the initial failure intermediate state M node becomes a substitute node until the residual bearing capacity of a certain node can completely absorb the released communication pressure; after the failure is finished, the PCN values of all the nodes are redistributed, and the network reaches a stable operation state again;
initial failure of primary state: when the primary state L node initially fails, the intermediate state M or the high-level state H node adjacent to the primary state L node directly absorbs the smaller communication bearing pressure originally born by the primary state L node.
6. The control method according to claim 1, comprising cluster network generation for performing communication withstand pressure distribution of all nodes and monitoring range distribution of each node;
the cluster network is generated by acquiring a communication bearing pressure index and a minimum load principle.
7. The control method according to claim 6, wherein the cluster network generation mechanism comprises:
initializing communication bearing pressure distribution of all nodes and monitoring range distribution of each node;
the method comprises the steps of performing initial connection, wherein newly-added nodes prefer to be connected with nodes with small communication bearing pressure in existing members, and after connection, the communication bearing pressure distribution of all nodes and the monitoring range distribution of each node are recalculated;
and networking the next node, and establishing connection between the newly added node and the updated node with low communication bearing pressure in the existing members until the monitoring network is generated.
8. The control method according to claim 1, comprising obtaining a stability degree of the communication network before and after the failure and obtaining a loss of a monitoring area of the network before and after the failure, and performing discrimination of the stability of the monitoring network itself and the reliability of the task to be executed.
9. A storage medium, wherein a program is stored in the storage medium, and when executed, the program implements the control method according to any one of claims 1 to 8.
10. The unmanned aerial vehicle cluster failure control system is characterized by comprising a storage medium, wherein a program is stored in the storage medium, and when the program is executed, the control method of any one of claims 1 to 8 is realized.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210033514.1A CN114363340B (en) | 2022-01-12 | 2022-01-12 | Unmanned aerial vehicle cluster failure control method, system and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210033514.1A CN114363340B (en) | 2022-01-12 | 2022-01-12 | Unmanned aerial vehicle cluster failure control method, system and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN114363340A true CN114363340A (en) | 2022-04-15 |
| CN114363340B CN114363340B (en) | 2023-12-26 |
Family
ID=81109060
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210033514.1A Active CN114363340B (en) | 2022-01-12 | 2022-01-12 | Unmanned aerial vehicle cluster failure control method, system and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114363340B (en) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021073083A1 (en) * | 2019-10-15 | 2021-04-22 | 南京莱斯网信技术研究院有限公司 | Node load-based dynamic data partitioning system |
| CN113068151A (en) * | 2021-03-15 | 2021-07-02 | 北京航空航天大学 | Unmanned aerial vehicle swarm task reliability analysis method and system considering system failure |
| CN113359480A (en) * | 2021-07-16 | 2021-09-07 | 中国人民解放军火箭军工程大学 | Multi-unmanned aerial vehicle and user cooperative communication optimization method based on MAPPO algorithm |
| WO2021213542A1 (en) * | 2020-11-09 | 2021-10-28 | 中国民航大学 | Logistics unmanned aerial vehicle failure risk assessment method based on bayesian network |
| CN113873466A (en) * | 2021-07-12 | 2021-12-31 | 东南大学 | Unmanned aerial vehicle network elasticity measurement method and system |
-
2022
- 2022-01-12 CN CN202210033514.1A patent/CN114363340B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2021073083A1 (en) * | 2019-10-15 | 2021-04-22 | 南京莱斯网信技术研究院有限公司 | Node load-based dynamic data partitioning system |
| WO2021213542A1 (en) * | 2020-11-09 | 2021-10-28 | 中国民航大学 | Logistics unmanned aerial vehicle failure risk assessment method based on bayesian network |
| CN113068151A (en) * | 2021-03-15 | 2021-07-02 | 北京航空航天大学 | Unmanned aerial vehicle swarm task reliability analysis method and system considering system failure |
| CN113873466A (en) * | 2021-07-12 | 2021-12-31 | 东南大学 | Unmanned aerial vehicle network elasticity measurement method and system |
| CN113359480A (en) * | 2021-07-16 | 2021-09-07 | 中国人民解放军火箭军工程大学 | Multi-unmanned aerial vehicle and user cooperative communication optimization method based on MAPPO algorithm |
Non-Patent Citations (3)
| Title |
|---|
| LU GUO等: "Intelligent Penetration of UAV Based on Moth Suppression Algorithm", 2018 CHINESE AUTOMATION CONGRESS (CAC) * |
| 何勇等: "基于状态切换的杂草测绘无人机集群失效控制算法", 控制与决策 * |
| 徐鑫: "基于状态切换的杂草测绘无人机群可靠性研究", 中国优秀硕士学位论文全文数据库 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN114363340B (en) | 2023-12-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN111193629B (en) | Fault propagation method for dynamic load cascade failure of multilayer information network | |
| CN105721228A (en) | Method for importance evaluation of nodes of power telecommunication network based on fast density clustering | |
| Shuvro et al. | Modeling cascading-failures in power grids including communication and human operator impacts | |
| CN110428191B (en) | Method for identifying fragile nodes of power distribution network | |
| CN111475899B (en) | Data link network reliability index system construction method based on hyper-network theory | |
| CN110543500A (en) | Power transmission and transformation equipment health assessment platform based on big data | |
| CN110018390B (en) | Hierarchical fuzzy petri network fault diagnosis method based on comprehensive variable weight | |
| CN115391963B (en) | Random supernetwork robustness research method and system based on superside internal structure | |
| CN114326806B (en) | Unmanned aerial vehicle cluster maintenance method considering task system | |
| WO2024119555A1 (en) | Regional transport network resilience evaluation method and system | |
| CN116016206A (en) | Elasticity evaluation method and system of unmanned cluster system in open environment | |
| CN113489605B (en) | Network node importance evaluation method based on health degree | |
| CN113962273B (en) | Multi-index-based time series anomaly detection method and system and storage medium | |
| CN114363340A (en) | Unmanned aerial vehicle cluster failure control method and system and storage medium | |
| CN114142923B (en) | Optical cable fault positioning method, device, equipment and readable medium | |
| CN113132163B (en) | Optimization method and system of distributed system | |
| CN112423308A (en) | Method for actively keeping unmanned cluster network connectivity based on k-hop reachability | |
| CN109858822B (en) | Information power fusion system reliability assessment method based on flow correlation analysis | |
| KR20080087571A (en) | Context prediction system and method thereof | |
| CN112866009B (en) | Virtual network fault diagnosis method and device for comprehensive service station | |
| CN109977567B (en) | Integrated equipment network elastic modeling method based on synchronous and asynchronous analysis | |
| CN110738326B (en) | Selection method and device of artificial intelligence service system model | |
| CN113569961A (en) | Power grid node classification method and computer readable medium | |
| CN113258567A (en) | Power grid key power transmission section determination method and system based on hybrid clustering algorithm | |
| CN105933147A (en) | Topology optimization method of satellite system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |