CN112929991A

CN112929991A - Sensor management method, device, equipment and storage medium

Info

Publication number: CN112929991A
Application number: CN202110184021.3A
Authority: CN
Inventors: 江潇潇; 杨佳峰; 马天鸣; 王永琦; 金婕
Original assignee: Shanghai University of Engineering Science
Current assignee: Shanghai University of Engineering Science
Priority date: 2021-02-10
Filing date: 2021-02-10
Publication date: 2021-06-08
Anticipated expiration: 2041-02-10
Also published as: CN112929991B

Abstract

The invention provides a management method, a device, equipment and a storage medium of a sensor, wherein the method comprises the following steps: acquiring a current divergence value corresponding to each preset sensor node in the wireless sensor network; determining a first sensor set based on the current divergence value corresponding to each preset sensor node; determining a first number of preset sensor nodes as a second sensor set based on the current residual energy of each preset sensor node in the first sensor set; and the preset sensor nodes in the second sensor set are adopted to track the target in the next time period, and the divergence value and the residual energy of each preset sensor node in the wireless sensor network are combined at each moment to determine that the preset sensor nodes of the activation part complete the multi-target tracking in the next time period, so that the energy consumption balance is realized on the basis of ensuring the multi-target tracking precision, and the service life of the wireless sensor network can be effectively prolonged.

Description

Sensor management method, device, equipment and storage medium

Technical Field

The present invention relates to the field of internet and communication technologies, and in particular, to a method, an apparatus, a device, and a storage medium for managing a sensor.

Background

In a wireless sensor network, a management technology of multiple sensors becomes a current research hotspot, at present, a management method of the sensors is mainly an optimization method for multiple indexes, such as target tracking precision, sensor network energy consumption, information safety and the like, which is provided on the basis of Kalman filtering and particle filtering, and the precision and energy optimization is realized by specifically establishing different models, but the methods are suitable for single-target tracking and are not suitable for multi-target tracking.

For multi-target tracking tasks, the prior art is usually realized based on single motion sensor nodes, however, the life cycle of a wireless sensor network is usually limited by the residual energy of the single sensor nodes, and the sensor management method in the existing multi-target tracking scene makes the life of the wireless sensor network shorter.

Disclosure of Invention

The embodiment of the invention provides a management method, a management device, management equipment and a storage medium of a sensor, and aims to solve the problem that a wireless sensor network in the prior art is short in service life.

In a first aspect, an embodiment of the present invention provides a management method for a sensor, including:

acquiring a current divergence value corresponding to each preset sensor node in the wireless sensor network;

determining a first sensor set based on the current divergence value corresponding to each preset sensor node;

determining a first number of preset sensor nodes as a second sensor set based on the current residual energy of each preset sensor node in the first sensor set;

and adopting preset sensor nodes in the second sensor set to track the target in the next time period.

In a second aspect, an embodiment of the present invention provides a management apparatus for a sensor, including:

the acquisition module is used for acquiring current divergence values corresponding to preset sensor nodes in the wireless sensor network;

the determining module is used for determining a first sensor set based on the current divergence value corresponding to each preset sensor node;

the first processing module is used for determining a first number of preset sensor nodes as a second sensor set based on the current residual energy of each preset sensor node in the first sensor set;

and the second processing module is used for tracking the target in the next time period by adopting the preset sensor nodes in the second sensor set.

In a third aspect, an embodiment of the present invention provides an electronic device, including: a memory, a transceiver, and at least one processor;

the processor, the memory and the transceiver are interconnected through a circuit;

the memory stores computer-executable instructions; the transceiver is used for receiving the measurement information sent by the sensor;

the at least one processor executes computer-executable instructions stored by the memory to cause the at least one processor to perform the method as set forth in the first aspect above and in various possible designs of the first aspect.

In a fourth aspect, the present invention provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the computer-executable instructions are executed by a processor, the method according to the first aspect and various possible designs of the first aspect is implemented.

According to the management method, the management device, the management equipment and the management storage medium of the sensor, the divergence value and the residual energy of each preset sensor node are combined at each moment, the fact that the preset sensor nodes of the activated part complete multi-target tracking in the next time period is determined, energy consumption balance is achieved on the basis of guaranteeing multi-target tracking accuracy, and the service life of a wireless sensor network can be effectively prolonged.

Drawings

In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings used in the description of the embodiments or the prior art will be briefly described below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and those skilled in the art can also obtain other drawings according to the drawings without creative efforts.

Fig. 1 is a schematic flow chart of a method for managing sensors according to an embodiment of the present invention;

FIG. 2 is a schematic flow chart illustrating step 103 of a method for managing sensors according to an embodiment of the present invention;

FIG. 3 is a schematic flow chart illustrating step 101 of a method for managing sensors according to an embodiment of the present invention;

FIG. 4 is a schematic flow chart illustrating step 104 of a method for managing sensors according to an embodiment of the present invention;

fig. 5 is a schematic overall flowchart of a management method for a sensor according to an embodiment of the present invention;

fig. 6 is a schematic diagram illustrating selection of a cluster head node according to an embodiment of the present invention;

fig. 7 is a schematic diagram illustrating selection of nodes in a cluster according to an embodiment of the present invention;

fig. 8 is a schematic diagram illustrating that a distance between a cluster node and a base station is smaller than a distance between the cluster node and a cluster head node according to an embodiment of the present invention;

FIG. 9 is an exemplary diagram of a default energy consumption model according to an embodiment of the present invention;

fig. 10 is a schematic diagram of dynamic clustering in a wireless sensor network according to an embodiment of the present invention;

fig. 11 is a schematic structural diagram of a management apparatus for a sensor according to an embodiment of the present invention;

fig. 12 is a schematic structural diagram of an electronic device according to an embodiment of the present invention.

With the above figures, certain embodiments of the invention have been illustrated and described in more detail below. The drawings and the description are not intended to limit the scope of the inventive concept in any way, but rather to illustrate it by those skilled in the art with reference to specific embodiments.

Detailed Description

In order to make the objects, technical solutions and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, but not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.

The terms to which the present invention relates will be explained first:

the GLMB filter: a Generalized Labeled Multi-Bernoulli Filter is a Filter based on a RFS random finite set, and provides the random finite set with labels on the basis of a CBMeMber Filter, and the state variables of the RFS are Labeled to realize the identification of the track of each target. Rather, it takes into account more of the assumptions and components of the bernoulli component than the labeled-many LMB, which also carries a label, and is a more general form of LMB filter. The GLMB filter strictly adheres to a Bayesian filtering framework, and the filtering process is divided into a prediction part and an updating part.

Furthermore, the terms "first", "second", etc. are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. In the description of the following examples, "plurality" means two or more unless specifically limited otherwise.

The following several specific embodiments may be combined with each other, and details of the same or similar concepts or processes may not be repeated in some embodiments. Embodiments of the present invention will be described below with reference to the accompanying drawings.

An embodiment of the present invention provides a method for managing a sensor, which is used for managing a sensor in a wireless sensor network. The execution subject of this embodiment is a management apparatus of a sensor, which may be provided in an electronic device, which may be a corresponding control device in a base station BS.

As shown in fig. 1, a schematic flow chart of a management method of a sensor provided in this embodiment is provided, where the method includes:

step 101, obtaining a current divergence value corresponding to each preset sensor node in a wireless sensor network.

Specifically, a certain number (for example, S) of Sensor nodes (which may be referred to as sensors or nodes for short, and may be referred to as preset Sensor nodes for distinction) may be set in a Wireless Sensor Network (WSN) monitoring area in advance, after the setting is completed, the positions of the preset Sensor nodes are unchanged, and the positions of the preset Sensor nodes are stored in a base station, which may specifically be corresponding electronic devices of the base station, so that the base station may acquire the positions of the preset Sensor nodes during Sensor management.

When one or some preset sensor nodes detect that the target appears in the monitoring area, the preset sensor nodes detecting the target can inform the base station, the base station can activate a preset number of preset sensor nodes detecting the target in a random selection mode, so that the measurement information of the target is collected through the activated preset sensor nodes, and then the posterior multi-target density pi is obtained by operating a preset filter (such as a GLMB filter)_k{X_k|Z_1:kAnd (e), where k is 1, implementing multi-target state estimation at an initial starting time, continuously tracking multiple targets by a subsequent wireless sensor network, selecting and activating some preset sensor nodes to complete a multi-target tracking task at each time according to the residual energy of each preset sensor node without activating all sensors in order to implement better energy consumption balance and prolong the service life of the network under the condition of obtaining better tracking accuracy, and then obtaining current dispersion values corresponding to each preset sensor node in the wireless sensor network at the current time (representing the current time as k time) for determining current dispersion values corresponding to each preset sensor node in the wireless sensor network for ensuring the multi-target tracking accuracy and prolonging the service life of the network at the same time of ensuring the multi-target tracking accuracyAnd determining the part of the sensor nodes needing to be activated. The target is an object monitored by a sensor node in a wireless sensor network, the wireless sensor network can comprise various sensor nodes, can detect various phenomena in the surrounding environment such as earthquake, electromagnetism, temperature, humidity, noise, light intensity, pressure, soil components, size, speed and direction of a moving object and the like, and can be applied to the fields of military affairs, aviation, explosion prevention, disaster relief, environment, medical treatment, health care, home furnishing, industry, commerce and the like.

The current divergence value corresponding to each preset sensor node may be any practicable divergence value, such as renyi divergence, CS divergence (Cauchy-Schwarz Cauchy schwatz) divergence, KL divergence (Kullback-Leibler divergence, also called information divergence, relative entropy), and the like, and may be specifically set according to an actual demand. Taking the renyi divergence value as an example, the renyi divergence value represents the information gain between the pre-experience and post-experience multi-target probability densities, and an increase in the renyi divergence value of the sensor can indicate that the sensor acquires more information from the future measurement set. The calculation of the renyi dispersion value may be performed according to the predicted target probability density and the pseudo-updated pseudo-posterior target probability density in a POMDP (probabilistic objective Markov Decision Process, partially Observable Markov Decision Process) framework, and details are not repeated.

And 102, determining a first sensor set based on the current divergence value corresponding to each preset sensor node.

Specifically, after the current divergence value corresponding to each preset sensor node is obtained, the preset sensor nodes may be sorted according to the current divergence value, and may be sorted from high to low, and may also be sorted from low to high, and specifically may be set according to an actual requirement, the sorted preset sensor nodes form a sensor set, which may be referred to as a first sensor set, where the first sensor set includes identification information of the sorted preset sensor nodes (which may be referred to as first sensor nodes for distinguishing), such as numbers of the sorted preset sensor nodes.

Illustratively, the total number of the preset sensor nodes is 5, the numbers are 1,2,3,4 and 5 respectively, and the first sensor set after the high-to-low sorting according to the divergence value is {4,3,5,1 and 2 }.

Step 103, determining a first number of preset sensor nodes as a second sensor set based on the current residual energy of each preset sensor node in the first sensor set.

Specifically, in order to equalize the energy consumption of each preset sensor node, after the first sensor set is determined, a first number of preset sensor nodes are determined from the first sensor set and activated as a second sensor set based on the current remaining energy of each preset sensor node, so as to be used for multi-target tracking in a next time period (for example, from the current k time to the k +1 time).

Optionally, the second sensor set may be determined according to a preset rule by combining the current divergence value and the current residual energy, and the preset rule may be set according to an actual requirement, which is not limited in this embodiment.

For example, an energy threshold may be set, and in the first sensor set, a first number of preset sensor nodes having a current remaining energy higher than the energy threshold may be selected as the second sensor set according to the current divergence value from high to low. The residual energy of each preset sensor node can be maintained in real time based on a preset energy consumption model.

For example, the first sensor set may be further divided into a first optimal cluster and a first candidate set according to the current divergence value, and a part of preset sensor nodes in the first optimal cluster and a part of sensor nodes in the first candidate set are exchanged with the current remaining energy of the preset sensor nodes in the first optimal cluster and the first candidate set to form a new optimal cluster (which may be referred to as a second optimal cluster) and a new candidate set, for example, if the total S of the first sensor set is 10 preset sensor nodes, and the first number is 3, then the 3 preset sensors with the largest current divergence value form the first optimal cluster, the remaining 7 preset sensors form the first candidate set, the preset sensor nodes in the first optimal cluster are traversed, and for each preset sensor node, whether the current remaining energy meets a certain energy requirement is determined, if not, the energy requirement is determined, replacing preset sensor nodes which can meet the energy requirement from the first candidate set to replace preset sensor nodes which cannot meet the energy requirement in the first optimal cluster to form a new optimal cluster and a new candidate set, selecting a cluster head node for a finally obtained second optimal cluster, taking other preset sensor nodes in the second optimal cluster as cluster-in nodes, wherein the cluster head node can be selected by setting a cluster head node energy threshold (which can be called as a first energy threshold), taking preset sensor nodes of which the current residual energy is greater than the first energy threshold in the second optimal cluster as cluster head nodes of the second optimal cluster, and judging whether the current residual energy of the preset sensor nodes in the first optimal cluster is greater than the first energy threshold or not after determining the first optimal cluster and the first candidate set, and if not, selecting any preset sensor node A of which the current residual energy is greater than the first energy threshold from the first candidate set (or selecting any preset sensor node A of which the current residual energy is greater than the first energy threshold from the first candidate set) Selecting a preset sensor node B with the largest current divergence value from a plurality of preset sensor nodes with the current residual energy larger than a first energy threshold value) to replace a preset sensor node C with the lowest current residual energy in a first optimal cluster, taking the preset sensor node A or the preset sensor node B as a cluster head node of the first optimal cluster (the cluster head node of the first optimal cluster is also the cluster head node of a finally obtained second optimal cluster), and then determining an intra-cluster node in the finally second optimal cluster based on an intra-cluster node energy threshold value (which can be called as a second energy threshold value), namely selecting a preset sensor node E with the current residual energy larger than or equal to the second energy threshold value from a first candidate set to replace the preset sensor node D in the first optimal cluster for the preset sensor node D with the intra-cluster node smaller than the second energy threshold value in the first optimal cluster, thereby forming the second optimal cluster, and finally, taking the second optimal cluster as a second sensor set for target tracking in the next time period.

And step 104, adopting preset sensor nodes in the second sensor set to track the target in the next time period.

Specifically, the next period may be a period from the current time k to the time k +1, and in practical application, the next period may also be a period from the current time k to the time k + H, where H may be set according to an actual requirement, for example, H may be set to any integer greater than 1, which is not specifically limited, and the period from the current time k to the time k +1 is taken as an example in the embodiment of the present invention for description; after the second sensor set is determined, each preset sensor node (which may be referred to as a second sensor node for distinguishing) in the second sensor set may be activated to perform target tracking for a next time period, specifically, the base station may send an activation control instruction to each second sensor node, so that each second sensor node enters a working state, monitors a target in a monitoring area, and continuously sends measurement information to the base station, or each second sensor node forms a cluster, one of the second sensor nodes serves as a cluster head node, the other second sensor nodes serve as intra-cluster nodes and send the measurement information to the cluster head node, and the cluster head node sends the measurement information of the other second sensor nodes and the measurement information of the cluster head node to the base station; the communication mode between each second sensor node and the base station may be set according to actual requirements, and this embodiment is not limited.

After the base station receives the measurement information of each second sensor node, the base station operates an updating step of a preset filter (such as GLMB filtering) based on the measurement information of each second sensor node, so that updated posterior multi-target density is obtained, and multi-target state extraction is performed based on the updated posterior multi-target density to obtain target number and target state estimation.

The preset filter may be any implementable filter, for example, any filter based on a random finite set, specifically, a GLMB filter, a multi-bernoulli filter, a label bernoulli filter, a multi-scan label bernoulli filter, and the like may be used, and may be specifically set according to actual requirements.

According to the management method of the sensor, the divergence value and the residual energy of each preset sensor node are combined at each moment, the fact that the preset sensor nodes of the activation part complete multi-target tracking in the next time period is determined, energy consumption balance is achieved on the basis of guaranteeing multi-target tracking accuracy, and the service life of a wireless sensor network can be effectively prolonged.

In order to make the technical solution of the present invention clearer, the method provided by the above embodiment is further described in an additional embodiment of the present invention.

As an implementable manner, as shown in fig. 2, an exemplary flowchart of step 103 of the management method for a sensor provided in this embodiment is shown, where step 103 may specifically include:

step 1031, obtaining a first number of preset sensor nodes from the first sensor set as a first optimal cluster, and taking other preset sensor nodes in the first sensor set as a first candidate set.

And 1032, exchanging the preset sensor nodes in the first optimal cluster and the first candidate set based on the current residual energy and the current divergence value of each preset sensor node in the first optimal cluster and the first candidate set to obtain an exchanged second optimal cluster so as to balance the residual energy of each preset sensor node in the second optimal cluster.

Step 1033, the preset sensor nodes in the second optimal cluster are used as a second sensor set.

Specifically, after the first sensor set is determined, a first number of first sensors with higher current divergence values may be selected as a first optimal cluster according to current divergence values of preset sensor nodes (i.e., first sensor nodes) in the first sensor set, and other first sensors in the first sensor set may be selected as a first candidate set; after the first optimal cluster and the first candidate set are determined, the preset sensor nodes in the first optimal cluster and the first candidate set can be exchanged based on a preset energy balance exchange rule, the current residual energy and the current divergence value of the preset sensor nodes in the first optimal cluster and the first candidate set, the exchanged first optimal cluster is called a second optimal cluster, the current residual energy of each preset sensor node in the second optimal cluster is balanced, and the preset sensor nodes in the second optimal cluster are used as the second sensor set and are used for being activated to perform target tracking in the next period, so that the residual energy of the sensors in the wireless sensor network is balanced, and the service life of the wireless sensor network is effectively prolonged; the energy balance exchange rule may be set according to actual requirements, for example, an energy threshold may be set, and in the first sensor set, the first number of preset sensor nodes whose current residual energy is higher than the energy threshold are selected as the second sensor set from high to low according to the current divergence value. The residual energy of each preset sensor node can be maintained in real time based on a preset energy consumption model.

Illustratively, the first sensor set includes 10 preset sensor nodes, the labels are sorted from high to low to 1-10 according to the current divergence value, and 4 preset sensors are selected as the second sensor set, which may be sequentially determined from 1-10, and for the jth (j ═ 1,2, …,10) preset sensor node, it is determined whether the current remaining energy of the preset sensor node is greater than an energy threshold, and if so, the preset sensor node is added to the second sensor set, and the 4 selected preset sensor nodes whose current remaining energy is greater than the energy threshold and whose current divergence value is higher form the second sensor set, where the energy threshold may be set according to actual requirements, for example, determined according to the energy consumed by communication between each preset sensor node and the base station, and details are not repeated.

In an embodiment, optionally, exchanging the preset sensor nodes in the first optimal cluster and the first candidate set based on the current remaining energy and the current variance of each preset sensor node in the first optimal cluster and the first candidate set, to obtain an exchanged second optimal cluster, includes:

if the current residual energy of each preset sensor node in the first optimal cluster is lower than a first energy threshold, exchanging a first preset sensor node with the highest current divergence value in the preset sensor nodes with the current residual energy higher than the first energy threshold in the first candidate set with a second preset sensor node with the lowest current residual energy in the first optimal cluster to obtain a third optimal cluster and a second candidate set, and taking the first preset sensor node as a cluster head node of the third optimal cluster;

and for each third preset sensor node with the current residual energy lower than the second energy threshold in the third optimal cluster, determining a fourth preset sensor node with the highest current divergence value from the preset sensor nodes with the current residual energy higher than the second energy threshold in the second candidate set, and exchanging the third preset sensor node with the fourth preset sensor node, or directly replacing the third preset sensor node in the third optimal cluster with the fourth preset sensor node to obtain the second optimal cluster.

Specifically, the first energy threshold is a cluster head node energy threshold, when the current residual energy of each preset sensor node in the first optimal cluster is lower than the first energy threshold, one preset sensor node (for example, preset sensor node a) with the current residual energy higher than the first energy threshold and the current divergence value optimal is selected from the first candidate set to replace a preset sensor node (for example, preset sensor node B) with the current residual energy lowest in the first optimal cluster, or replace any preset sensor node in the first optimal cluster, a third optimal cluster and a second candidate set are formed, one preset sensor node (for example, preset sensor node a) higher than the first energy threshold in the third optimal cluster is used as a cluster head node of the third optimal cluster, and for each preset sensor node in the third optimal cluster, it is determined that the preset sensor node candidate set needs to be exchanged from the second sensor node based on the second energy threshold (the second energy threshold is an intra-cluster node energy threshold) Finally, a second optimal cluster is formed, the current residual energy of each preset sensor node in the second optimal cluster is higher than a second energy threshold, and the current residual energy of the cluster head node of the second optimal cluster is higher than the first energy threshold; the determination of the nodes in the second optimal cluster may specifically be exchanged by combining the current remaining energy and the current divergence value of the preset sensor nodes in the third optimal cluster and the second candidate set, specifically, for any preset sensor node x in the third optimal cluster, if the current remaining energy of the preset sensor node x is less than the second energy threshold, the preset sensor node y with the current remaining energy greater than or equal to the second energy threshold and the current divergence value being optimal is selected from the second candidate set to replace the preset sensor node x in the third optimal cluster, the preset sensor node x is removed from the third optimal cluster and added to the second candidate set, and the preset sensor y is removed from the second candidate set and added to the third optimal cluster, or the preset sensor node y is directly replaced to the preset sensor node x in the third optimal cluster, and removing the preset sensor node x from the third optimal cluster to obtain a second optimal cluster, wherein the second optimal cluster does not need to be processed, and the preset sensor node x can be specifically set according to actual requirements, and the embodiment of the invention is not limited.

It is to be understood that, in an embodiment, when the current remaining energy of at least 2 preset sensor nodes in the third optimal cluster is lower than the second energy threshold, taking 2 as an example, the third optimal cluster and the second candidate set based on the 2 nd replacement are the third optimal cluster and the second candidate set after the 1 st replacement, for example, the preset sensor nodes included in the third optimal cluster are {1,2,3,4}, the preset sensor nodes included in the second candidate set are {5,6,7,8,9,10}, the preset sensor nodes in the third optimal cluster and the second candidate set are all sorted according to the current divergence value, and the current divergence value of each preset sensor node in the third optimal cluster is greater than that of each preset sensor node in the second candidate set, the preset sensor nodes in the third optimal cluster are traversed to the preset sensor node 2, when the current residual energy of the preset sensor node 2 is found to be lower than the second energy threshold, the preset sensor nodes in the second candidate set are traversed, and when the preset sensor node 6 is traversed, the current residual energy of the preset sensor node 6 is judged to be larger than the second energy threshold (the current residual energy of the preset sensor 5 is lower than the second energy threshold), and the current divergence value of the preset sensor node 6 is optimal, the preset sensor node 2 in the third optimal cluster is exchanged with the preset sensor node 6 in the second candidate set to form an exchanged third optimal cluster {1,3,4,6} (each preset sensor node in the exchange process can be always sequenced from high to low according to the current divergence value, or can not be sequenced, and can be specifically set according to actual requirements, in this example, the preset sensor nodes are displayed in a sequence) and the exchanged second candidate set {2,5,7,8,9,10}, continuously traversing the preset sensor nodes 3 in the third optimal cluster {1,3,4,6} (no judgment is needed for the preset sensor nodes to be exchanged), judging that the current residual energy of the preset sensor nodes 3 is also lower than the second energy threshold, selecting the preset sensor nodes 9 with the current residual energy larger than the second energy threshold from the second candidate set {2,5,7,8,9,10} (the current residual energy of the preset sensor nodes 7 and 8 is lower than the second energy threshold) to replace the preset sensor nodes 3 in the third optimal cluster, and the current residual energy of the preset sensors 4 in the third optimal cluster is larger than the second energy threshold, so as to obtain the final second optimal cluster {1,4,6,9} and the third candidate set {2,3,5,7,8,10 }.

In another embodiment, optionally, at least two (for example, 2) preset sensor nodes of which the current remaining energy is lower than the second energy threshold in the third optimal cluster may be determined first, then 3 preset sensor nodes are selected from the second candidate set according to the current divergence value from high to low, and finally, the second optimal cluster and the third candidate set are formed by performing uniform swapping.

Optionally, if the first optimal cluster originally satisfies that the current residual energy of at least one preset sensor node is higher than the first energy threshold, and the current residual energy of other preset sensor nodes is higher than the second energy threshold, the first optimal cluster may be directly used as the second optimal cluster without being changed, one preset sensor node is selected from the at least one preset sensor node whose current residual energy is higher than the first energy threshold as a cluster head node of the second optimal cluster, and other preset sensor nodes except the cluster head node are used as intra-cluster nodes.

In an embodiment, optionally, after acquiring a first number of preset sensor nodes from the first sensor set as the first optimal cluster, the method further includes:

and determining one preset sensor node from the first optimal cluster as a cluster head node based on the current residual energy of each preset sensor node in the first optimal cluster.

Specifically, if the current residual energy of at least one preset sensor node in the first optimal cluster is higher than the first energy threshold, one preset sensor node in the at least one preset sensor node may be used as a cluster head node of the first optimal cluster, otherwise, according to the above manner, the preset sensor node whose current residual energy is higher than the first energy threshold and whose current divergence value is optimal is exchanged from the first candidate set as a cluster head node of the exchanged third optimal cluster, which is specifically referred to the above description and is not described herein again.

In an embodiment, after exchanging the preset sensor nodes in the first optimal cluster and the first candidate set based on the current residual energy and the current divergence value of each preset sensor node in the first optimal cluster and the first candidate set to obtain an exchanged second optimal cluster, the method further includes:

aiming at each intra-cluster node in the second optimal cluster, determining a first distance between the intra-cluster node and a cluster head node and a second distance between the intra-cluster node and a base station; if the first distance is larger than the second distance, the intra-cluster node is determined to be in direct communication with the base station, otherwise, the intra-cluster node is in communication with the cluster head node, and the cluster head node is in communication with the base station.

Specifically, after the second optimal cluster is determined, the second optimal cluster may have a situation that the second distance between one or some intra-cluster nodes and the base station is smaller or far smaller than the first distance between the intra-cluster nodes and the cluster head nodes, in order to reduce energy consumption, the base station may control the intra-cluster nodes to directly perform information exchange with the base station, other intra-cluster nodes perform information interaction with the cluster head nodes, and the cluster head nodes send the measurement information of each intra-cluster node interacted with the cluster head nodes and the measurement of the cluster head nodes to the base station; specifically, after determining the intra-cluster nodes which need to directly perform information interaction with the base station, the base station may send a control instruction to the intra-cluster nodes to notify the intra-cluster nodes of directly performing information interaction with the base station.

When the distance between the cluster node and the base station is smaller than the distance between the cluster node and the cluster head node, the cluster node is controlled to directly communicate with the base station without communicating with the cluster head node, so that the transmission energy consumption is reduced, and the service life of the network is further prolonged.

As another implementable manner, as shown in fig. 3, which is an exemplary flowchart of step 101 of the management method for a sensor provided in this embodiment, step 101 may specifically include:

and step 1011, obtaining the predicted multi-target density at the k +1 moment according to the posterior multi-target density at the k moment, wherein the k moment is the current moment.

Specifically, when one or some preset sensor nodes detect that a target appears in the monitoring area, the sensor nodes detecting the target notify the base station, the base station can activate a preset number (for example, N) of the sensor nodes detecting the target in a random selection manner, so as to collect measurement information of the target, and then the GLMB filter is operated to obtain the posterior multi-target density pi_k{X_k|Z_1:kRealizing multi-target state estimation at the initial starting moment, wherein k is 1; the subsequent wireless sensor network needs to continuously track multiple targets; and then, at the current time k (k is 1,2, …), predicting the posterior multi-target density at the time k +1 in advance to obtain the predicted multi-target density at the time k + 1.

The prediction of the multi-target density at the k +1 moment according to the posterior multi-target density at the k moment can be performed according to any practicable prediction mode in the prior art, and the embodiment is not limited. For example, the prediction is performed by using a prediction step of the GLMB filter, which is specifically as follows:

posterior multiple target density pi at k time_k{X_k|Z_1:kSubstituting the predicted target density of the moment k +1 into a prediction step of a GLMB filter to obtain the predicted target density of the moment k +1

k being used to denote the sampling instant, X_kRepresenting the set of target states at time k, X_k+1Represents a set of target predicted states at time k +1, Z_1:kRepresenting all the measurement sets from the starting time to the current k time.

And 1012, performing state extraction on the predicted multi-target density at the moment k +1 to obtain a multi-target predicted state set at the moment k + 1.

Specifically, according to the state extraction mode provided by the GLMB filter, multiple targets can be predicted at the k +1 momentDensity of

Performing state extraction to obtain a k +1 moment multi-target prediction state set

Where M denotes the number of predicted objects and i denotes the index of the object.

And 1013, determining a prediction ideal measurement set corresponding to each preset sensor node aiming at each prediction target state in the multi-target prediction state set at the moment k + 1.

Specifically, for each predicted target state X⁽ⁱ⁾Each sensor node j (j ═ 1, …, S) can calculate its corresponding predicted ideal measurement set according to the measurement equation

It differs from the calculation of ordinary measurements in that it has no clutter, no process noise, no measurement noise, and a detection probability of 1.

And 1014, pseudo updating the predicted multi-target density at the k +1 moment based on the predicted ideal measurement set corresponding to each preset sensor node to obtain the pseudo-updated pseudo-posterior multi-target density at the k +1 moment corresponding to each preset sensor node.

Specifically, according to the updating step of the GLMB filter, the predicted ideal measurement set of each sensor node j is utilized

Predicting multi-target density for k +1 time

Pseudo-updating (when pseudo-updating is carried out, the detection probability used in the GLMB updating step is 1, no process noise and clutter exist), and the pseudo-updated k +1 moment pseudo-posterior multi-target density corresponding to each sensor node j can be obtained

And step 1015, determining the current divergence value corresponding to each preset sensor node based on the predicted multi-target density at the k +1 moment and the pseudo-updated pseudo-posterior multi-target density at the k +1 moment corresponding to each preset sensor node.

Specifically, in order to obtain an analytic solution of the renyi (rayleigh) divergence, a monte carlo method can be adopted for approximate calculation, and the predicted multi-target density and the pseudo-posterior multi-target density obtained according to the GLMB filtering process both contain corresponding target state probability density information, namely, the information of the target state probability density is obtained

The predicted target probability density p (X) can be known_k+1|Z_1:k) From

Knowledge of the probability density of a pseudo-posterior target

And determining the current divergence value corresponding to each preset sensor node based on the predicted target probability density and the pseudo posterior target probability density.

As another practical way, on the basis of the foregoing embodiment, optionally, determining the first sensor set based on the current divergence value corresponding to each preset sensor node includes:

and sequencing the preset sensor nodes according to the current divergence values corresponding to the preset sensor nodes to obtain a first sensor set.

Illustratively, all preset sensor nodes are sorted in descending order according to the dispersion value of R < nyi >, and a new sensor set Sen is generated_aAs a first set of sensors.

As a further implementable manner, on the basis of the above embodiment, optionally, after performing target tracking for the next period by using preset sensor nodes in the second sensor set, the method further includes:

and updating the residual energy information of each preset sensor node.

Specifically, the energy consumption of each preset sensor node is calculated according to the preset energy consumption model, so that the residual energy of each preset sensor node is further obtained

The default residual energy of the sensor nodes which do not participate in target tracking is unchanged, so that the real-time maintenance of the residual energy of each preset sensor node is realized, and the accuracy of the residual energy used in the management process of the sensor is ensured.

Optionally, the preset energy consumption model may adopt any implementable energy consumption model, and this embodiment is not limited.

As another implementable manner, as shown in fig. 4, which is an exemplary flowchart of step 104 of the management method for a sensor provided in this embodiment, step 104 may specifically include:

step 1041, obtaining measurement information of each preset sensor node in the second sensor set.

1042, based on the measurement information of each preset sensor node in the second sensor set and the predicted multi-target density at the moment k +1, updating the posterior multi-target density at the moment k by adopting a preset filter to obtain the posterior multi-target density updated at the moment k + 1;

and 1043, determining the target number and target state estimation based on the posterior multi-target density updated at the k +1 moment.

Specifically, all sensor nodes in the second optimal cluster (i.e. the second sensor set) are activated and then transmit the measurement information to the base station, and the base station performs the updating step of the GLMB filtering according to the measurement information and the predicted multi-target density obtained previously, so as to obtain the updated posterior multi-target density pi_k+1{X_k+1|Z_1:k+1And then, extracting a multi-target state, obtaining the target number and target state estimation, realizing target tracking in the next time period from the time k to the time k +1, and then entering the next time, namely when the time k is equal to the time k +1, circularly executing the processes until the tracking is finished or the service life of the network is finished.

As an exemplary implementation manner, as shown in fig. 5, an overall flowchart of a management method of a sensor provided in this embodiment is shown, where a POMDP framework refers to a partially observable markov decision process framework, and the management method of the sensor specifically includes:

distribution of sensor nodes

Before a target enters a monitoring area of a wireless sensor network, S sensor nodes (also called preset sensor nodes, referred to as nodes for short) are placed in the monitoring area of the whole wireless sensor network in advance, the positions of all the sensor nodes are unchanged, and the positions of all the sensor nodes are known by a base station.

Second, the object appears

When the sensor nodes detect that the target appears in the monitoring area, the sensor nodes detecting the target inform the base station, the base station can activate a preset number (such as N) of sensor nodes detecting the target in a random selection mode, so that the measurement information of the target is collected, and then the posterior multi-target density pi is obtained by operating the GLMB filter_k{X_K|Z_1:kRealizing multi-target state estimation at the initial starting moment, wherein k is 1; the subsequent wireless sensor network needs to continuously track multiple targets.

Information theory-based first round sensor management

1. Pseudo prediction

k being used to denote the sampling instant, X_kRepresenting the set of target states at time k, X_k+1Represents a set of target predicted states at time k +1, Z_1:kRepresenting all the measurement sets from the starting time to the current time k, i.e., the predicted multi-target density at the time k +1 is predicted at time k.

2. Multi-target state extraction

According to the state extraction mode provided by the GLMB filter, the multi-target density can be predicted at the k +1 moment

3. Calculating and predicting Ideal Measurement Set (Predicted Ideal Measurement Set, PIMS for short)

For each predicted target state X⁽ⁱ⁾Each sensor node j (j ═ 1, …, S) can calculate its corresponding predicted ideal measurement set according to the measurement equation

The method is different from the common measurement in that the method has no clutter, no process noise and no measurement noise, the detection probability is 1, and the specific calculation mode is as follows:

wherein, [ t ]_x,k+1,t_y,k+1]Is in the predicted target state X⁽ⁱ⁾The position components in (a) represent the position of the target in the x-axis and y-axis, respectively. [ s ] of_x,j,s_y,j]And the positions of the sensor nodes j on the x axis and the y axis respectively are represented, namely the whole wireless sensor network target tracking model is in a two-dimensional coordinate space.

4. Pseudo update

According to the updating step of the GLMB filter, the predicted ideal measurement set of each sensor node j is utilized

Predicting multi-target density for k +1 time

5. Calculation of the divergence value

Predicting multi-target density based on k +1 moment

And pseudo-updated k +1 moment pseudo-posterior multi-target density corresponding to each preset sensor node

And calculating the current divergence value corresponding to each preset sensor node.

The predicted target probability density p (X) can be known_k+1|Z_1:k) From

Knowledge of the probability density of a pseudo-posterior target

Assuming that each target state probability density can be represented by a series of weighted particles

De-approximation, PN represents the number of particles, x_cDenotes the c-th particle perEach particle corresponding to a possible target state, w^cDenotes the particle x^cCorresponding weight, then the predicted target probability density p (X) at time k +1_k+1|Z_1:k) Particles with weights may be used

To approximate, it is specifically expressed in the following form:

where δ is the dirac delta function.

Similarly, the pseudo posterior target probability density obtained after pseudo-updating at the time k +1

Also using weighted particles

To approximate, then can be expressed as follows:

then, the following formula can be used to calculate the predicted target state X of the sensor node j⁽ⁱ⁾The renyi divergence.

Wherein, alpha is the preset parameter, can set up according to actual demand, for example alpha gets 0.5, and when 0.5, the tracking effect is better.

For sensor node j, it predicts the target state X for different⁽ⁱ⁾The mean value of the dispersion of the renyi as the sensingThe renyi divergence value of the device node j, namely:

m denotes the predicted target number.

6. According to the Rinyi divergence value

Sequencing all the sensor nodes in a descending order to generate a new sensor set Sen_a。

Fourth, second round sensor management based on energy consumption model

That is, the sensor nodes used for target tracking in the period from the k time to the k +1 time are determined based on the current remaining energy of each sensor node, which specifically includes the following steps:

1. acquiring a sensor set Sen obtained by the previous round of sensor management_aFrom Sen_aScheduling (selecting N with the largest R < N > yi divergence value_k+1Individual sensor nodes) form an optimal cluster (i.e., the first optimal cluster)

For the optimal cluster node, the subscript k +1 is the time. Wherein N is_k+1Is the number of sensor nodes in the optimal cluster. Then using the rest sensor nodes as a candidate node set (namely a first candidate set)

And collecting nodes as candidate nodes. Wherein L is_k+1The number of sensor nodes as a candidate node set, the candidate node set Sen_cArranged in descending order of the divergence of the ryi.

2. Setting a suitable cluster head node energy threshold ChThre (i.e. the first energy threshold),all sensor nodes in the cyclic optimal cluster

The current remaining energy of (c) is compared to ChThre and is divided into two cases:

if the current residual energy of at least one sensor node (the initial energy base station of the sensor node knows that the subsequent residual energy base station of the sensor node is calculated and maintained through an energy consumption model) is higher than ChThre, the sensor node with the current residual energy higher than ChThre and the optimal Rinyi divergence (namely the maximum divergence value) is used

As cluster head node ch (cluster head).

② if no sensor node has the current residual energy higher than ChThre, Sen_cOptimal Ranyi divergence sensor node with current residual energy higher than ChThre

Replacement of optimal cluster Sen_uSensor node with lowest medium energy

The current best cluster (i.e., the third best cluster) is obtained as follows:

the set of candidate nodes (i.e., the second candidate set) obtained is:

and will be

The nodes act as cluster head nodes.

Illustratively, as shown in fig. 6, a schematic diagram of selecting cluster head nodes is provided for the present embodiment, wherein the nodesPoint A (Black node) as the optimal Cluster Sen_u(solid circled part) with the lowest energy node, node B (gray node) is the candidate node set Sen_cThe current remaining energy is higher than the node with the best Rnyi divergence of ChThre. If the optimal cluster Sen_uIf the current residual energy of none of the inner nodes is higher than the cluster head node energy threshold value ChThre, then Sen is used_cReplacing the optimal cluster Sen by the node (node B) with the current Rinyi divergence with the current residual energy higher than ChThre_uThe middle lowest energy node (node A) to form a new optimal cluster Sen_u' (dotted circle).

3. Setting a proper intra-cluster node energy threshold value CmThre (namely a second energy threshold value), selecting intra-cluster nodes, and circularly traversing the optimal cluster Sen_u'. Two cases are distinguished:

firstly, as long as the current residual energy of a preset sensor node is lower than a cluster node threshold value CmThre, the preset sensor node is used

Removing Sen_uFrom the set of candidate nodes Sen_c' selecting the node with the optimal Rinyi divergence and the current residual energy higher than CmThre

Substitute for optimal cluster Sen_u' inner node

Forming an optimal cluster (i.e., the second optimal cluster)

Set of current candidate nodes as

Or no need to carry out the searching on the candidate node set Sen_c' treating.

Wherein, the energy threshold value CmThre of the nodes in the cluster can be based on the preset transmission coefficient R in the cluster_ncTo set up R_ncFor the maximum distance of transmission in a cluster, e.g. according to the predictionAnd (4) calculating the energy consumed in the sending process by using a formula of the energy consumption in the energy consumption model as CmThre.

Exemplarily, as shown in fig. 7, a schematic diagram for selecting nodes in a cluster provided by the present embodiment is provided, where black nodes a and B are optimal clusters Sen_u' nodes with current residual energy lower than the energy threshold of nodes in the cluster, and gray nodes C and D are candidate node set Sen_cIn the method, the nodes with the optimal Ranyi divergence and the second best nodes are selected from a candidate node set Sen_c' Gray nodes C and D replace the optimal cluster Sen_u' intermediate black nodes A and B, forming a new optimal cluster Sen_u". Setting the new optimal cluster Sen_uThe measured information of the sensor node in' is used for target tracking at the moment k + 1.

When the optimal cluster Sen_u' all nodes have current residual energy higher than CmThre, and the original optimal cluster Sen is maintained_u' invariant, i.e. the optimal cluster Sen_u' as the final second optimal cluster Sen_u″。

4. Finally formed optimal cluster Sen_u"there may be a case that a distance R1 (i.e. a first distance) between the intra-cluster node and the base station BS (base station) is far smaller than a distance R2 (i.e. a second distance) between the intra-cluster node and the cluster head node, in order to reduce a transmission distance and achieve the purpose of reducing energy consumption, the base station may inform the intra-cluster node of directly performing information interaction with the base station BS, the remaining intra-cluster nodes in the cluster continue to perform information transmission with the cluster head node Ch, and the cluster head node Ch sends data to the base station BS together, for example, as shown in fig. 8, a schematic diagram that the distance between the intra-cluster node and the base station is smaller than the distance between the intra-cluster node and the cluster head node is provided in this embodiment.

Five, GLMB filter update

Optimal cluster Sen_uAfter all the sensor nodes are activated, the measurement information is transmitted to the base station, and the base station operates the updating step of carrying out GLMB filtering according to the measurement information and the predicted multi-target density at the moment of k +1, so as to obtain the updated posterior multi-target density pi_k+1{X_k+1|Z_1:k+1Then to the updated posterior multi-target density pi_k+1{X_k+1|Z_1:k+1And extracting the target state, acquiring the target number and target state estimation, realizing target tracking from the time k to the time k +1, then entering the next time, namely the time k is the time k +1, and circularly executing the three, four and five processes until the tracking is finished or the service life of the network is finished.

It should be noted that the optimal cluster Sen is determined at the time k_u"thereafter, the optimal cluster Sen_uThe sensor nodes in the GLMB filter are activated, the sensor nodes monitor targets and collect measurement information according to a specified time, then the time of collecting the measurement information can be regarded as the time of k +1, that is, the base station operates the updating step of the GLMB filter based on the measurement information at the time of k +1 and the predicted multi-target density at the time of k +1 obtained by the prediction process, the predicted multi-target density at the time of k +1 is corrected by using the measurement information at the time of k +1, so as to obtain the posterior multi-target density at the time of k +1, and in order to represent the cycle of the above process, it is regarded that the current time enters the time of k +1 after obtaining the target number and the target state at the time of k + 1. That is, at the current k-1 moment, the posterior multi-target density at the k moment is obtained according to the process based on the posterior multi-target density at the k-1 moment, after the posterior multi-target density at the moment k is obtained and the number of targets and the target state at the moment k are obtained based on the posterior multi-target density at the moment k, the current moment enters the moment k, at the current k moment, the posterior multi-target density at the k +1 moment is obtained according to the process on the basis of the posterior multi-target density at the k moment, after obtaining the a posteriori multi-object density at the time k +1 and obtaining the number of objects and the state of the objects at the time k +1 based on the a posteriori multi-object density at the time k +1, the current time enters the time k +1, and so on, the current time changes continuously along with the time, and the embodiment of the invention only takes the current time as the time k as an example for explanation.

After the GLMB filter is updated at the moment k and the target number and the target state estimation are obtained, the energy consumption of each sensor node is calculated according to a preset energy consumption model, and therefore the residual energy of each sensor node is further obtained

And the default residual energy of the sensor nodes which do not participate in target tracking is not changed.

Illustratively, as shown in fig. 9, an exemplary diagram of the preset energy consumption model provided for the embodiment is shown, wherein E_elecIs the coefficient of the transmitting device (i.e. Transmit Electronics), Receive Electronics denotes the receiving device, Tx Amplifier denotes the power Amplifier, packet denotes the data packet, α denotes the path loss exponent, ε denotes the power amplification factor, l is the length of the transmitted data, the unit is bits, d is the distance from the sender to the receiver, when d is less than d₀When, alpha is 2, parameter epsilon is epsilon_fs(ii) a When d is greater than or equal to d₀When alpha is 4, the parameter epsilon is epsilon_amp，

Denotes the reference distance, ε_fsAnd ε_ampThe parameters are preset constants which respectively represent a power amplification coefficient of a free energy attenuation model and a power amplification coefficient of a multipath attenuation channel model, and specifically, an energy consumption calculation formula in a transmission process is as follows:

wherein, free space (d)²) And multipath fading (d)⁴) Is based on the distance between two points, i.e. the sender and the receiver. When a sensor node receives lbits-sized data, the energy consumption of the received data is:

E_Rx(l)＝E_Rx-elec(l)＝lE_elecformula seven

Based on a preset energy consumption model, the residual energy of each sensor node is maintained in real time, and the accuracy of the residual energy used in the sensor management process is ensured.

For example, as shown in fig. 10, for the dynamic clustering schematic diagram in the wireless sensor network provided in this embodiment, after clustering, only one cluster head node Ch exists in a cluster, and other nodes are all cluster internal nodes, and the cluster internal nodes transmit data to the cluster head node, and then the cluster head node transmits the data of all cluster internal nodes to a base station (the base station is a computing center) together, a black arrow indicates a data transmission direction, and a gray arrow indicates a target track.

It should be noted that the respective implementable modes in the embodiment may be implemented individually, or may be implemented in combination in any combination without conflict, and the present invention is not limited thereto.

Compared with the existing method for tracking the target only by using a single sensor node, the method for managing the sensor provided by the embodiment of the invention has the advantages that the multi-sensor node is simultaneously activated to serve as the task node to perform multi-target tracking, so that the target tracking precision is remarkably improved; moreover, the existing sensor management of multi-target tracking only considers the tracking precision and ignores the influence of energy consumption of sensor nodes on the service life of the network, the invention adds an energy consumption model on the basis of adopting the multi-sensor nodes to carry out multi-target tracking, comprehensively considers the information gain and the energy consumption brought by the sensor nodes, the wireless sensor network realizes the energy balance of the sensor nodes on the basis of ensuring the tracking precision of multi-target tracking, effectively improves the service life of the network, controls the nodes in the cluster to directly communicate with the base station without communicating with the cluster head node when the distance between the nodes in the cluster and the base station is less than the distance between the nodes in the cluster and the cluster head node, therefore, transmission energy consumption is reduced, the service life of the network is further prolonged, residual energy of each sensor node can be maintained in real time, and the accuracy of the residual energy used in the sensor management process is guaranteed.

Still another embodiment of the present invention provides a management apparatus for a sensor, which is used to perform the method of the above embodiment.

As shown in fig. 11, a schematic structural diagram of a management device of a sensor provided in this embodiment is shown. The device 30 comprises: an acquisition module 31, a determination module 32, a first processing module 33 and a second processing module 34.

The system comprises an acquisition module, a processing module and a processing module, wherein the acquisition module is used for acquiring current divergence values corresponding to preset sensor nodes in a wireless sensor network; the determining module is used for determining a first sensor set based on the current divergence value corresponding to each preset sensor node; the first processing module is used for determining a first number of preset sensor nodes as a second sensor set based on the current residual energy of each preset sensor node in the first sensor set; and the second processing module is used for tracking the target in the next time period by adopting the preset sensor nodes in the second sensor set.

Specifically, the acquisition module acquires current divergence values corresponding to preset sensor nodes in the wireless sensor network and sends the current divergence values to the determination module, the determination module determines a first sensor set and sends the first sensor set to the first processing module based on the current divergence values corresponding to the preset sensor nodes, the first processing module determines a first number of preset sensor nodes as a second sensor set based on current residual energy of the preset sensor nodes in the first sensor set and sends the second sensor set to the second processing module, and the second processing module performs target tracking in a next period by using the preset sensor nodes in the second sensor set.

The specific manner in which each module performs the operation has been described in detail in the embodiment of the method, and the same technical effect can be achieved, and will not be described in detail herein.

In order to make the device of the present invention clearer, the device provided by the above embodiment is further described in an additional embodiment of the present invention.

As a practical manner, on the basis of the foregoing embodiment, optionally, the first processing module is specifically configured to:

acquiring a first number of preset sensor nodes from a first sensor set as a first optimal cluster, and taking other preset sensor nodes in the first sensor set as a first candidate set; exchanging the preset sensor nodes in the first optimal cluster and the first candidate set based on the current residual energy and the current divergence value of each preset sensor node in the first optimal cluster and the first candidate set to obtain an exchanged second optimal cluster so as to balance the residual energy of each preset sensor node in the second optimal cluster; and taking the preset sensor nodes in the second optimal cluster as a second sensor set.

Optionally, the first processing module may specifically include an obtaining submodule and an energy balancing submodule, where the obtaining submodule is configured to obtain a first number of preset sensor nodes from the first sensor set as a first optimal cluster, use other preset sensor nodes in the first sensor set as a first candidate set, and send the first optimal cluster and the first candidate set to the energy balancing submodule; and the energy balance submodule is used for exchanging the preset sensor nodes in the first optimal cluster and the first candidate set based on the current residual energy and the current divergence value of each preset sensor node in the first optimal cluster and the first candidate set to obtain an exchanged second optimal cluster so as to balance the residual energy of each preset sensor node in the second optimal cluster, and the energy balance submodule further sends the preset sensor nodes in the second optimal cluster to the second processing module as a second sensor set.

Optionally, the first processing module is specifically configured to:

if the current residual energy of each preset sensor node in the first optimal cluster is lower than a first energy threshold, exchanging a first preset sensor node with the highest current divergence value in the preset sensor nodes with the current residual energy higher than the first energy threshold in the first candidate set with a second preset sensor node with the lowest current residual energy in the first optimal cluster to obtain a third optimal cluster and a second candidate set, and taking the first preset sensor node as a cluster head node of the third optimal cluster; and for each third preset sensor node of which the current residual energy is lower than the second energy threshold in the third optimal cluster, determining a fourth preset sensor node with the highest current divergence value from the preset sensor nodes of which the current residual energy is higher than the second energy threshold in the second candidate set, and exchanging the third preset sensor node with the fourth preset sensor node, or replacing the third preset sensor node of the third optimal cluster with the fourth preset sensor node to obtain the second optimal cluster.

Optionally, the second processing module is further configured to:

As another implementable manner, on the basis of the foregoing embodiment, optionally, the obtaining module is specifically configured to:

obtaining predicted multi-target density at the k +1 moment according to the posterior multi-target density at the k moment, wherein the k moment is the current moment; performing state extraction on the k +1 moment predicted multi-target density to obtain a k +1 moment multi-target predicted state set; aiming at each predicted target state in the multi-target predicted state set at the moment k +1, determining a predicted ideal measurement set corresponding to each preset sensor node; pseudo-updating the predicted multi-target density at the k +1 moment based on the predicted ideal measurement set corresponding to each preset sensor node to obtain pseudo-updated pseudo-posterior multi-target density at the k +1 moment corresponding to each preset sensor node; and determining the current divergence value corresponding to each preset sensor node based on the k +1 moment predicted multi-target density and the pseudo-updated k +1 moment pseudo-posterior multi-target density corresponding to each preset sensor node.

Optionally, the obtaining module may specifically include a first sub-module, a second sub-module, a third sub-module, a fourth sub-module, and a fifth sub-module, where the first sub-module obtains the predicted multi-target density at the k +1 time according to the posterior multi-target density at the k time, and sends the predicted multi-target density to the second sub-module, and the k time is the current time; the second submodule performs state extraction on the k +1 moment predicted multi-target density to obtain a k +1 moment multi-target predicted state set and sends the k +1 moment multi-target predicted state set to the third submodule; the third sub-module determines a prediction ideal measurement set corresponding to each preset sensor node and sends the prediction ideal measurement set to the fourth sub-module aiming at each prediction target state in the multi-target prediction state set at the moment k + 1; the fourth sub-module performs pseudo-updating on the k +1 moment predicted multi-target density based on the predicted ideal measurement set corresponding to each preset sensor node to obtain pseudo-updated k +1 moment pseudo-posterior multi-target density corresponding to each preset sensor node and sends the pseudo-updated k +1 moment pseudo-posterior multi-target density to the fifth sub-module; and the fifth sub-module determines the current divergence value corresponding to each preset sensor node based on the k +1 moment predicted multi-target density and the pseudo-updated k +1 moment pseudo-posterior multi-target density corresponding to each preset sensor node.

As another implementable manner, on the basis of the foregoing embodiment, optionally, the determining module is specifically configured to:

As another implementable manner, on the basis of the foregoing embodiment, optionally, the second processing module is further configured to update the remaining energy information of each preset sensor node.

As another implementable manner, on the basis of the foregoing embodiment, optionally, the second processing module is specifically configured to:

measuring information of each preset sensor node in the second sensor set is obtained; based on the measurement information of each preset sensor node in the second sensor set and the k +1 moment predicted multi-target density, updating the k moment posterior multi-target density by adopting a preset filter to obtain the k +1 moment updated posterior multi-target density; and determining the number of targets and target state estimation based on the posterior multi-target density updated at the k +1 moment.

Optionally, the second processing module may specifically include a receiving submodule, a processing submodule, and a determining submodule, where the receiving submodule acquires measurement information of each preset sensor node in the second sensor set and sends the measurement information to the processing submodule; the processing submodule predicts the multi-target density at the moment k +1 based on the measurement information of each preset sensor node in the second sensor set, updates the posterior multi-target density at the moment k by adopting a preset filter, obtains the posterior multi-target density updated at the moment k +1 and sends the posterior multi-target density to the determining submodule; and the determining submodule determines the target number and the target state estimation based on the posterior multi-target density updated at the k +1 moment.

Still another embodiment of the present invention provides an electronic device, configured to perform the method provided by the foregoing embodiment. The electronic device may be a server or other implementable computer device.

As shown in fig. 12, is a schematic structural diagram of the electronic device provided in this embodiment. The electronic device 50 includes: memory 51, transceiver 52, and at least one processor 53.

The processor, the memory and the transceiver are interconnected through a circuit; the memory stores computer-executable instructions; the transceiver is used for receiving the measurement information sent by the sensor node; the at least one processor executes computer-executable instructions stored by the memory to cause the at least one processor to perform a method as provided by any of the embodiments above.

Specifically, the receiver receives measurement information sent by a sensor node (which may be an activated sensor node, or a cluster head node selected from activated sensor nodes), and the receiver sends the received measurement information to the processor, and the processor reads and executes the computer-executable instructions from the memory, thereby implementing the method provided in any of the above embodiments.

The electronic equipment can be applied to target tracking scenes of wireless sensor networks in any fields, such as environment monitoring and forecasting, health care, intelligent home, building state monitoring, complex machinery monitoring, urban traffic, space exploration, large workshop and warehouse management, safety monitoring of airports and large industrial parks and the like, and can be specifically set according to actual requirements.

It should be noted that the electronic device of this embodiment can implement the method provided in any of the above embodiments, and can achieve the same technical effect, which is not described herein again.

Yet another embodiment of the present invention provides a computer-readable storage medium, in which computer-executable instructions are stored, and when the processor executes the computer-executable instructions, the method provided in any one of the above embodiments is implemented.

It should be noted that the computer-readable storage medium of this embodiment can implement the method provided in any of the above embodiments, and can achieve the same technical effects, which are not described herein again.

In the embodiments provided in the present invention, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative, and for example, the division of the units is only one logical division, and other divisions may be realized in practice, for example, a plurality of units or components may be combined or integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may be in an electrical, mechanical or other form.

Other embodiments of the invention will be apparent to those skilled in the art from consideration of the specification and practice of the invention disclosed herein. This invention is intended to cover any variations, uses, or adaptations of the invention following, in general, the principles of the invention and including such departures from the present disclosure as come within known or customary practice within the art to which the invention pertains. It is intended that the specification and examples be considered as exemplary only, with a true scope and spirit of the invention being indicated by the following claims.

It will be understood that the invention is not limited to the precise arrangements described above and shown in the drawings and that various modifications and changes may be made without departing from the scope thereof. The scope of the invention is limited only by the appended claims.

Claims

1. A method for managing sensors, comprising:

2. The method of claim 1, wherein determining a first number of preset sensor nodes as a second sensor set based on a current remaining energy of each preset sensor node in the first sensor set comprises:

acquiring a first number of preset sensor nodes from the first sensor set as a first optimal cluster, and taking other preset sensor nodes in the first sensor set as a first candidate set;

exchanging the preset sensor nodes in the first optimal cluster and the first candidate set based on the current residual energy and the current divergence value of each preset sensor node in the first optimal cluster and the first candidate set to obtain an exchanged second optimal cluster so as to balance the residual energy of each preset sensor node in the second optimal cluster;

and taking preset sensor nodes in the second optimal cluster as the second sensor set.

3. The method of claim 2, wherein the swapping the preset sensor nodes in the first optimal cluster and the first candidate set based on the current remaining energy of the preset sensor nodes in the first optimal cluster and the first candidate set to obtain a swapped second optimal cluster comprises:

if the current residual energy of each preset sensor node in the first optimal cluster is lower than a first energy threshold, exchanging a first preset sensor node with the highest current divergence value among preset sensor nodes with the current residual energy higher than the first energy threshold in the first candidate set with a second preset sensor node with the lowest current residual energy in the first optimal cluster to obtain a third optimal cluster and a second candidate set, and taking the first preset sensor node as a cluster head node of the third optimal cluster;

and aiming at each third preset sensor node of which the current residual energy is lower than a second energy threshold value in the third optimal cluster, determining a fourth preset sensor node with the highest current divergence value from preset sensor nodes of which the current residual energy is higher than the second energy threshold value in the second candidate set, exchanging the third preset sensor node with the fourth preset sensor node, or replacing the third preset sensor node of the third optimal cluster with the fourth preset sensor node to obtain the second optimal cluster.

4. The method of claim 2, wherein after the preset sensor nodes in the first optimal cluster and the first candidate set are transposed based on the current remaining energy and the current variance of the preset sensor nodes in the first optimal cluster and the first candidate set, and a transposed second optimal cluster is obtained, the method further comprises:

determining a first distance between each cluster node in the second optimal cluster and a cluster head node and a second distance between each cluster node in the second optimal cluster and a base station;

if the first distance is larger than the second distance, the in-cluster node is determined to be in direct communication with the base station, otherwise, the in-cluster node is in communication with the cluster head node, and the cluster head node is in communication with the base station.

5. The method according to claim 1, wherein the obtaining the current divergence value corresponding to each preset sensor node in the wireless sensor network comprises:

obtaining predicted multi-target density at the k +1 moment according to the posterior multi-target density at the k moment, wherein the k moment is the current moment;

performing state extraction on the k +1 moment predicted multi-target density to obtain a k +1 moment multi-target predicted state set;

aiming at each predicted target state in the multi-target predicted state set at the moment k +1, determining a predicted ideal measurement set corresponding to each preset sensor node;

pseudo-updating the predicted multi-target density at the k +1 moment based on the predicted ideal measurement set corresponding to each preset sensor node to obtain pseudo-updated pseudo-posterior multi-target density at the k +1 moment corresponding to each preset sensor node;

and determining the current divergence value corresponding to each preset sensor node based on the k +1 moment predicted multi-target density and the pseudo-updated k +1 moment pseudo-posterior multi-target density corresponding to each preset sensor node.

6. The method of claim 1, wherein determining the first set of sensors based on the current divergence value corresponding to each preset sensor node comprises:

and sequencing the preset sensor nodes according to the current divergence values corresponding to the preset sensor nodes to obtain the first sensor set.

7. The method of claim 1, wherein after performing target tracking for a next period of time using preset sensor nodes in the second set of sensors, the method further comprises:

and updating the residual energy information of each preset sensor node.

8. The method according to any one of claims 1-7, wherein said using a predetermined sensor node in said second sensor set for target tracking for a next time period comprises:

measuring information of each preset sensor node in the second sensor set is obtained;

predicting the multi-target density at the k +1 moment based on the measurement information of each preset sensor node in the second sensor set, and updating the posterior multi-target density at the k moment by adopting a preset filter to obtain the posterior multi-target density updated at the k +1 moment;

and determining the number of targets and target state estimation based on the posterior multi-target density updated at the k +1 moment.

9. A management device for a sensor, comprising:

10. An electronic device, comprising: a memory, a transceiver, and at least one processor;

the memory stores computer-executable instructions; the transceiver is used for receiving the measurement information sent by the sensor node;

the at least one processor executing the computer-executable instructions stored by the memory causes the at least one processor to perform the method of any one of claims 1-8.

11. A computer-readable storage medium having computer-executable instructions stored thereon which, when executed by a processor, implement the method of any one of claims 1-8.