WO2021243767A1 - 应用于海洋信息网络的分层式数据采集系统及方法 - Google Patents

应用于海洋信息网络的分层式数据采集系统及方法 Download PDF

Info

Publication number
WO2021243767A1
WO2021243767A1 PCT/CN2020/097722 CN2020097722W WO2021243767A1 WO 2021243767 A1 WO2021243767 A1 WO 2021243767A1 CN 2020097722 W CN2020097722 W CN 2020097722W WO 2021243767 A1 WO2021243767 A1 WO 2021243767A1
Authority
WO
WIPO (PCT)
Prior art keywords
cluster
node
sse
head node
sensor nodes
Prior art date
Application number
PCT/CN2020/097722
Other languages
English (en)
French (fr)
Inventor
秦川
Original Assignee
秦川
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 秦川 filed Critical 秦川
Priority to US17/436,591 priority Critical patent/US11305848B2/en
Publication of WO2021243767A1 publication Critical patent/WO2021243767A1/zh

Links

Images

Classifications

    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B63—SHIPS OR OTHER WATERBORNE VESSELS; RELATED EQUIPMENT
    • B63B—SHIPS OR OTHER WATERBORNE VESSELS; EQUIPMENT FOR SHIPPING 
    • B63B35/00—Vessels or similar floating structures specially adapted for specific purposes and not otherwise provided for
    • 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
    • H04L67/12—Protocols specially adapted for proprietary or special-purpose networking environments, e.g. medical networks, sensor networks, networks in vehicles or remote metering networks
    • G—PHYSICS
    • G08—SIGNALLING
    • G08G—TRAFFIC CONTROL SYSTEMS
    • G08G3/00—Traffic control systems for marine craft
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
    • G16Y30/00—IoT infrastructure
    • G—PHYSICS
    • G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
    • G16Y—INFORMATION AND COMMUNICATION TECHNOLOGY SPECIALLY ADAPTED FOR THE INTERNET OF THINGS [IoT]
    • G16Y40/00—IoT characterised by the purpose of the information processing
    • G16Y40/10—Detection; Monitoring
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
    • H04L45/00—Routing or path finding of packets in data switching networks
    • H04L45/02—Topology update or discovery
    • H04L45/08—Learning-based routing, e.g. using neural networks or artificial intelligence
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W40/00—Communication routing or communication path finding
    • H04W40/02—Communication route or path selection, e.g. power-based or shortest path routing
    • H04W40/04—Communication route or path selection, e.g. power-based or shortest path routing based on wireless node resources
    • H04W40/10—Communication route or path selection, e.g. power-based or shortest path routing based on wireless node resources based on available power or energy
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W84/00—Network topologies
    • H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B63—SHIPS OR OTHER WATERBORNE VESSELS; RELATED EQUIPMENT
    • B63B—SHIPS OR OTHER WATERBORNE VESSELS; EQUIPMENT FOR SHIPPING 
    • B63B35/00—Vessels or similar floating structures specially adapted for specific purposes and not otherwise provided for
    • B63B2035/006—Unmanned surface vessels, e.g. remotely controlled
    • B63B2035/007—Unmanned surface vessels, e.g. remotely controlled autonomously operating
    • B—PERFORMING OPERATIONS; TRANSPORTING
    • B63—SHIPS OR OTHER WATERBORNE VESSELS; RELATED EQUIPMENT
    • B63B—SHIPS OR OTHER WATERBORNE VESSELS; EQUIPMENT FOR SHIPPING 
    • B63B2211/00—Applications
    • B63B2211/02—Oceanography
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W4/00—Services specially adapted for wireless communication networks; Facilities therefor
    • H04W4/30—Services specially adapted for particular environments, situations or purposes
    • H04W4/38—Services specially adapted for particular environments, situations or purposes for collecting sensor information
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W40/00—Communication routing or communication path finding
    • H04W40/02—Communication route or path selection, e.g. power-based or shortest path routing
    • H04W40/20—Communication route or path selection, e.g. power-based or shortest path routing based on geographic position or location
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W40/00—Communication routing or communication path finding
    • H04W40/24—Connectivity information management, e.g. connectivity discovery or connectivity update
    • H04W40/32—Connectivity information management, e.g. connectivity discovery or connectivity update for defining a routing cluster membership
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04W—WIRELESS COMMUNICATION NETWORKS
    • H04W84/00—Network topologies
    • H04W84/18—Self-organising networks, e.g. ad-hoc networks or sensor networks
    • H04W84/20—Master-slave selection or change arrangements
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
    • Y02D30/00—Reducing energy consumption in communication networks
    • Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks

Definitions

  • the present invention relates to the technical field of marine data information collection, in particular, to a layered data acquisition system applied to a marine information network, and also to a layered data acquisition method applied to a marine information network.
  • the current conventional data acquisition system is to arrange multiple sensor nodes at different depths of the ocean to collect and transmit underwater information.
  • the ocean data is finally aggregated to the fixed base station (vessel) on the water surface.
  • the multi-hop communication method not only increases the data packet loss rate, but also has a large system delay;
  • the data transmission link is relatively fixed. If the energy of each node is exhausted, the entire connection will face the possibility of failure, resulting in poor system robustness.
  • the present invention provides a hierarchical data collection system and method applied to a marine information network, so as to solve the technical problems of large energy consumption and low efficiency in the current marine information network-oriented data collection system.
  • a hierarchical data acquisition system applied to a marine information network, which includes a surface base station, a data acquisition layer for collecting seabed data, and a The seabed data information collected by the data acquisition layer is transmitted to the data transmission layer of the surface base station.
  • the data acquisition layer includes several sensor nodes anchored on the seabed, and the several sensor nodes are arranged in clusters, and each cluster includes a cluster head Node and multiple ordinary nodes, multiple ordinary nodes collect seabed data information and then transmit it to the cluster head node for data information aggregation;
  • the data transmission layer includes underwater robots, which traverse every base station after starting from the surface base station.
  • the cluster head node of a cluster collects the aggregated data information and then drives back to the surface base station, and transmits the collected seabed data information to the surface base station; after each data collection cycle is completed, a new sensor node is reselected in each cluster As the cluster head node of the next data collection cycle.
  • T(s) represents the probability that the sensor node s is selected as the new cluster head node
  • G represents the set of all ordinary nodes of the cluster in this round
  • w 1 and w 2 represent the constant coefficient of the weight
  • d s represents The reciprocal of the distance between the sensor node s and its corresponding cluster center
  • E j is the ratio of the remaining energy of the sensor node s to the initial energy
  • ⁇ is any constant.
  • the underwater robot adopts an ant algorithm that considers distance and angle at the same time and has a rewarding property to plan its running path, and the algorithm includes the following content:
  • the first ant starts from the location of the base station and calculates the transition probability from location i to location j at that moment Then select the next location point, update the taboo table at the same time, calculate the transition probability, select a new location point, and update the taboo table synchronously until the ant dies after traversing N ant locations, where the transition probability of the ant is :
  • s represents the position of the next moment
  • n(i) represents the set of remaining position points adjacent to the position point i
  • ⁇ and ⁇ respectively represent the weight constant
  • ⁇ ij (t) represents the pheromone concentration
  • ⁇ ij (t) represents the heuristic factor
  • d ij represents the distance from the position point i to j
  • r ij represents the offset angle of the position point i to j
  • Q 1 and Q 2 are arbitrary constants
  • ⁇ is the reward factor, which indicates the degree of influence of the pheromone concentration of the subsequent path on the current path selection, and its value range is 0 to 1
  • P js represents the transition probability of the ant at the next moment
  • ⁇ ij (t+1) (1- ⁇ ) ⁇ ij (t)+ ⁇ ij ;
  • ⁇ is the pheromone evaporation coefficient, which represents the amount of pheromone reduction
  • (1- ⁇ ) is the pheromone residual coefficient, which represents the remaining pheromone concentration
  • the offset angle r ij of the position point i to j is calculated by the following formula:
  • xi and yi respectively represent the abscissa and ordinate of the position point i
  • xj and yj respectively represent the abscissa and ordinate of the position point j
  • ⁇ represents the vector The angle with the horizontal direction.
  • sensor nodes are arranged in clusters in the following manner:
  • C i represents the set of all sensor nodes in a cluster
  • x represents the coordinate value of the sensor nodes
  • k represents the number of clusters.
  • ⁇ S represents the difference of SSE generated by two adjacent clusters
  • the sensor node closest to the cluster center is selected as the initial cluster head node.
  • the present invention also provides a layered data collection method applied to marine information network, including the following steps:
  • Step S1 Arrange and anchor several sensor nodes in subsea clusters.
  • Each cluster includes a cluster head node and multiple common nodes. After multiple common nodes collect subsea data information, they will be transmitted to the cluster head node for data information aggregation. ;
  • Step S2 Release the underwater robot from the position of the surface base station, and control the underwater robot to traverse the cluster head node of each cluster to collect the aggregated data information and then drive back to the surface base station;
  • Step S3 Control the underwater robot to transmit the collected seabed data information to the surface base station;
  • Step S4 After each data collection cycle is executed, a new sensor node is reselected in each cluster as the cluster head node of the next data collection cycle.
  • step S4 the following formula is used to calculate the probability of each sensor node in the cluster as a new cluster head node, and the new cluster head node is selected according to the calculation result:
  • T(s) represents the probability that the sensor node s is selected as the new cluster head node
  • G represents the set of all ordinary nodes of the cluster in this round
  • w 1 and w 2 represent the constant coefficient of the weight
  • d s represents the sensor node
  • E j is the ratio of the remaining energy of the sensor node s to the initial energy
  • ⁇ is any constant.
  • step S2 an ant algorithm that considers distance and angle at the same time and has a rewarding property is used to plan the running path, and the algorithm includes the following steps:
  • Step S21 Suppose there are a total of m ants and N ant location points. All ants are placed at the base station location on the surface of the water. At the same time, a taboo table is set to indicate the set of location points passed by the ants. Each time the ant passes by a location point Then add this point to the taboo table;
  • Step S22 The first ant starts from the position of the base station and calculates the transition probability from position i to position j at that moment Then select the next location point, update the taboo table at the same time, calculate the transition probability, select a new location point, and update the taboo table synchronously until the ant dies after traversing N ant locations, where the transition probability of the ant is :
  • s represents the position of the next moment
  • n(i) represents the set of remaining position points adjacent to the position point i
  • ⁇ and ⁇ respectively represent the weight constant
  • ⁇ ij (t) represents the pheromone concentration
  • ⁇ ij (t) represents the heuristic factor
  • d ij represents the distance from the position point i to j
  • r ij represents the offset angle of the position point i to j
  • Step S23 Calculate the pheromone concentration of the ant on each path
  • Q 1 and Q 2 are arbitrary constants
  • ⁇ is the reward factor, which indicates the degree of influence of the pheromone concentration of the subsequent path on the current path selection, and its value range is 0 to 1
  • P js represents the transition probability of the ant at the next moment
  • Step S24 Repeat the above steps S22 and S23 until m ants traverse all the position points, and calculate the pheromone increment ⁇ ij and the total pheromone ⁇ ij (t+1) on each path,
  • ⁇ ij (t+1) (1- ⁇ ) ⁇ ij (t)+ ⁇ ij ;
  • ⁇ is the pheromone evaporation coefficient, which represents the amount of pheromone reduction
  • (1- ⁇ ) is the pheromone residual coefficient, which represents the remaining pheromone concentration
  • Step S25 Record the sum of distances and offset angles generated in this iteration, and update the current optimal path according to the recording results;
  • Step S26 Iteration is continuously performed. When the preset number of iterations is reached or a stagnation phenomenon occurs, the iteration is terminated, and the optimal path updated after the last iteration is taken as the running path, otherwise the iteration is continued.
  • step S22 the offset angle r ij in step S22 is calculated by the following formula:
  • xi and yi respectively represent the abscissa and ordinate of the position point i
  • x j and y j respectively represent the abscissa and ordinate of the position point j
  • ⁇ represents the vector The angle with the horizontal direction.
  • step S1 includes the following steps:
  • Step S11 When a number of sensor nodes are divided into 1 to n clusters, the corresponding squared distance error SSE from the sensor node in each cluster to the cluster center is calculated,
  • C i represents the set of all sensor nodes in a cluster
  • x represents the coordinate value of the sensor nodes
  • k represents the number of clusters.
  • Step S12 Based on the k value obtained above, randomly select k sensor nodes as the initial cluster centers, and then calculate the SSE values from the remaining sensor nodes to the k cluster centers, and divide the remaining sensor nodes separately based on the principle of closest distance Allocate to each initial cluster center, and continuously update the cluster center and the corresponding SSE value until the convergence condition is met.
  • the update equation of the cluster center is: The convergence condition is:
  • Step S13 After the iteration, the sensor node closest to the cluster center is selected as the initial cluster head node.
  • the hierarchical data acquisition system applied to the marine information network of the present invention arranges a number of sensor nodes in clusters, each cluster includes a cluster head node and a plurality of ordinary nodes, after collecting seabed data information through a plurality of ordinary sensor nodes Unified transmission to the cluster head node, the cluster head node is responsible for summarizing the data from each common node in the same cluster and transmitting the summarized data to the underwater robot.
  • the common node mainly performs the data collection function
  • the cluster head node mainly Performing the data transmission function greatly reduces the energy consumption of each sensor node, prolongs the service life of the sensors in the data acquisition layer, and greatly improves the data collection efficiency of the data acquisition layer.
  • a new sensor node in each cluster will be re-selected as the cluster head node for the next data collection cycle.
  • This system takes into account that the energy consumption of the cluster head node is larger than that of ordinary nodes. If a cluster head node is fixedly used, the energy loss of the sensor as the cluster head node will be very fast, and it will soon die, resulting in a very low data collection efficiency for the entire cluster. Replacing a new cluster head node can well balance the energy loss of all sensor nodes in the same cluster. The old cluster head node with a low energy value can be converted to a normal node for use, and it can still be used under low energy values.
  • the layered data collection method applied to the marine information network of the present invention also has the above-mentioned advantages.
  • Fig. 1 is a schematic diagram of the layout of a hierarchical data collection system applied to a marine information network according to a preferred embodiment of the present invention.
  • FIG. 2 is a schematic diagram of the comparison of the results of the sensor clustering arrangement algorithm used in the hierarchical data acquisition system applied to the marine information network in the preferred embodiment of the present invention with the classic K-means algorithm and LEACH algorithm in terms of energy consumption.
  • FIG. 3 is a schematic diagram of the comparison of the results of the sensor clustering arrangement algorithm used in the hierarchical data acquisition system applied to the marine information network of the preferred embodiment of the present invention with the classic K-means algorithm and LEACH algorithm in terms of node survival rate.
  • Fig. 4 is a schematic diagram of calculating the offset angle in the hierarchical data collection system applied to the marine information network according to the preferred embodiment of the present invention.
  • Fig. 5 is a schematic diagram showing the comparison of the shortest path results respectively obtained by the path planning algorithm used in the hierarchical data collection system applied to the marine information network and the classic ant algorithm according to the preferred embodiment of the present invention.
  • Fig. 6 is a schematic diagram showing the comparison of the offset angle results respectively obtained by the path planning algorithm and the classic ant algorithm used in the hierarchical data collection system applied to the marine information network according to the preferred embodiment of the present invention.
  • Fig. 7 is a schematic flowchart of a layered data collection method applied to a marine information network according to another embodiment of the present invention.
  • Fig. 8 is a schematic diagram of a sub-process of clustering and arranging several sensor nodes in step S1 in Fig. 7.
  • FIG. 9 is a schematic diagram of a sub-flow of the path planning algorithm adopted in step S2 in FIG. 7.
  • a preferred embodiment of the present invention provides a hierarchical data acquisition system applied to a marine information network, which includes a surface base station, a data acquisition layer for collecting seabed data information, and a data acquisition layer located on the data acquisition layer.
  • the upper part is used to transmit the seabed data information collected by the data acquisition layer to the data transmission layer of the surface base station, where the data acquisition layer includes several sensor nodes anchored on the seabed, and several sensor nodes are arranged in clusters, Each cluster includes a cluster head node and multiple common nodes, and multiple common nodes collect seabed data information and then transmit it to the cluster head node for data information aggregation;
  • the data transmission layer includes an underwater robot, the underwater robot After starting from the surface base station, it traverses the cluster head node of each cluster to collect the aggregated data information and then drives back to the surface base station, and transmits the collected seabed data information to the surface base station; after each data collection cycle is completed, in each cluster Select the new sensor node as the cluster head node in the next data
  • the underwater robot when the underwater robot moves to a certain height above the cluster head node, for example, at a height of 20 m to 50 m, preferably at a height of 30 m, the underwater robot will quickly establish a communication link with the cluster head node and perform data transmission. , After completing the data transmission of the current cluster head node, drive to the next cluster head node, and perform data transmission with each cluster head node in turn, and finally bring the data of all cluster head nodes back to the surface base station.
  • the cluster head node communicates with other ordinary nodes, and each sensor node has an autonomous control function, which can switch autonomously between the data collection function and the data transmission function.
  • the specific way to achieve autonomous switching belongs to the existing technology. I won't repeat it here.
  • the underwater robot in the data transmission layer can directly sink directly from the plane where the surface base station is located to a certain height above the data acquisition layer. Then collect data in the horizontal plane according to the running path; in addition, the underwater robot can sink directly from the base station on the surface to a certain height above the first cluster head node, and then collect data along the plane according to the running path.
  • the specific route along which the underwater robot moves from the horizontal plane where the surface base station is located to the horizontal plane at a certain height above the data collection layer is not specifically limited here.
  • the hierarchical data collection system applied to the marine information network of this embodiment arranges several sensor nodes in clusters, and each cluster includes a cluster head node and multiple common nodes, which are collected by multiple common sensor nodes.
  • the cluster head node After the seabed data information is uniformly transmitted to the cluster head node, the cluster head node is responsible for summarizing the data from each common node in the same cluster and transmitting the summarized data to the underwater robot.
  • the common node mainly performs the data collection function
  • the cluster head node mainly performs the data transmission function, which greatly reduces the energy consumption of each sensor node, prolongs the service life of the sensors in the data acquisition layer, and greatly improves the data collection efficiency of the data acquisition layer.
  • a new sensor node in each cluster will be re-selected as the cluster head node for the next data collection cycle.
  • This system takes into account that the energy consumption of the cluster head node is larger than that of ordinary nodes. If a cluster head node is fixedly used, the energy loss of the sensor as the cluster head node will be very fast, and it will soon die, resulting in a very low data collection efficiency for the entire cluster. Replacing a new cluster head node can well balance the energy loss of all sensor nodes in the same cluster. The old cluster head node with a low energy value can be converted to a normal node for use, and it can still be used under low energy values.
  • the cluster head node will also process and fuse the aggregated information to eliminate duplicate data information, and then the cluster head node will transmit the processed and fused data information to the underwater robot , That is, the cluster head node has the ability of data preprocessing.
  • the cluster head node has the ability of data preprocessing.
  • several sensor nodes are arranged in clusters in the following manner:
  • C i represents the set of all sensor nodes in a cluster
  • x represents the coordinate value of the sensor nodes
  • k represents the number of clusters
  • the value range of k is 1 to n.
  • ⁇ S represents the difference between the SSE generated by two adjacent clusters, and thus the number of clusters k is determined. Then, based on the k value obtained above, randomly select k sensor nodes as the initial cluster centers, and then calculate the SSE values from the remaining sensor nodes to the k cluster centers, and allocate the remaining sensor nodes separately based on the principle of closest distance For each initial cluster center, the iterative cluster center and the corresponding SSE value are continuously updated during the allocation process until the convergence condition is met.
  • the update equation of the cluster center is:
  • the convergence condition is:
  • C i represents the number of sensor nodes in the cluster
  • ⁇ is the minimum threshold
  • SSE 1 and SSE 2 respectively represent the SSE value generated by the current iteration and the SSE value generated by the previous iteration
  • the final The coordinates of the cluster center and the sensor nodes contained in each cluster.
  • the sensor node closest to the final cluster center is selected as the initial cluster head node, and the remaining sensor nodes are regarded as ordinary nodes.
  • Ordinary nodes begin to collect seabed data information, and then transmit it to the cluster head node in the same cluster.
  • the cluster head node gathers the data collected by all ordinary nodes in the cluster and transmits it to the underwater robot. This process is a data collection cycle.
  • the specific clustering arrangement plan is transmitted to the underwater robot, and the underwater robot dives to traverse all the sensor nodes and arranges them in clusters
  • the information is correspondingly transmitted to each sensor node, and each sensor node completes the clustering arrangement after receiving the clustering arrangement information.
  • the following formula is used to calculate the probability of each remaining sensor node in the cluster as the new cluster head node, and the new cluster head node is selected according to the calculation result.
  • the new cluster head node performs the data aggregation and data transmission functions.
  • the cluster head node of is converted to a normal node to perform the data collection function, and the calculation formula is as follows:
  • T(s) represents the probability that the sensor node s is selected as the new cluster head node
  • G represents the set of all ordinary nodes of the cluster in this round
  • w 1 and w 2 represent the constant coefficient of the weight
  • d s represents The reciprocal of the distance between the sensor node s and its corresponding cluster center
  • E j is the ratio of the remaining energy of the sensor node s to the initial energy
  • ⁇ is any constant.
  • each ordinary node after performing a data collection cycle, each ordinary node performs the above probability calculation, and then each ordinary node transmits its own probability calculation result to the old cluster head node, and the old cluster head node The node with the highest probability value is compared as the new cluster head node, and a feedback signal is generated and transmitted to the ordinary node corresponding to the highest probability value.
  • the cluster head node is used, and the old cluster head node is automatically switched to perform the data collection function, that is, used as a normal node.
  • the old cluster head node after executing a data collection cycle, automatically switches to the data collection function, that is, it is used as an ordinary node, and an ordinary node is selected for the above probability calculation based on the random principle or the principle of proximity.
  • the surface base station can select the new cluster head node in each cluster through data simulation calculation, and then add the new cluster head node The cluster head node information is transmitted to the underwater robot.
  • the underwater robot collects the data information summarized by the old cluster head node, and at the same time transmits the new cluster head node information to the old cluster head node, and the old cluster head node informs the new cluster For the head node, the new cluster head node performs the data transmission function, and the old cluster head node is switched to the normal node for use.
  • the surface base station may select a new cluster head node in each cluster through data simulation calculation, and then transmit the new cluster head node information to the underwater robot.
  • the underwater robot dives according to the updated operating path, and transmits the new cluster head node information to the sensor node in turn.
  • the sensor node that receives the new cluster head node information is switched to the new cluster head node for use, and The old cluster head node is automatically switched to normal node for use.
  • the above algorithm comprehensively considers the distance between the sensor node and the cluster center and the remaining energy of the sensor node itself, which not only ensures that the remaining energy of the sensor node itself is sufficient to support the data aggregation and data transmission functions, but also that the remaining ordinary nodes reach the new cluster head
  • the distance between the nodes is relatively short, which reduces the energy loss caused by the transmission distance between other ordinary nodes and the new cluster head node, further realizes energy saving and consumption reduction, and improves the service life and overall of the sensor nodes in the entire cluster. The efficiency of data collection.
  • This application names the aforementioned sensor clustering algorithm ECBIK algorithm, as shown in Figure 2 and Figure 3.
  • This application compares the classic K-means algorithm and LEACH algorithm with respect to node survival rate and energy consumption. It can be seen from the comparison result that the ECBIK algorithm of this application has the longest time for the first dead node and the time for all nodes to die, maintaining a high node survival rate throughout the process, and maintaining a high residual energy throughout the process. The whole process consumes the least energy and has the best energy-saving effect.
  • the underwater robot uses an ant algorithm that considers distance and angle at the same time and has a rewarding nature for path planning.
  • the algorithm includes the following:
  • the first ant starts from the location of the base station and calculates the transition probability from location i to location j at that moment Then select the next location point, update the taboo table at the same time, calculate the transition probability, select a new location point, and update the taboo table synchronously until the ant dies after traversing N ant locations, where the transition probability of the ant is :
  • s represents the position of the next moment
  • n(i) represents the set of remaining position points adjacent to the position point i
  • ⁇ and ⁇ respectively represent the weight constant
  • ⁇ ij (t) represents the pheromone concentration
  • ⁇ ij (t) represents the heuristic factor
  • di j represents the distance from the position point i to j
  • r ij represents the offset angle of the position point i to j
  • the underwater robot is at position i at the moment and heading to position j, then its state matrix can be expressed as:
  • ⁇ represents the position matrix
  • ⁇ represents the velocity matrix
  • u, v, ⁇ respectively represent the surge speed, yaw speed and yaw speed of the underwater robot
  • ⁇ in Figure 4 represents the vector
  • the angle with the horizontal direction defines The difference with ⁇ is the offset angle of the underwater robot from the position point i to the position point j, namely r ij .
  • the offset angle r ij of the position point i to j can be calculated by the following formula:
  • x i and y i respectively represent the abscissa and ordinate of the position point i
  • x j and y j respectively represent the abscissa and ordinate of the position point j.
  • J( ⁇ ) is the motion transition matrix, which can be expressed as:
  • the coordinates of the next position point can be calculated according to formulas (10) and (11), and then the offset angle r between the two position points can be calculated ij .
  • Q 1 and Q 2 are arbitrary constants
  • ⁇ is the reward factor, which indicates the degree of influence of the pheromone concentration of the subsequent path on the current path selection, and its value range is 0 to 1
  • P js represents the transition probability of the ant at the next moment .
  • ⁇ ij (t+1) (1- ⁇ ) ⁇ ij (t)+ ⁇ ij ;
  • ⁇ represents the pheromone evaporation coefficient, which represents the amount of pheromone reduction
  • (1- ⁇ ) is the pheromone residual coefficient, which represents the remaining pheromone concentration
  • this application names the above-mentioned improved path planning algorithm R-ACO algorithm, which comprehensively considers the influence of distance and angle offset, and the running distance and offset angle of the solved path are both small and large.
  • the energy consumption of the underwater robot is reduced, and the influence of the future state pheromone concentration on the current path selection is also considered in formula (12). It has the effect of hyperopia and the characteristics of global optimization, and the solution efficiency is higher and the convergence is faster.
  • this application compares the calculated shortest path and minimum deviation angle with the classic ant algorithm ACO. It can be clearly seen from the comparison result that the R- The running distance and deviation angle solved by the ACO algorithm are smaller than the classic ACO algorithm, and the convergence time is shorter.
  • another embodiment of the present invention also provides a hierarchical data collection method applied to a marine information network, which preferably adopts the hierarchical data collection system of the above-mentioned embodiment.
  • the method of data collection includes the following steps:
  • Step S1 Arrange and anchor several sensor nodes in subsea clusters.
  • Each cluster includes a cluster head node and multiple common nodes. After multiple common nodes collect subsea data information, they will be transmitted to the cluster head node for data information aggregation. ;
  • Step S2 Release the underwater robot from the position of the surface base station, and control the underwater robot to traverse the cluster head node of each cluster to collect the aggregated data information and then drive back to the surface base station;
  • Step S3 Control the underwater robot to transmit the collected seabed data information to the surface base station;
  • Step S4 After each data collection cycle is executed, a new sensor node is reselected in each cluster as the cluster head node of the next data collection cycle.
  • the hierarchical data collection method applied to the marine information network of this embodiment is arranged by clustering a number of sensor nodes, each cluster includes a cluster head node and multiple common nodes, through multiple common sensor nodes After collecting the seabed data information, it is uniformly transmitted to the cluster head node.
  • the cluster head node is responsible for summarizing the data from each common node in the same cluster and transmitting the summarized data to the underwater robot.
  • the common node mainly performs the data collection function.
  • the cluster head node mainly performs the data transmission function, which greatly reduces the energy consumption of each sensor node, prolongs the service life of the sensors in the data acquisition layer, and greatly improves the data collection efficiency of the data acquisition layer.
  • a new sensor node in each cluster will be re-selected as the cluster head node for the next data collection cycle.
  • This system takes into account that the energy consumption of the cluster head node is larger than that of ordinary nodes. If a cluster head node is fixedly used, the energy loss of the sensor as the cluster head node will be very fast, and it will soon die, resulting in a very low data collection efficiency for the entire cluster. Replacing a new cluster head node can well balance the energy loss of all sensor nodes in the same cluster. The old cluster head node with a low energy value can be converted to a normal node for use, and it can still be used under low energy values.
  • step S4 the following formula is used to calculate the probability of each sensor node in the cluster as a new cluster head node, and the new cluster head node is selected according to the calculation result:
  • T(s) represents the probability that the sensor node s is selected as the new cluster head node
  • G represents the set of all ordinary nodes of the cluster in this round
  • w 1 and w 2 represent the constant coefficient of the weight
  • d s represents the sensor node
  • E j is the ratio of the remaining energy of the sensor node s to the initial energy
  • ⁇ is any constant.
  • step S1 the specific process of clustering and arranging several sensor nodes in step S1 includes the following steps:
  • Step S11 When a number of sensor nodes are divided into 1 to n clusters, the corresponding squared distance error SSE from the sensor node in each cluster to the cluster center is calculated,
  • C i represents the set of all sensor nodes in a cluster
  • x represents the coordinate value of the sensor nodes
  • k represents the number of clusters.
  • ⁇ S represents the difference of SSE generated by two adjacent clusters
  • Step S12 Based on the k value obtained above, randomly select k sensor nodes as the initial cluster centers, and then calculate the SSE values from the remaining sensor nodes to the k cluster centers, and divide the remaining sensor nodes separately based on the principle of closest distance Allocate to each initial cluster center, and continuously update the cluster center and the corresponding SSE value until the convergence condition is met.
  • the update equation of the cluster center is:
  • Step S13 After the iteration, the sensor node closest to the cluster center is selected as the initial cluster head node.
  • step S2 an ant algorithm that considers distance and angle at the same time and has a rewarding property is used for running path planning, and the algorithm includes the following steps:
  • Step S21 Suppose there are a total of m ants and N ant location points. All ants are placed at the base station location on the surface of the water. At the same time, a taboo table is set to indicate the set of location points passed by the ants. Each time the ant passes by a location point Then add this point to the taboo table;
  • Step S22 The first ant starts from the position of the base station and calculates the transition probability from position i to position j at that moment Then select the next location point, update the taboo table at the same time, calculate the transition probability, select a new location point, and update the taboo table synchronously until the ant dies after traversing N ant locations, where the transition probability of the ant is :
  • s represents the position of the next moment
  • n(i) represents the set of remaining position points adjacent to the position point i
  • ⁇ and ⁇ respectively represent the weight constant
  • ⁇ ij (t) represents the pheromone concentration
  • ⁇ ij (t) represents the heuristic factor
  • d ij represents the distance from the position point i to j, and r ij represents the offset angle of the position point i to j;
  • Step S23 Calculate the pheromone concentration of the ant on each path
  • Q 1 and Q 2 are arbitrary constants
  • ⁇ is the reward factor, which indicates the degree of influence of the pheromone concentration of the subsequent path on the current path selection, and its value range is 0 to 1
  • P js represents the transition probability of the ant at the next moment
  • Step S24 Repeat the above steps S22 and S23 until m ants traverse all the position points, and calculate the pheromone increment ⁇ ij and the total pheromone ⁇ ij (t+1) on each path,
  • ⁇ ij (t+1) (1- ⁇ ) ⁇ ij (t)+ ⁇ ij ;
  • ⁇ represents the pheromone evaporation coefficient, which represents the amount of pheromone reduction
  • (1- ⁇ ) is the pheromone residual coefficient, which represents the remaining pheromone concentration
  • Step S25 Record the sum of distances and offset angles generated in this iteration, and update the current optimal path according to the recording results;
  • Step S26 Iteration is continuously performed. When the preset number of iterations is reached or a stagnation phenomenon occurs, the iteration is terminated, and the optimal path updated after the last iteration is taken as the running path, otherwise the iteration is continued.
  • the offset angle r ij in the step S22 is calculated by the following formula:
  • x i and y i represent the abscissa and ordinate of the position point i respectively
  • x j and y j represent the abscissa and ordinate of the position point j respectively
  • ⁇ represents the vector The angle with the horizontal direction.

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Computing Systems (AREA)
  • Ocean & Marine Engineering (AREA)
  • Medical Informatics (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Radar, Positioning & Navigation (AREA)
  • Physics & Mathematics (AREA)
  • Remote Sensing (AREA)
  • Health & Medical Sciences (AREA)
  • Chemical & Material Sciences (AREA)
  • Combustion & Propulsion (AREA)
  • Mechanical Engineering (AREA)
  • Artificial Intelligence (AREA)
  • Evolutionary Computation (AREA)
  • Manipulator (AREA)
  • Feedback Control In General (AREA)
  • Arrangements For Transmission Of Measured Signals (AREA)

Abstract

一种应用于海洋信息网络的分层式数据采集系统及方法，其将若干个传感器节点分簇布置，每一簇包括一簇首节点和多个普通节点，通过多个普通传感器节点采集海底数据信息后统一传输至簇首节点，由簇首节点负责数据汇总并将其传输至水下机器人，降低了每个传感器节点的使用能耗，延长了数据采集层的传感器的使用寿命，提高了数据采集层的数据收集效率。并且，每完成一个数据采集周期后，都会在每个簇内重新选择一个新的传感器节点作为下一个数据采集周期的簇首节点，很好地均衡同一个簇内所有传感器节点的能量损耗，通过不断地轮回更换簇首节点，提升了整个簇的数据采集效率，实现了簇内所有传感器节点的能量利用最大化，更加节能高效。

Description

应用于海洋信息网络的分层式数据采集系统及方法 技术领域
本发明涉及海洋数据信息采集技术领域，特别地，涉及一种应用于海洋信息网络的分层式数据采集系统，另外还涉及一种应用于海洋信息网络的分层式数据采集方法。
背景技术
近年来，海洋通信网络受到广泛关注，认识海洋，开发海洋，经略海洋，全面建设海洋强国被提升到新的战略高度，而支撑海洋强国建设，海洋信息网络必不可少，如何设计一种面向海洋信息网络的稳定、高效、低能耗的数据采集系统成为核心要务。
当前常规的数据采集系统是在海洋不同深度布置多个传感器节点以收集传递水下信息，通过由水底向水面的方向以节点多跳的通信方式，使海洋数据最终汇聚到水面的固定基站(船只)。这样的方式存在三个问题，第一，受制于水声信道较弱，通信速率较低的现实问题，多跳的通信方式不仅增大了数据丢包率，且系统时延较大；第二，每个传感器作为固定中继节点，一直处在高负荷工作状态，以致传感器节点耗能过大，死亡率较高，数据采集效率较低；第三，数据传输的链路相对固定，一旦某个节点能量耗尽，整条联络将面临失效的可能，致使系统的鲁棒性较差。
另外，还有一种数据采集系统，它利用布置在海底的传感器来收集数据信息，然后通过水下机器人(AUV)下潜遍历所有的传感器，收集完所有传感器的数据信息后再返回至水面基站，并将收集到的数据信息传输至水面基站，其虽然可以解决常规数据采集系统存在的通信传输不稳定的问题，但是，这种数据采集系统的传感器需要同时兼顾数据收集、融合和传输等多项工作，能量消耗较高，使用寿命较短，数据收集效率也较低。此外其针对水下机器人(AUV)的路径规划也存在不足，普遍存在两个缺陷：第一，其通常采用蚂蚁算法来解决旅行商问题，而常用的蚂蚁算法仅考虑了当前状态下信息素浓度，没有考虑之后状态信息素浓度的影响，因而存在求解效率不高，易陷入局部最优等问题；第二，传统的蚂蚁算法在路径规划问题上仅考虑到了距离的影响，而没有考虑到角度改变带来的影响，因而在AUV运动过程中会产生大量的能耗，影响了整体的数据收集效率。
发明内容
本发明提供了一种应用于海洋信息网络的分层式数据采集系统及方法，以解决目前的面向海洋信息网络的数据采集系统存在的能耗大、效率低的技术问题。
根据本发明的一个方面，提供一种应用于海洋信息网络的分层式数据采集系统，包括水面基站、用于采集海底数据信息的数据采集层、位于所述数据采集层的上方并用于将所述数据采集层采集到的海底数据信息传输至水面基站的数据传输层，所述数据采集层包括若干个锚定在海底的传感器节点，若干个传感器节点分簇布置，每一簇包括一个簇首节点和多个普通节点，多个普通节点采集海底数据信息后将其传输至簇首节点进行数据信息汇聚；所述数据传输层包括水下机器人，所述水下机器人从水面 基站出发后遍历每一个簇的簇首节点以收集汇聚的数据信息后驶回水面基站，并将收集到的海底数据信息传输至水面基站；每完成一个数据采集周期后，在每个簇中重新选择新的传感器节点作为下一个数据采集周期的簇首节点。
进一步地，具体采用以下公式计算簇内其余每个传感器节点作为新的簇首节点的概率，根据计算结果来选择新的簇首节点：
Figure PCTCN2020097722-appb-000001
上式中，T(s)表示传感器节点s被选择为新的簇首节点的概率，G表示该簇本轮次所有普通节点的集合，w 1和w 2表示权重的常系数，d s表示传感器节点s到其对应簇中心的距离的倒数，E j是传感器节点s的剩余能量和初始能量的比值，λ为任一常数。
进一步地，所述水下机器人采用同时考虑距离和角度并具有奖励性质的蚂蚁算法进行运行路径规划，该算法包括以下内容：
假设一共有m个蚂蚁，N ant个位置点，将所有蚂蚁均置于水面基站位置点出发，同时设置一个禁忌表，表示蚂蚁所经过的位置点的集合，蚂蚁每经过一个位置点则将该点添加进禁忌表；
第一个蚂蚁从基站位置点出发，计算该时刻从位置点i到位置点j的转移概率
Figure PCTCN2020097722-appb-000002
然后选择下一个位置点，同时更新禁忌表，再计算转移概率，再选择新的位置点，再同步更新禁忌表，直至遍历N ant个位置点一次，该蚂蚁死亡，其中，蚂蚁的转移概率为：
Figure PCTCN2020097722-appb-000003
上式中，s表示下一时刻的位置，n(i)表示与位置点i相邻的其余位置点的集合，α和β分别表示权重常数，τ ij(t)表示信息素浓度，η ij(t)表示启发因子，且
Figure PCTCN2020097722-appb-000004
d ij表示位置点i到j的距离，r ij表示位置点i向j的偏移角度；
计算该蚂蚁留在各条路径上的信息素浓度
Figure PCTCN2020097722-appb-000005
Figure PCTCN2020097722-appb-000006
Q 1和Q 2为任意常数，γ为奖励因子，表示之后路径的信息素浓度对当前路径选择的影响程度，其取值范围为0～1，P js表示下一时刻蚂蚁的转移概率；
重复执行上述内容，直至m个蚂蚁均遍历所有的位置点，计算各条路径上的信息素增量Δτ ij和信息素总量τ ij(t+1)，
Figure PCTCN2020097722-appb-000007
τ ij(t+1)＝(1-ρ)τ ij(t)+Δτ ij；
其中，ρ为信息素蒸发系数，表示信息素减少的数量，(1-ρ)为信息素残留系数，表示余下的信息素浓度；
记录本次迭代产生的距离之和、偏移角度之和，并根据记录结果更新当前最优路径；
不断进行迭代，当达到预设的迭代次数或者出现停滞现象，则迭代终止，并将最后一次迭代后更新的最优路径作为运行路径，否则继续迭代。
进一步地，位置点i向j的偏移角度r ij通过以下公式计算得到：
Figure PCTCN2020097722-appb-000008
xi和yi分别表示位置点i的横纵坐标，xj和yj分别表示位置点j的横纵坐标，
Figure PCTCN2020097722-appb-000009
表示水下机器人当前的朝向角度，θ表示向量
Figure PCTCN2020097722-appb-000010
与水平方向的夹角。
进一步地，采用以下方式对若干个传感器节点进行分簇布置：
分别计算若干个传感器节点被分为1至n个簇时，对应的每个簇内传感器节点到簇中心的平方距离误差值SSE，
Figure PCTCN2020097722-appb-000011
上式中，C i表示一个簇内所有传感器节点的集合，x表示传感器节点的坐标值，
Figure PCTCN2020097722-appb-000012
表示簇中心的坐标值，k表示簇的数量，当相邻的两个SSE值变化最明显时，对应的簇的值即为需分簇的数量k，求k的过程为：
Figure PCTCN2020097722-appb-000013
其中，ΔS表示两个相邻的簇产生的SSE的差值；
基于上述求出的k值，随机选取k个传感器节点作为初始的簇中心，然后分别计算剩余的传感器节点到这k个簇中心的SSE值，基于距离最近原则将剩余的传感器节点分别分配给每个初始的簇中心，不断更新簇中心及对应的SSE值，直至满足收敛条件，簇中心的更新方程为：
Figure PCTCN2020097722-appb-000014
收敛条件为：|SSE 1-SSE 2|＜ε；
其中，|C i|表示簇内传感器节点的个数，ε为最小阈值，SSE 1和SSE 2分别表示当前迭代产生的SSE值和上一次迭代产生的SSE值；
迭代结束后，选择距离簇中心最近的传感器节点作为初始簇首节点。
另一方面，本发明还提供一种应用于海洋信息网络的分层式数据采集方法，包括以下步骤：
步骤S1：在海底分簇布置若干个传感器节点并进行锚定，每一簇包括簇首节点和多个普通节点，多个普通节点采集海底数据信息后将其传输至簇首节点进行数据信息汇聚；
步骤S2：从水面基站位置处释放水下机器人，控制水下机器人遍历每一个簇的簇首节点以收集汇聚的数据信息后驶回水面基站；
步骤S3：控制水下机器人将收集的海底数据信息传输至水面基站；
步骤S4：每执行一个数据采集周期后，在每个簇中重新选择新的传感器节点作为下一个数据采集周期的簇首节点。
进一步地，所述步骤S4中采用以下公式计算簇内每个传感器节点作为新的簇首节点的概率，根据计算结果来选择新的簇首节点：
Figure PCTCN2020097722-appb-000015
其中，T(s)表示传感器节点s被选择为新的簇首节点的概率，G表示该簇本轮次所有普通节点的集合，w 1和w 2表示权重的常系数，d s表示传感器节点s到其对应簇中心的距离的倒数，E j是传感器节点s的剩余能量和初始能量的比值，λ为任一常数。
进一步地，所述步骤S2中采用同时考虑距离和角度并具有奖励性质的蚂蚁算法进行运行路径规划，该算法包括以下步骤：
步骤S21：假设一共有m个蚂蚁，N ant个位置点，将所有蚂蚁均置于水面基站位置点出发，同时设置一个禁忌表，表示蚂蚁所经过的位置点的集合，蚂蚁每经过一个位置点则将该点添加进禁忌表；
步骤S22：第一个蚂蚁从基站位置点出发，计算该时刻从位置点i到位置点j的转移概率
Figure PCTCN2020097722-appb-000016
然后选择下一个位置点，同时更新禁忌表，再计算转移概率，再选择新的位置点，再同步更新禁忌表，直至遍历N ant个位置点一次，该蚂蚁死亡，其中，蚂蚁的转移概率为：
Figure PCTCN2020097722-appb-000017
上式中，s表示下一时刻的位置，n(i)表示与位置点i相邻的其余位置点的集合，α和β分别表示权重常数，τ ij(t)表示信息素浓度，η ij(t)表示启发因子， 且
Figure PCTCN2020097722-appb-000018
d ij表示位置点i到j的距离，r ij表示位置点i向j的偏移角度；
步骤S23：计算该蚂蚁留在各条路径上的信息素浓度
Figure PCTCN2020097722-appb-000019
Figure PCTCN2020097722-appb-000020
Q 1和Q 2为任意常数，γ为奖励因子，表示之后路径的信息素浓度对当前路径选择的影响程度，其取值范围为0～1，P js表示下一时刻蚂蚁的转移概率；
步骤S24：重复执行上述步骤S22和步骤S23，直至m个蚂蚁均遍历所有的位置点，计算各条路径上的信息素增量Δτ ij和信息素总量τ ij(t+1)，
Figure PCTCN2020097722-appb-000021
τ ij(t+1)＝(1-ρ)τ ij(t)+Δτ ij；
其中，ρ为信息素蒸发系数，表示信息素减少的数量，(1-ρ)为信息素残留系数，表示余下的信息素浓度；
步骤S25：记录本次迭代产生的距离之和、偏移角度之和，并根据记录结果更新当前最优路径；
步骤S26：不断进行迭代，当达到预设的迭代次数或者出现停滞现象，则迭代终止，并将最后一次迭代后更新的最优路径作为运行路径，否则继续迭代。
进一步地，所述步骤S22中的偏移角度r ij通过以下公式计算得到：
Figure PCTCN2020097722-appb-000022
xi和yi分别表示位置点i的横纵坐标，x j和y j分别表示位置点j的横纵坐标，
Figure PCTCN2020097722-appb-000023
表示水下机器人当前的朝向角度，θ表示向量
Figure PCTCN2020097722-appb-000024
与水平方向的夹角。
进一步地，所述步骤S1中对若干个传感器节点进行分簇布置的具体过程包括以下步骤：
步骤S11：分别计算若干个传感器节点被分为1至n个簇时，对应的每个簇内传感器节点到簇中心的平方距离误差值SSE，
Figure PCTCN2020097722-appb-000025
上式中，C i表示一个簇内所有传感器节点的集合，x表示传感器节点的坐标值，
Figure PCTCN2020097722-appb-000026
表示簇中心的坐标值，k表示簇的数量，当相邻的两个SSE值变化最明显时，对应的簇的值即为需分簇的数量k，求k的过程为：
Figure PCTCN2020097722-appb-000027
其中，ΔS表示两个相邻的簇产生的SSE的差值；
步骤S12：基于上述求出的k值，随机选取k个传感器节点作为初始的簇中心，然后分别计算剩余的传感器节点到这k个簇中心的SSE值，基于距离最近原则将剩余的传感器节点分别分配给每个初始的簇中心，不断更新簇中心及对应的SSE值，直至满足收敛条件，簇中心的更新方程为：
Figure PCTCN2020097722-appb-000028
收敛条件为：|SSE 1-SSE 2|＜ε；
其中，|C i|表示簇内传感器节点的个数，ε为最小阈值，SSE 1和SSE 2分别表示当前迭代产生的SSE值和上一次迭代产生的SSE值；
步骤S13：迭代结束后，选择距离簇中心最近的传感器节点作为初始簇首节点。
本发明具有以下效果：
本发明的应用于海洋信息网络的分层式数据采集系统，将若干个传感器节点分簇布置，每一簇包括一个簇首节点和多个普通节点，通过多个普通传感器节点采集海底数据信息后统一传输至簇首节点，由簇首节点负责汇总来自于同一个簇内的各个普通节点的数据并将汇总后的数据传输至水下机器人，普通节点主要执行数据采集功能，而簇首节点主要执行数据传输功能，大大降低了每个传感器节点的使用能耗，延长了数据采集层的传感器的使用寿命，大幅度提高了数据采集层的数据收集效率。并且，每完成一个数据采集周期后，都会在每个簇内重新选择一个新的传感器节点作为下一个数据采集周期的簇首节点，本系统考虑到了簇首节点的能量消耗比普通节点要大，如果固定使用一个簇首节点，那么作为簇首节点的传感器，其能量损耗会非常快，很快就趋于死亡，从而导致整个簇的数据采集效率很低，而本系统在每个采集周期均更换新的簇首节点，可以很好地均衡同一个簇内所有传感器节点的能量损耗，旧的、能量值低的簇首节点可以转变为普通节点使用，在低能量值的情况下也仍然可执行数据采集功能，而新的、能量值高的簇首节点则可以很好地执行数据传输功能，通过不断地轮回更换簇首节点，最终待所有传感器节点的能量值耗尽、均趋近于死亡的情况再对其进行一次性更换，大大提升了整个簇的数据采集效率，实现了簇内所有传感器节点的能量利用最大化，更加节能高效。
另外，本发明的应用于海洋信息网络的分层式数据采集方法同样具有上述优点。
除了上面所描述的目的、特征和优点之外，本发明还有其它的目的、特征和优点。下面将参照图，对本发明作进一步详细的说明。
附图说明
构成本申请的一部分的附图用来提供对本发明的进一步理解，本发明的示意性实施例及其说明用于解释本发明，并不构成对本发明的不当限定。在附图中：
图1是本发明优选实施例的应用于海洋信息网络的分层式数据采集系统的布局示意图。
图2是本发明的优选实施例的应用于海洋信息网络的分层式数据采集系统采用的传感器分簇布置算法与经典的K-means算法、LEACH算法在能耗方面的结果比对示意图。
图3是本发明的优选实施例的应用于海洋信息网络的分层式数据采集系统采用的传感器分簇布置算法与经典的K-means算法、LEACH算法在节点存活率方面的结果比对示意图。
图4是本发明优选实施例的应用于海洋信息网络的分层式数据采集系统中计算偏移角度的示意图。
图5是本发明的优选实施例的应用于海洋信息网络的分层式数据采集系统采用的路径规划算法与经典的蚂蚁算法分别求解出的最短路径结果比较示意图。
图6是本发明的优选实施例的应用于海洋信息网络的分层式数据采集系统采用的路径规划算法与经典的蚂蚁算法分别求解出的偏移角度结果比较示意图。
图7是本发明另一实施例的应用于海洋信息网络的分层式数据采集方法的流程示意图。
图8是图7中的步骤S1中对若干个传感器节点进行分簇布置的子流程示意图。
图9是图7中的步骤S2中采用的路径规划算法的子流程示意图。
具体实施方式
以下结合附图对本发明的实施例进行详细说明，但是本发明可以由下述所限定和覆盖的多种不同方式实施。
如图1所示，本发明的优选实施例提供一种应用于海洋信息网络的分层式数据采集系统，包括水面基站、用于采集海底数据信息的数据采集层、位于所述数据采集层的上方并用于将所述数据采集层采集到的海底数据信息传输至水面基站的数据传输层，其中，所述数据采集层包括若干个锚定在海底的传感器节点，若干个传感器节点分簇布置，每一簇包括一个簇首节点和多个普通节点，多个普通节点采集海底数据信息后将其传输至簇首节点进行数据信息汇聚；所述数据传输层包括水下机器人，所述水下机器人从水面基站出发后遍历每一个簇的簇首节点以收集汇聚的数据信息后驶回水面基站，并将收集到的海底数据信息传输至水面基站；每完成一个数据采集周期后，在每个簇中重新选择新的传感器节点作为下一个数据采集周期的簇首节点。可以理解，所述水下机器人运动至簇首节点上方一定高度处时，例如20m～50m高度处，优选为30m高度处，水下机器人会快速地与簇首节点建立通信链路并进行数据传输，完成当前簇首节点的数据传输后驶向下一个簇首节点，并依次与每一个簇首节点进行数据传输，最终将所有簇首节点的数据带回水面基站。另外，簇首节点与其余普通节点之间通信连接，每个传感器节点均具有自主控制功能，可以在数据采集功能和数据传输功 能之间进行自主切换，具体实现自主切换的方式属于现有技术，不在此不再赘述。可以理解，由于水面基站、数据传输层和数据采集层三者分层设置，故而数据传输层的水下机器人可以直接从水面基站所在平面直接垂直下沉到数据采集层上方一定高度的水平面处，然后按照运行路径在该水平面内进行数据收集；另外，水下机器人还可以直接从水面基站位置处下沉至第一个簇首节点上方的一定高度处，然后按照运行路径沿平面进行数据收集。当然，水下机器人具体沿怎样的路线从水面基站所在水平面移动至数据采集层上方一定高度水平面的，在此不做具体限定。
可以理解，本实施例的应用于海洋信息网络的分层式数据采集系统，将若干个传感器节点分簇布置，每一簇包括一个簇首节点和多个普通节点，通过多个普通传感器节点采集海底数据信息后统一传输至簇首节点，由簇首节点负责汇总来自于同一个簇内的各个普通节点的数据并将汇总后的数据传输至水下机器人，普通节点主要执行数据采集功能，而簇首节点主要执行数据传输功能，大大降低了每个传感器节点的使用能耗，延长了数据采集层的传感器的使用寿命，大幅度提高了数据采集层的数据收集效率。并且，每完成一个数据采集周期后，都会在每个簇内重新选择一个新的传感器节点作为下一个数据采集周期的簇首节点，本系统考虑到了簇首节点的能量消耗比普通节点要大，如果固定使用一个簇首节点，那么作为簇首节点的传感器，其能量损耗会非常快，很快就趋于死亡，从而导致整个簇的数据采集效率很低，而本系统在每个采集周期均更换新的簇首节点，可以很好地均衡同一个簇内所有传感器节点的能量损耗，旧的、能量值低的簇首节点可以转变为普通节点使用，在低能量值的情况下也仍然可执行数据采集功能，而新的、能量值高的簇首节点则可以很好地执行数据传输功能，通过不断地轮回更换簇首节点，最终待所有传感器节点的能量值耗尽、均趋近于死亡的情况再对其进行一次性更换，大大提升了整个簇的数据采集效率，实现了簇内所有传感器节点的能量利用最大化，更加节能高效。
可以理解，为了避免同一簇内产生冗余信息，簇首节点还将对汇聚后的信息进行处理融合，剔除掉重复的数据信息，然后簇首节点将处理融合后的数据信息传输至水下机器人，即簇首节点具有数据预处理的能力。具体地，采用以下方式对若干个传感器节点进行分簇布置：
首先，在海底布放若干个传感器节点，分别计算若干个传感器节点被分为1至n个簇时，对应的每个簇内传感器节点到簇中心的平方距离误差值SSE，其中，每个簇内包含的传感器节点数量为随机分配，平方距离误差值SSE的求解公式如下：
Figure PCTCN2020097722-appb-000029
上式中，C i表示一个簇内所有传感器节点的集合，x表示传感器节点的坐标值，
Figure PCTCN2020097722-appb-000030
表示簇中心的坐标值，k表示簇的数量，k的取值范围为1～n。当相邻的两个SSE值变化最明显时，对应的簇的值即为需分簇的数量k，求k的过程可表示为：
Figure PCTCN2020097722-appb-000031
其中，ΔS表示两个相邻的簇产生的SSE的差值，从而就确定了簇的数量k。然后，基于上述求出的k值，随机选取k个传感器节点作为初始的簇中心，然后分别计算剩余的传感器节点到这k个簇中心的SSE值，基于距离最近原则将剩余的传感器节点分别分配给每个初始的簇中心，在分配的过程中不断更新迭代簇中心及对应的SSE值，直至满足收敛条件，其中，簇中心的更新方程为：
Figure PCTCN2020097722-appb-000032
收敛条件为：|SSE 1-SSE 2＜ε  (4)；
其中，|C i|表示簇内传感器节点的个数，ε为最小阈值，SSE 1和SSE 2分别表示当前迭代产生的SSE值和上一次迭代产生的SSE值，通过不断迭代更新即可得到最终的簇中心的坐标值和每个簇包含的传感器节点。在迭代结束后，选择距离最终的簇中心最近的传感器节点作为初始簇首节点，其余传感器节点则作为普通节点。普通节点开始收集海底数据信息，然后传送给同一簇内的簇首节点，簇首节点汇聚簇内所有普通节点收集的数据后传输至水下机器人，此过程即为一个数据采集周期。
所述水面基站通过上述过程计算出若干个传感器节点的分簇布置方式时，将具体的分簇布置方案传输至水下机器人，由水下机器人下潜遍历所有的传感器节点，并将分簇布置信息对应地传输至每个传感器节点，每个传感器节点都接收到分簇布置信息后即完成分簇布置。
具体地，采用以下公式计算簇内其余每个传感器节点作为新的簇首节点的概率，根据计算结果来选择新的簇首节点，由新的簇首节点来执行数据汇聚和数据传输功能，旧的簇首节点则转换为普通节点执行数据采集功能，计算公式具体如下：
Figure PCTCN2020097722-appb-000033
上式中，T(s)表示传感器节点s被选择为新的簇首节点的概率，G表示该簇本轮次所有普通节点的集合，w 1和w 2表示权重的常系数，d s表示传感器节点s到其对应簇中心的距离的倒数，E j是传感器节点s的剩余能量和初始能量的比值，λ为任一常数。
作为一种选择，在执行完一个数据采集周期后，每个普通节点均进行上述概率计算，然后每个普通节点将各自的概率计算结果均传输至旧的簇首节点，由旧的簇首节点比较出概率值最高者作为新的簇首节点，并生成反馈信号传输至概率值最高者对应的普通节点，该普通节点即从执行数据采集功能自主切换到执行数据汇总和传输功能，即作为新的簇首节点使用，而旧的簇首节点则自动切换到执行数据采集功能，即作为普通节点使用。而作为另一种选择，在执行完一个数据采集周期后，旧的簇首节点自动切换至数据采集功能，即作为普通节点使用，同时基于随机原则或者就近原则选择一个普通节点进行上述概率计算，若计算结果大于预设值，则将其作为新的簇首节点，否则再进行下一个普通节点的概率计算，直至某一个普通节点的概率计算值大于预设 值。作为另一种选择，所述数据采集层完成数据采集和汇总后，在水下机器人下潜之前，所述水面基站可以通过数据仿真计算选出每个簇中新的簇首节点，然后将新的簇首节点信息传输至水下机器人。当水下机器人下潜之后，水下机器人在收集旧的簇首节点汇总的数据信息的同时，将新的簇首节点信息传输至旧的簇首节点，由旧的簇首节点通知新的簇首节点，由新的簇首节点执行数据传输功能，旧的簇首节点则切换至普通节点使用。作为另一种选择，在完成一个数据采集周期后，所述水面基站可以通过数据仿真计算选出每个簇中新的簇首节点，然后将新的簇首节点信息传输至水下机器人。水下机器人按照更新后的运行路径下潜，并将新的簇首节点信息依次对应地传输至传感器节点，接收到新的簇首节点信息的传感器节点即切换为新的簇首节点使用，而旧的簇首节点则自动切换成普通节点使用。
可以理解，上述算法综合考虑了传感器节点与簇中心的距离和传感器节点自身的剩余能量，既确保了传感器节点自身的剩余能量足以支撑数据汇总和数据传输功能，而且其余普通节点到新的簇首节点的距离相当且较短，降低了其它普通节点到新的簇首节点之间的传输距离带来的能量损耗，进一步实现了节能降耗，提高了整个簇内的传感器节点的使用寿命和整体的数据采集效率。
本申请将上述传感器分簇布置算法命名为ECBIK算法，如图2和图3所示，本申请针对节点存活率和能耗两方面分别与经典的K-means算法、LEACH算法进行了比对，从比对结果可以看出，本申请的ECBIK算法首次出现死亡节点的时间和全部节点死亡的时间均最长，全程保持了较高的节点存活率，并且全程都维持了较高的剩余能量，全程能量消耗最少，具有最佳的节能功效。
可以理解，在水下机器人的运动过程中，需要考虑到每两个目的地之间的运行距离和偏移角度，因为运行距离越长、角度改变越大，意味着水下机器人的能量消耗就越多，因而需要选择距离最短且角度变化最小的路径作为运行路径，以降低水下机器人的能量损耗。因此，本系统通过对现有的蚂蚁算法进行了改进，具体地，所述水下机器人采用同时考虑距离和角度并具有奖励性质的蚂蚁算法进行运行路径规划，该算法包括以下内容：
(1)、假设一共有m个蚂蚁，N ant个位置点，将所有蚂蚁均置于水面基站位置点出发，同时设置一个禁忌表，表示蚂蚁所经过的位置点的集合，蚂蚁每经过一个位置点则将该点添加进禁忌表，意味着之后将不会再选择禁忌表内的位置点作为目的地；
(2)、第一个蚂蚁从基站位置点出发，计算该时刻从位置点i到位置点j的转移概率
Figure PCTCN2020097722-appb-000034
然后选择下一个位置点，同时更新禁忌表，再计算转移概率，再选择新的位置点，再同步更新禁忌表，直至遍历N ant个位置点一次，该蚂蚁死亡，其中，蚂蚁的转移概率为：
Figure PCTCN2020097722-appb-000035
上式中，s表示下一时刻的位置，n(i)表示与位置点i相邻的其余位置点的集合，α和β分别表示权重常数，τ ij(t)表示信息素浓度，η ij(t)表示启发因子， 其表达式为：
Figure PCTCN2020097722-appb-000036
其中，di j表示位置点i到j的距离，r ij表示位置点i向j的偏移角度。
另外，如图4所示，水下机器人此刻处于位置点i处，驶向位置点j，则其状态矩阵可表示为：
Figure PCTCN2020097722-appb-000037
其中，η表示位置矩阵，υ表示速度矩阵，u，v，ω分别表示水下机器人的喘振速度、摇摆速度和偏航速度，
Figure PCTCN2020097722-appb-000038
表示水下机器人当前的朝向角度，图4中的θ表示向量
Figure PCTCN2020097722-appb-000039
与水平方向的夹角，定义
Figure PCTCN2020097722-appb-000040
与θ的差值为水下机器人从位置点i向位置点j的偏移角度，即r ij。而位置点i向j的偏移角度r ij可以通过以下公式计算得到：
Figure PCTCN2020097722-appb-000041
其中，x i和y i分别表示位置点i的横纵坐标，x j和y j分别表示位置点j的横纵坐标。
将水下机器人下一时刻位置设为η t+1，则η t+1＝J(η)υ  (10)；
其中，J(η)是运动转移矩阵，其可以表示为：
Figure PCTCN2020097722-appb-000042
因此，由于初始位置点和水下机器人运行速度为已知，故而根据公式(10)和(11)即可算出下一个位置点的坐标，进而求出两个位置点之间的偏移角度r ij。
(3)、计算该蚂蚁留在各条路径上的信息素浓度
Figure PCTCN2020097722-appb-000043
Figure PCTCN2020097722-appb-000044
其中，Q 1和Q 2为任意常数，γ为奖励因子，表示之后路径的信息素浓度对当前路径选择的影响程度，其取值范围为0～1，P js表示下一时刻蚂蚁的转移概率。
(4)、重复执行上述内容(2)至(3)，直至m个蚂蚁均遍历所有的位置点，计算各条路径上的信息素增量Δτ ij和信息素总量τ ij(t+1)，
Figure PCTCN2020097722-appb-000045
τ ij(t+1)＝(1-ρ)τ ij(t)+Δτ ij；
其中，ρ表示信息素蒸发系数，表示信息素减少的数量，(1-ρ)为信息素残留系数，表示余下的信息素浓度。
(5)、记录本次迭代产生的距离之和、偏移角度之和，并根据记录结果更新当前最优路径，具体选择距离之和最小且偏移角度之和最小的路径来作为更新后的最优路径。
(6)、不断进行迭代，当达到预设的迭代次数或者出现停滞现象，则迭代终止，并将最后一次迭代后更新的最优路径作为运行路径，否则回到(1)继续迭代。
可以理解，本申请将上述改进后的路径规划算法命名为R-ACO算法，其综合考虑了距离和角度偏移的影响，求解出的运行路径的运行距离和偏移角度均较小，大幅度降低了水下机器人的能耗，并且在公式(12)中还考虑了未来状态信息素浓度对当前路径选择的影响，具有远视的效果和全局最优的特性，求解效率更高，收敛更快。具体地，如图5和图6所示，本申请针对求解出的最短路径和最小偏离角度两方面与经典的蚂蚁算法ACO进行了对比，从对比结果可以明显地看出，本申请的R-ACO算法求解出的运行距离和偏离角度均比经典的ACO算法更小，且收敛时间更短。
另外，如图7所示，本发明的另一实施例还提供一种应用于海洋信息网络的分层式数据采集方法，其优选采用上述实施例的分层式数据采集系统，所述分层式数据采集方法包括以下步骤：
步骤S1：在海底分簇布置若干个传感器节点并进行锚定，每一簇包括簇首节点和多个普通节点，多个普通节点采集海底数据信息后将其传输至簇首节点进行数据信息汇聚；
步骤S2：从水面基站位置处释放水下机器人，控制水下机器人遍历每一个簇的簇首节点以收集汇聚的数据信息后驶回水面基站；
步骤S3：控制水下机器人将收集的海底数据信息传输至水面基站；
步骤S4：每执行一个数据采集周期后，在每个簇中重新选择新的传感器节点作为下一个数据采集周期的簇首节点。
可以理解，本实施例的应用于海洋信息网络的分层式数据采集方法，通过将若干个传感器节点分簇布置，每一簇包括一个簇首节点和多个普通节点，通过多个普通传感器节点采集海底数据信息后统一传输至簇首节点，由簇首节点负责汇总来自于同一个簇内的各个普通节点的数据并将汇总后的数据传输至水下机器人，普通节点主要执行数据采集功能，而簇首节点主要执行数据传输功能，大大降低了每个传感器节点的使用能耗，延长了数据采集层的传感器的使用寿命，大幅度提高了数据采集层的数据收集效率。并且，每完成一个数据采集周期后，都会在每个簇内重新选择一个新的传感器节点作为下一个数据采集周期的簇首节点，本系统考虑到了簇首节点的能量消耗比普通节点要大，如果固定使用一个簇首节点，那么作为簇首节点的传感器，其能量损耗会非常快，很快就趋于死亡，从而导致整个簇的数据采集效率很低，而本系统在每个采集周期均更换新的簇首节点，可以很好地均衡同一个簇内所有传感器节点的能量损耗，旧的、能量值低的簇首节点可以转变为普通节点使用，在低能量值的情况下也仍然可执行数据采集功能，而新的、能量值高的簇首节点则可以很好地执行数据传 输功能，通过不断地轮回更换簇首节点，最终待所有传感器节点的能量值耗尽、均趋近于死亡的情况再对其进行一次性更换，大大提升了整个簇的数据采集效率，实现了簇内所有传感器节点的能量利用最大化，更加节能高效。
可以理解，所述步骤S4中采用以下公式计算簇内每个传感器节点作为新的簇首节点的概率，根据计算结果来选择新的簇首节点：
Figure PCTCN2020097722-appb-000046
其中，T(s)表示传感器节点s被选择为新的簇首节点的概率，G表示该簇本轮次所有普通节点的集合，w 1和w 2表示权重的常系数，d s表示传感器节点s到其对应簇中心的距离的倒数，E j是传感器节点s的剩余能量和初始能量的比值，λ为任一常数。
可以理解，如图8所示，所述步骤S1中对若干个传感器节点进行分簇布置的具体过程包括以下步骤：
步骤S11：分别计算若干个传感器节点被分为1至n个簇时，对应的每个簇内传感器节点到簇中心的平方距离误差值SSE，
Figure PCTCN2020097722-appb-000047
上式中，C i表示一个簇内所有传感器节点的集合，x表示传感器节点的坐标值，
Figure PCTCN2020097722-appb-000048
表示簇中心的坐标值，k表示簇的数量，当相邻的两个SSE值变化最明显时，对应的簇的值即为需分簇的数量k，求k的过程为：
Figure PCTCN2020097722-appb-000049
其中，ΔS表示两个相邻的簇产生的SSE的差值；
步骤S12：基于上述求出的k值，随机选取k个传感器节点作为初始的簇中心，然后分别计算剩余的传感器节点到这k个簇中心的SSE值，基于距离最近原则将剩余的传感器节点分别分配给每个初始的簇中心，不断更新簇中心及对应的SSE值，直至满足收敛条件，簇中心的更新方程为：
Figure PCTCN2020097722-appb-000050
收敛条件为：|SSE 1-SSE 2|＜ε   (4)；
其中，|C i|表示簇内传感器节点的个数，ε为最小阈值，SSE 1和SSE 2分别表示当前迭代产生的SSE值和上一次迭代产生的SSE值；
步骤S13：迭代结束后，选择距离簇中心最近的传感器节点作为初始簇首节点。
可以理解，如图9所示，所述步骤S2中采用同时考虑距离和角度并具有奖励性质的蚂蚁算法进行运行路径规划，该算法包括以下步骤：
步骤S21：假设一共有m个蚂蚁，N ant个位置点，将所有蚂蚁均置于水面基站位置点出发，同时设置一个禁忌表，表示蚂蚁所经过的位置点的集合，蚂蚁每经过一个位置点则将该点添加进禁忌表；
步骤S22：第一个蚂蚁从基站位置点出发，计算该时刻从位置点i到位置点j的转移概率
Figure PCTCN2020097722-appb-000051
然后选择下一个位置点，同时更新禁忌表，再计算转移概率，再选择新的位置点，再同步更新禁忌表，直至遍历N ant个位置点一次，该蚂蚁死亡，其中，蚂蚁的转移概率为：
Figure PCTCN2020097722-appb-000052
上式中，s表示下一时刻的位置，n(i)表示与位置点i相邻的其余位置点的集合，α和β分别表示权重常数，τ ij(t)表示信息素浓度，η ij(t)表示启发因子，其表达式为：
Figure PCTCN2020097722-appb-000053
d ij表示位置点i到j的距离，r ij表示位置点i向j的偏移角度；
步骤S23：计算该蚂蚁留在各条路径上的信息素浓度
Figure PCTCN2020097722-appb-000054
Figure PCTCN2020097722-appb-000055
Q 1和Q 2为任意常数，γ为奖励因子，表示之后路径的信息素浓度对当前路径选择的影响程度，其取值范围为0～1，P js表示下一时刻蚂蚁的转移概率；
步骤S24：重复执行上述步骤S22和步骤S23，直至m个蚂蚁均遍历所有的位置点，计算各条路径上的信息素增量Δτ ij和信息素总量τ ij(t+1)，
Figure PCTCN2020097722-appb-000056
τ ij(t+1)＝(1-ρ)τ ij(t)+Δτ ij；
其中，ρ表示信息素蒸发系数，表示信息素减少的数量，(1-ρ)为信息素残留系数，表示余下的信息素浓度；
步骤S25：记录本次迭代产生的距离之和、偏移角度之和，并根据记录结果更新当前最优路径；
步骤S26：不断进行迭代，当达到预设的迭代次数或者出现停滞现象，则迭代终止，并将最后一次迭代后更新的最优路径作为运行路径，否则继续迭代。
其中，所述步骤S22中的偏移角度r ij通过以下公式计算得到：
Figure PCTCN2020097722-appb-000057
x i和y i分别表示位置点i的横纵坐标，x j和y j分别表示位置点j的横纵坐标，
Figure PCTCN2020097722-appb-000058
表示水下机器人当前的朝向角度，θ表示向量
Figure PCTCN2020097722-appb-000059
与水平方向的夹角。
另外，还可以理解，所述方法实施例中各个子步骤的具体执行内容与上述系统实施例相一致，故在此不再赘述。
以上所述仅为本发明的优选实施例而已，并不用于限制本发明，对于本领域的技术人员来说，本发明可以有各种更改和变化。凡在本发明的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本发明的保护范围之内。

Claims (10)

  1. 一种应用于海洋信息网络的分层式数据采集系统，包括水面基站、用于采集海底数据信息的数据采集层、位于所述数据采集层的上方并用于将所述数据采集层采集到的海底数据信息传输至水面基站的数据传输层，其特征在于，所述数据采集层包括若干个锚定在海底的传感器节点，若干个传感器节点分簇布置，每一簇包括一个簇首节点和多个普通节点，多个普通节点采集海底数据信息后将其传输至簇首节点进行数据信息汇聚；所述数据传输层包括水下机器人，所述水下机器人从水面基站出发后遍历每一个簇的簇首节点以收集汇聚的数据信息后驶回水面基站，并将收集到的海底数据信息传输至水面基站；每完成一个数据采集周期后，在每个簇中重新选择新的传感器节点作为下一个数据采集周期的簇首节点。
  2. 如权利要求1所述的应用于海洋信息网络的分层式数据采集系统，其特征在于，具体采用以下公式计算簇内其余每个传感器节点作为新的簇首节点的概率，根据计算结果来选择新的簇首节点：
    Figure PCTCN2020097722-appb-100001
    上式中，T(s)表示传感器节点s被选择为新的簇首节点的概率，G表示该簇本轮次所有普通节点的集合，w 1和w 2表示权重的常系数，d s表示传感器节点s到其对应簇中心的距离的倒数，E j是传感器节点s的剩余能量和初始能量的比值，λ为任一常数。
  3. 如权利要求1所述的应用于海洋信息网络的分层式数据采集系统，其特征在于，所述水下机器人采用同时考虑距离和角度并具有奖励性质的蚂蚁算法进行运行路径规划，该算法包括以下内容：
    假设一共有m个蚂蚁，N ant个位置点，将所有蚂蚁均置于水面基站位置点出发，同时设置一个禁忌表，表示蚂蚁所经过的位置点的集合，蚂蚁每经过一个位置点则将该点添加进禁忌表；第一个蚂蚁从基站位置点出发，计算该时刻从位置点i到位置点j的转移概率
    Figure PCTCN2020097722-appb-100002
    然后选择下一个位置点，同时更新禁忌表，再计算转移概率，再选择新的位置点，再同步更新禁忌表，直至遍历N ant个位置点一次，该蚂蚁死亡，其中，蚂蚁的转移概率为：
    Figure PCTCN2020097722-appb-100003
    上式中，s表示下一时刻的位置，n(i)表示与位置点i相邻的其余位置点的集合，α和β分别表示权重常数，τ ij(t)表示信息素浓度，η ij(t)表示启发因子，且
    Figure PCTCN2020097722-appb-100004
    d ij表示位置点i到j的距离，r ij表示位置点i向j的偏移角度；
    计算该蚂蚁留在各条路径上的信息素浓度
    Figure PCTCN2020097722-appb-100005
    Figure PCTCN2020097722-appb-100006
    Q 1和Q 2为任意常数，γ为奖励因子，表示之后路径的信息素浓度对当前路径选择的影响程度，其取值范围为0～1，P js表示下一时刻蚂蚁的转移概率；
    重复执行上述内容，直至m个蚂蚁均遍历所有的位置点，计算各条路径上的信息素增量Δτ ij和信息素总量τ ij(t+1)，
    Figure PCTCN2020097722-appb-100007
    τ ij(t+1)＝(1-ρ)τ ij(t)+Δτ ij；
    其中，ρ为信息素蒸发系数，表示信息素减少的数量，(1-ρ)为信息素残留系数，表示余下的信息素浓度；
    记录本次迭代产生的距离之和、偏移角度之和，并根据记录结果更新当前最优路径；
    不断进行迭代，当达到预设的迭代次数或者出现停滞现象，则迭代终止，并将最后一次迭代后更新的最优路径作为运行路径，否则继续迭代。
  4. 如权利要求3所述的应用于海洋信息网络的分层式数据采集系统，其特征在于，位置点i向j的偏移角度r ij通过以下公式计算得到：
    Figure PCTCN2020097722-appb-100008
    x i和y i分别表示位置点i的横纵坐标，x j和y j分别表示位置点j的横纵坐标，
    Figure PCTCN2020097722-appb-100009
    表示水下机器人当前的朝向角度，θ表示向量
    Figure PCTCN2020097722-appb-100010
    与水平方向的夹角。
  5. 如权利要求1所述的应用于海洋信息网络的分层式数据采集系统，其特征在于，采用以下方式对若干个传感器节点进行分簇布置：
    分别计算若干个传感器节点被分为1至n个簇时，对应的每个簇内传感器节点到簇中心的平方距离误差值SSE，
    Figure PCTCN2020097722-appb-100011
    上式中，C i表示一个簇内所有传感器节点的集合，x表示传感器节点的坐标值，
    Figure PCTCN2020097722-appb-100012
    表示簇中心的坐标值，k表示簇的数量，当相邻的两个SSE值变化最明显时，对应的簇的值即为需分簇的数量k，求k的过程为：
    Figure PCTCN2020097722-appb-100013
    其中，ΔS表示两个相邻的簇产生的SSE的差值；
    基于上述求出k值，随机选取k个传感器节点作为初始的簇中心，然后分别计算剩余的传感器节点到这k个簇中心的SSE值，基于距离最近原则将剩余的传感器节点分别分配给每个初始的簇中心，不断更新簇中心及对应的SSE值，直至满足收敛条件，簇中心的更新方程为：
    Figure PCTCN2020097722-appb-100014
    收敛条件为：|SSE 1-SSE 2|＜ε；
    其中，|C i|表示簇内传感器节点的个数，ε为最小阈值，SSE 1和SSE 2分别表示当前迭代产生的SSE值和上一次迭代产生的SSE值；
    迭代结束后，选择距离簇中心最近的传感器节点作为初始簇首节点。
  6. 一种应用于海洋信息网络的分层式数据采集方法，其特征在于，包括以下步骤：
    步骤S1：在海底分簇布置若干个传感器节点并进行锚定，每一簇包括簇首节点和多个普通节点，多个普通节点采集海底数据信息后将其传输至簇首节点进行数据信息汇聚；
    步骤S2：从水面基站位置处释放水下机器人，控制水下机器人遍历每一个簇的簇首节点以收集汇聚的数据信息后驶回水面基站；
    步骤S3：控制水下机器人将收集的海底数据信息传输至水面基站；
    步骤S4：每执行一个数据采集周期后，在每个簇中重新选择新的传感器节点作为下一个数据采集周期的簇首节点。
  7. 如权利要求6所述的一种应用于海洋信息网络的分层式数据采集方法，其特征在于，所述步骤S4中采用以下公式计算簇内每个传感器节点作为新的簇首节点的概率，根据计算结果来选择新的簇首节点：
    Figure PCTCN2020097722-appb-100015
    其中，T(s)表示传感器节点s被选择为新的簇首节点的概率，G表示该簇本轮次所有普通节点的集合，w 1和w 2表示权重的常系数，d s表示传感器节点s到其对应簇中心的距离的倒数，E j是传感器节点s的剩余能量和初始能量的比值，λ为任一常数。
  8. 如权利要求6所述的一种应用于海洋信息网络的分层式数据采集方法，其特征在于，所述步骤S2中采用同时考虑距离和角度并具有奖励性质的蚂蚁算法进行运行路径规划，该算法包括以下步骤：
    步骤S21：假设一共有m个蚂蚁，N ant个位置点，将所有蚂蚁均置于水面基站位 置点出发，同时设置一个禁忌表，表示蚂蚁所经过的位置点的集合，蚂蚁每经过一个位置点则将该点添加进禁忌表；
    步骤S22：第一个蚂蚁从基站位置点出发，计算该时刻从位置点i到位置点j的转移概率
    Figure PCTCN2020097722-appb-100016
    然后选择下一个位置点，同时更新禁忌表，再计算转移概率，再选择新的位置点，再同步更新禁忌表，直至遍历N ant个位置点一次，该蚂蚁死亡，其中，蚂蚁的转移概率为：
    Figure PCTCN2020097722-appb-100017
    上式中，s表示下一时刻的位置，n(i)表示与位置点i相邻的其余位置点的集合，α和β分别表示权重常数，τ ij(t)表示信息素浓度，η ij(t)表示启发因子，且
    Figure PCTCN2020097722-appb-100018
    d ij表示位置点i到j的距离，r ij表示位置点i向j的偏移角度；
    步骤S23：计算该蚂蚁留在各条路径上的信息素浓度
    Figure PCTCN2020097722-appb-100019
    Figure PCTCN2020097722-appb-100020
    Q 1和Q 2为任意常数，γ为奖励因子，表示之后路径的信息素浓度对当前路径选择的影响程度，其取值范围为0～1，P js表示下一时刻蚂蚁的转移概率；
    步骤S24：重复执行上述步骤S22和步骤S23，直至m个蚂蚁均遍历所有的位置点，计算各条路径上的信息素增量Δτ ij和信息素总量τ ij(t+1)，
    Figure PCTCN2020097722-appb-100021
    τ ij(t+1)＝(1-ρ)τ ij(t)+Δτ ij；
    其中，ρ表示信息素蒸发系数，表示信息素减少的数量，(1-ρ)为信息素残留系数，表示余下的信息素浓度；
    步骤S25：记录本次迭代产生的距离之和、偏移角度之和，并根据记录结果更新当前最优路径；
    步骤S26：不断进行迭代，当达到预设的迭代次数或者出现停滞现象，则迭代终止，并将最后一次迭代后更新的最优路径作为运行路径，否则继续迭代。
  9. 如权利要求8所述的一种应用于海洋信息网络的分层式数据采集方法，其特征在于，所述步骤S22中的偏移角度r ij通过以下公式计算得到：
    Figure PCTCN2020097722-appb-100022
    x i和y i分别表示位置点i的横纵坐标，x j和y j分别表示位置点j的横纵坐标，
    Figure PCTCN2020097722-appb-100023
    表示水下机器人当前的朝向角度，θ表示向量
    Figure PCTCN2020097722-appb-100024
    与水平方向的夹角。
  10. 如权利要求6所述的一种应用于海洋信息网络的分层式数据采集方法，其特征在于，所述步骤S1中对若干个传感器节点进行分簇布置的具体过程包括以下步骤：
    步骤S11：分别计算若干个传感器节点被分为1至n个簇时，对应的每个簇内传感器节点到簇中心的平方距离误差值SSE，
    Figure PCTCN2020097722-appb-100025
    上式中，C i表示一个簇内所有传感器节点的集合，x表示传感器节点的坐标值，
    Figure PCTCN2020097722-appb-100026
    表示簇中心的坐标值，k表示簇的数量，当相邻的两个SSE值变化最明显时，对应的簇的值即为需分簇的数量k，求k的过程为：
    Figure PCTCN2020097722-appb-100027
    其中，ΔS表示两个相邻的簇产生的SSE的差值；
    步骤S12：基于上述求出的k值，随机选取k个传感器节点作为初始的簇中心，然后分别计算剩余的传感器节点到这k个簇中心的SSE值，基于距离最近原则将剩余的传感器节点分别分配给每个初始的簇中心，不断更新簇中心及对应的SSE值，直至满足收敛条件，簇中心的更新方程为：
    Figure PCTCN2020097722-appb-100028
    收敛条件为：|SSE 1-SSE 2|＜ε；
    其中，|C i|表示簇内传感器节点的个数，ε为最小阈值，SSE 1和SSE 2分别表示当前迭代产生的SSE值和上一次迭代产生的SSE值；
    步骤S13：迭代结束后，选择距离簇中心最近的传感器节点作为初始簇首节点。
PCT/CN2020/097722 2020-06-01 2020-06-23 应用于海洋信息网络的分层式数据采集系统及方法 WO2021243767A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US17/436,591 US11305848B2 (en) 2020-06-01 2020-06-23 Layered data acquisition system applied to marine information network and method thereof

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010486637.1A CN111641930B (zh) 2020-06-01 2020-06-01 应用于海洋信息网络的分层式数据采集系统及方法
CN202010486637.1 2020-06-01

Publications (1)

Publication Number Publication Date
WO2021243767A1 true WO2021243767A1 (zh) 2021-12-09

Family

ID=72333327

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/097722 WO2021243767A1 (zh) 2020-06-01 2020-06-23 应用于海洋信息网络的分层式数据采集系统及方法

Country Status (3)

Country Link
US (1) US11305848B2 (zh)
CN (1) CN111641930B (zh)
WO (1) WO2021243767A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115103318A (zh) * 2022-08-24 2022-09-23 江西怡杉环保股份有限公司 一种多节点在线监测方法和系统
CN116828002A (zh) * 2023-08-29 2023-09-29 北京南天智联信息科技股份有限公司 一种基于物联网数据中台的数据处理方法及系统
CN117453447A (zh) * 2023-12-21 2024-01-26 临沂大学 一种汽车电驱动系统的模拟测试方法

Families Citing this family (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112506215B (zh) * 2020-11-18 2022-08-09 广州工程技术职业学院 物联网数据采集方法和无人机
CN112613640B (zh) * 2020-12-07 2024-08-09 清华大学 异构auv协同的水下信息采集系统及能量优化方法
CN113178937B (zh) * 2021-05-27 2023-01-06 珠海创旗科技有限公司 一种自供电海洋物联网节点和系统
CN115243212B (zh) * 2022-07-20 2023-08-08 青岛科技大学 一种基于auv辅助和改进跨层聚类的海洋数据采集方法
CN115865965B (zh) * 2022-11-22 2023-12-19 中山大学 一种基于层次感知的移动目标探测方法、系统及设备
CN115550194B (zh) * 2022-12-01 2023-04-28 中国科学院合肥物质科学研究院 基于类最远采样的区块链网络传输方法及存储介质
CN116506239B (zh) * 2023-06-28 2023-09-19 豪越科技有限公司 一种节能数据处理方法及存储服务器
CN117395273B (zh) * 2023-09-25 2024-05-17 湖北华中电力科技开发有限责任公司 一种基于云数据对比的安全检测方法及系统
CN117555341B (zh) * 2024-01-12 2024-05-24 中国石油大学(华东) 基于改进蚁群算法的深海采矿车路径规划方法及系统
CN118168602A (zh) * 2024-03-11 2024-06-11 自然资源部第一海洋研究所 一种海洋环境数据采集方法及系统
CN117941673B (zh) * 2024-03-26 2024-06-14 中国科学院水生生物研究所 基于多源发声器的声驱网智能监测驱赶方法
CN118310526A (zh) * 2024-04-09 2024-07-09 哈尔滨工业大学（威海） 一种动态拓扑下基于因子图的多auv量测滤波方法及系统

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105072656A (zh) * 2015-07-10 2015-11-18 桂林电子科技大学 基于K-means聚类和蚁群算法的多级异构无线传感器网络分簇路由方法
US20190106867A1 (en) * 2017-10-10 2019-04-11 Michael Antonio Mariano Electronic Water Distribution Center
CN109769222A (zh) * 2019-02-27 2019-05-17 天津城建大学 基于多水下自主航行器的水下传感器网络路由方法
CN110784842A (zh) * 2019-11-04 2020-02-11 常州信息职业技术学院 水声传感器网络中基于多auv位置预测的数据收集方法

Family Cites Families (25)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US5257241A (en) * 1991-05-08 1993-10-26 Atlantic Richfield Company Method and system for acquisition of 3-dimensional marine seismic data
US7319411B2 (en) * 2002-07-18 2008-01-15 Kmg2 Sensors Corporation Network of sensor nodes assemblies and method of remote sensing within liquid environments
GB0321096D0 (en) * 2003-09-09 2003-10-08 British Telecomm Hierarchical routing in ad-hoc networks
JP4592392B2 (ja) * 2004-11-10 2010-12-01 株式会社エヌ・ティ・ティ・ドコモ 制御装置、移動端末及び移動通信方法
US7747710B1 (en) * 2005-02-03 2010-06-29 Dj Inventions, Llc System for detecting changes in preselected measurable conditions
US8953647B1 (en) * 2007-03-21 2015-02-10 Lockheed Martin Corporation High-power laser using thulium-doped fiber amplifier and frequency quadrupling for blue output
EP2140627B1 (en) * 2007-03-30 2012-02-29 BRITISH TELECOMMUNICATIONS public limited company Ad hoc communication system
US8913488B2 (en) * 2008-12-23 2014-12-16 Bce Inc. Methods and systems for enabling end-user equipment at an end-user premise to effect communications having certain origins when an ability of the end-user equipment to communicate via a communication link connecting the end-user equipment to a communications network is disrupted
CA2748266C (en) * 2008-12-23 2017-11-14 Jonathan Allan Arsenault Methods and systems for enabling end-user equipment at an end-user premise to effect communications when an ability of the end-user equipment to communicate via a communication link connecting the end-user equipment to a communications network is disrupted
CA2689884C (en) * 2008-12-23 2017-12-19 Bce Inc. Methods and systems for enabling end-user equipment at an end-user premise to effect communications having certain destinations when an ability of the end-user equipment to communicate via a communication link connecting the end-user equipment to a communications network is disrupted
CA2748272A1 (en) * 2008-12-24 2010-07-01 Bce Inc. System for end user premise event notification
US8686849B2 (en) * 2010-08-10 2014-04-01 Robert Bosch Gmbh Method of alarm handling in wireless sensor networks
US9094119B2 (en) * 2011-07-06 2015-07-28 Huei Meng Chang Communications network for retransmission of signals
KR102109911B1 (ko) * 2013-10-15 2020-05-12 삼성전자주식회사 토폴로지 제어를 위한 방법 및 그 전자 장치
CN104010336B (zh) * 2014-06-12 2017-07-21 河海大学常州校区 一种两级异构分簇的水下无线传感器网络路由方法
US10310126B2 (en) * 2014-12-01 2019-06-04 Subvision Ab System and method for sea bed surveying
CN108180914B (zh) * 2018-01-09 2021-06-18 昆明理工大学 一种基于蚁群改进和尖峰平滑的移动机器人路径规划方法
US10425788B2 (en) * 2018-02-08 2019-09-24 King Fahd University Of Petroleum And Minerals Equal distance different members node placement method and system
GB2575292B (en) * 2018-07-04 2020-07-08 Graphcore Ltd Code Compilation for Scaling Accelerators
CN109040954B (zh) * 2018-07-30 2021-02-02 西南科技大学 一种基于无线传感网络的杆塔状态监测系统及其路由方法
CN110160546B (zh) * 2019-05-10 2022-05-20 安徽工程大学 一种移动机器人路径规划方法
CN110989612A (zh) * 2019-12-17 2020-04-10 哈工大机器人（合肥）国际创新研究院 一种基于蚁群算法的机器人路径规划方法及装置
CN111026126A (zh) * 2019-12-27 2020-04-17 哈尔滨工程大学 一种基于改进蚁群算法的无人艇全局路径多目标规划方法
CN111200856B (zh) * 2020-02-19 2022-02-22 重庆邮电大学 一种无线传感器的多跳最优路径选择方法
CN112351466B (zh) * 2020-10-27 2022-08-16 常州信息职业技术学院 水声传感网中基于auv时空均分部署的数据收集方法

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105072656A (zh) * 2015-07-10 2015-11-18 桂林电子科技大学 基于K-means聚类和蚁群算法的多级异构无线传感器网络分簇路由方法
US20190106867A1 (en) * 2017-10-10 2019-04-11 Michael Antonio Mariano Electronic Water Distribution Center
CN109769222A (zh) * 2019-02-27 2019-05-17 天津城建大学 基于多水下自主航行器的水下传感器网络路由方法
CN110784842A (zh) * 2019-11-04 2020-02-11 常州信息职业技术学院 水声传感器网络中基于多auv位置预测的数据收集方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
HOU MENGTING, ZHAO ZUOPENG, GAO MENG, ZHANG NANA: "Ant Colony Optimization Multipath Routing Algorithm Adopted Angle Factor", COMPUTER ENGINEERING AND APPLICATIONS, vol. 53, no. 1, 18 August 2016 (2016-08-18), pages 107 - 112, XP055877124, ISSN: 1002-8331, DOI: 10.3778/j.issn.1002-8331.1604-0176 *

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115103318A (zh) * 2022-08-24 2022-09-23 江西怡杉环保股份有限公司 一种多节点在线监测方法和系统
CN115103318B (zh) * 2022-08-24 2022-11-01 江西怡杉环保股份有限公司 一种多节点在线监测方法和系统
CN116828002A (zh) * 2023-08-29 2023-09-29 北京南天智联信息科技股份有限公司 一种基于物联网数据中台的数据处理方法及系统
CN117453447A (zh) * 2023-12-21 2024-01-26 临沂大学 一种汽车电驱动系统的模拟测试方法
CN117453447B (zh) * 2023-12-21 2024-03-22 临沂大学 一种汽车电驱动系统的模拟测试方法

Also Published As

Publication number Publication date
US11305848B2 (en) 2022-04-19
US20220041255A1 (en) 2022-02-10
CN111641930B (zh) 2021-04-13
CN111641930A (zh) 2020-09-08

Similar Documents

Publication Publication Date Title
WO2021243767A1 (zh) 应用于海洋信息网络的分层式数据采集系统及方法
Donta et al. Data collection and path determination strategies for mobile sink in 3D WSNs
Liu et al. DRL-UTPS: DRL-based trajectory planning for unmanned aerial vehicles for data collection in dynamic IoT network
CN111862579A (zh) 一种基于深度强化学习的出租车调度方法及系统
CN106054875B (zh) 一种分布式多机器人动态网络连通性控制方法
Khedr et al. Successors of PEGASIS protocol: A comprehensive survey
CN110430547B (zh) UASNs中基于Q-learning的多AUV协作数据收集方法
CN110784842A (zh) 水声传感器网络中基于多auv位置预测的数据收集方法
CN112738752A (zh) 一种基于强化学习的wrsn多移动充电器优化调度方法
Alasarpanahi et al. Energy‐efficient void avoidance geographic routing protocol for underwater sensor networks
Han et al. Sleep-scheduling-based hierarchical data collection algorithm for gliders in underwater acoustic sensor networks
Zhang et al. Energy-aware data gathering mechanism for mobile sink in wireless sensor networks using particle swarm optimization
CN113965948A (zh) 一种基于自适应分簇网络的传感器数据采集方法
Hajiakhondi-Meybodi et al. Deep reinforcement learning for trustworthy and time-varying connection scheduling in a coupled UAV-based femtocaching architecture
CN107708086B (zh) 一种无线传感器和执行器网络的移动能量补充方法
CN117042083A (zh) 一种面向无人集群组网的分布式可靠传输保障方法
CN115915160A (zh) 海上移动边缘计算无盲区网络构建方法
CN111093216A (zh) 一种基于改进二进制粒子群优化的无线传感器网络节点调度方法
Guang et al. A joint optimized data collection algorithm based on dynamic cluster-head selection and value of information in UWSNs
CN115334165B (zh) 一种基于深度强化学习的水下多无人平台调度方法及系统
CN107995114A (zh) 基于密度聚类的容迟网络路由方法
CN112351466B (zh) 水声传感网中基于auv时空均分部署的数据收集方法
CN115243212B (zh) 一种基于auv辅助和改进跨层聚类的海洋数据采集方法
Tarif et al. A review of Energy Efficient Routing Protocols in Underwater Internet of Things
CN115589625A (zh) 一种auv辅助的水声传感网动态分层路由方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20938694

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20938694

Country of ref document: EP

Kind code of ref document: A1