Unmanned aerial vehicle cluster communication positioning fusion device and method in GNSS denial environment

Abstract

The invention discloses an unmanned aerial vehicle cluster communication positioning fusion device and method in a GNSS denial environment, and relates to the technical field of unmanned aerial vehicle cooperative localization, and the device comprises a sensor sensing unit which obtains distance information, inertial data information and / or GNSS information; the cooperative positioning iterative optimization unit is used for carrying out self-position estimation and cooperative positioning performance evaluation according to the information in the neighbor table, and taking self-position estimation meeting an evaluation result as self-position information; the network construction unit is used for broadcasting a Hello data packet to an adjacent unmanned aerial vehicle regularly and storing data in the received Hello data packet in a neighbor table; the routing decision-making unit is used for carrying out routing strategy selection based on the position information to obtain an optimal transmission path; and the data transmission unit transmits data along the optimal transmission path. According to the invention, through the design of the cooperative positioning algorithm and the routing strategy, the positioning function and the communication function of the unmanned aerial vehicle are promoted mutually, and the autonomous cooperative communication and positioning capability of the unmanned aerial vehicle cluster in a specific environment is improved.

Description

Technical Field

[0001] The present invention relates to the technical field of unmanned aerial vehicle cluster self-organizing network and unmanned aerial vehicle collaborative positioning, and in particular to a unmanned aerial vehicle cluster communication positioning fusion device and method in a GNSS denied environment. Background Art

[0002] Drones have been widely used in emergency rescue, forest search, military operations and other tasks due to their high-speed movement, wide field of view and flexible operation characteristics. In these complex tasks, drones usually work together in clusters, so efficient ad hoc routing technology and accurate location information are crucial to the success of the mission. On the one hand, data transmission between clustered drones depends on ad hoc networks, but in complex drone environments, mutual interference and changing conditions will lead to a significant decrease in communication efficiency. On the other hand, electromagnetic interference, terrain complexity and other external factors will affect the positioning stability and accuracy of drones, thereby adversely affecting mission execution.

[0003] Compared with traditional topology-based routing, routing technology based on location information can better cope with the dynamic environment of drone clusters. Location information helps optimize and make decisions in routing algorithms and improve the efficiency of cluster communication. For example, Yajuan Cui et al. [Y. Cui, H. Tian, ​​C. Chen, W. Ni, H. Wu and G. Nie, "New Geographical Routing Protocol for Three-Dimensional Flying Ad-Hoc Network Based on New Effective Transmission Range," in IEEE Transactions on Vehicular Technology, vol. 72, no. 12, pp. 16135-16147, Dec. 2023, doi: 10.1109 / TVT.2023.3296082.] proposed to select a reliable transmission distance according to the time-varying channel conditions, and comprehensively consider the duration of the link, the distance between nodes and the angle to determine the best next-hop node, so as to enhance the routing protocol based on location information and adapt to the changing environmental requirements in UAV missions; similarly, when the positioning of the UAV cluster is unstable or inaccurate, the communication network, mutual distance and other information of the UAVs can be used to build a collaborative positioning system to provide positioning information and improve positioning accuracy, such as Xiaobo Gu et al. [X.Gu, C.Zheng, Z.Li, G.Zhou, H.Zhou and L.Zhao,

[0004] Based on radio ranging measurements between UAVs, a framework is proposed to jointly estimate clock errors and relative distances for relative positioning, and a closed loop consisting of synchronized two-way ranging, parameter estimation, and clock tuning is constructed to improve positioning accuracy.

[0005] In complex environments, drones can improve positioning accuracy and reliability by integrating communication and positioning technologies, thereby utilizing multi-source data, and ensuring efficient and reliable data transmission under complex conditions. Lu Sun et al. [L.Sun, J.Wang, L.Wan, K.Li, X.Wang and Y.Lin,"Human-UAV Interaction Assisted Heterogeneous UAVSwarm Scheduling for Target Searching in Communication Denial Environment," in IEEE Transactions on Automation Science and Engineering, doi: 10.1109 / TASE.2024.3412005.] proposed a scheduling time slot model specifically for scenarios with poor communication quality. The model allows drones to transmit data during communication time slots, and uses route fitting and Kalman filtering to predict drone positions during communication denial time slots, thereby optimizing drone scheduling management. Chinese patent CN118488379A proposes a drone ultra-wideband positioning coordination method based on a distributed out-of-band network. This method realizes dynamic management of drone positioning nodes by establishing a distributed out-of-band network of drones, and coordinates the ranging and positioning between drone nodes to achieve efficient communication and collaborative positioning between multiple drones. Chinese patent CN117685953A proposes a multi-drone UWB and visual fusion collaborative positioning method, which realizes mutual cooperation and resource sharing through drone networks, uses visual sensors and UWB sensors to perceive the status of companions, and adopts the method of constructing and solving pose graphs to achieve multi-drone collaborative positioning in the absence of a global reference system.

[0006] UAVs have shown great potential in collaborative positioning and data transmission, especially in scenarios with inaccurate positioning, where UAVs can be used for collaborative positioning and communication to perform tasks. However, on the one hand, traditional topology-based routing protocols are difficult to adapt to the dynamic environment of high-speed movement and rapid changes in topology structure of UAVs. Routing protocols based on location information can better adapt to UAV networks, but inaccurate or frequent changes in location information may lead to routing failures, thereby affecting communication quality. On the other hand, UAV collaborative positioning relies on efficient communication between UAVs to ensure real-time synchronization of node location information. How to maintain stable communication in a dynamic environment and ensure the accuracy and timely update of location information has become a key challenge to improve the performance of UAV collaborative positioning systems. At present, research in the field of UAV communication and positioning mainly focuses on routing strategies and information sharing collaborative positioning of UAV networking. Few studies focus on the impact of inaccurate location information on UAV routing, and how to improve the collaborative positioning effect of UAV swarms by optimizing routing strategies. Therefore, in-depth integration of communication and positioning systems to optimize the communication and positioning performance of UAV swarms has broad development prospects. Summary of the invention

[0007] In view of the above problems, the present invention proposes a UAV cluster communication positioning fusion device and method in a GNSS-denied environment. In the collaborative positioning process, the collaborative positioning iterative optimization unit of the device uses its own attitude, speed, proximity, historical position information collected by the sensor perception unit and the attitude, speed, position, and connection status information of the neighboring nodes shared by the network construction unit to iteratively optimize the UAV position information, thereby realizing the precise collaborative positioning of the UAV cluster; in the routing process, the routing decision unit selects the optimal transmission path of the data packet based on the collaborative positioning result and the shared information of the network construction unit, and then the data transmission unit completes the transmission of the data packet along the path to realize the routing function. At the same time, the routing information will be fed back to the network construction unit, and the accuracy of the topology information will be further improved by optimizing the neighbor node table and eliminating routing holes. In this way, the collaborative optimization of the communication and positioning functions of the UAV cluster is realized.

[0008] On the one hand, a UAV cluster communication positioning fusion device in a GNSS denial environment includes the following:

[0009] Positioning module, including a sensor perception unit and a collaborative positioning iterative optimization unit;

[0010] The sensor sensing unit acquires inertial data information in real time; and when outside the GNSS-denied environment area, acquires GNSS information as its own position information;

[0011] The collaborative positioning iteration optimization unit, when the sensor perception unit is within the GNSS denied environment area, calculates the credibility of each neighbor node according to the node degree, and marks the neighbor nodes with credibility higher than the preset credibility threshold as trusted neighbor nodes; constructs a local factor graph according to the trusted neighbor nodes, and iteratively updates the local factor graph by minimizing the objective function based on the inertial data information and the distance information, and obtains its own position estimate from the local factor graph finally obtained; performs collaborative positioning performance evaluation on the own position estimate, and if the collaborative positioning performance evaluation result exceeds the preset performance threshold, uses the own position estimate as its own position information; the objective function is based on the inertial data information;

[0012] Routing module, including network construction unit, routing decision unit and data transmission unit;

[0013] The network construction unit receives a Hello data packet and regularly broadcasts the Hello data packet to neighboring drone nodes; the Hello data packet includes its own location information and node degree; when the network construction unit broadcasts the Hello data packet, the sensor perception unit obtains distance information;

[0014] The routing decision unit selects a routing strategy to obtain an optimal transmission path: before the collaborative positioning iterative optimization unit performs collaborative positioning, an on-demand routing protocol is used, and the obtained routing path is the optimal transmission path; during the collaborative positioning iterative optimization unit performs collaborative positioning, it checks whether the location information is complete; if the location information is incomplete, an on-demand routing protocol is used, and the obtained routing path is the optimal transmission path; if the location information is complete, a neighbor node with the largest utility function value is selected as a forwarding node as the optimal transmission path; the location information includes its own location information, the neighbor node's own location information, and the target node's own location information;

[0015] The data transmission unit performs data transmission according to the optimal transmission path of the routing decision unit.

[0016] Preferably, the distance information and the data in the received Hello data packet are stored in a neighbor table.

[0017] Preferably, the network construction unit receives routing information from the routing decision unit, and uses the routing information to eliminate routing holes and optimize the neighbor table.

[0018] Preferably, the credibility of each neighbor node is calculated according to the node degree, specifically as follows:

[0019] For the current drone i, the credibility of its neighboring node drone j is expressed as:

[0020]

[0021] Among them, C ij represents the credibility of drone j relative to drone i; D j represents the node degree of drone j, delay ij represents the delay of sending Hello packets between UAV i and UAV j; BER ij represents the bit error rate; ω1, ω2, ω3, and ω4 represent weight coefficients, ω1+ω2+ω3+ω4=1; Represents the Fisher information value of UAV j.

[0022] Preferably, the Fisher information value of the drone j is expressed as:

[0023]

[0024] Among them, x j represents the position state of UAV j; E represents the expected value; n(j) represents the neighbor of UAV j; z jk represents the measured distance between UAVs j and k in the distance information; h(·) represents the relative distance between two nodes; ||·|| 2 represents the Euclidean distance; σ jk represents the standard deviation of the distance measurement between UAVs j and k; represents the expected position of UAV k; The standard deviation of the position of drone k.

[0025] Preferably, the weighted least squares method based on the time window is used to iteratively update the local factor graph by minimizing the objective function, which is expressed as:

[0026]

[0027] in, Represents the optimal solution of the UAV position state in the local factor graph; It means finding the parameter x that makes the objective function reach the minimum value; x i represents the position state of UAV i; L represents the sliding time window; N represents the number of factor nodes of the local factor graph constructed by the current UAV; C ij represents the credibility of drone j relative to drone i; z ij Represents the distance observation value between UAV i and UAV j in the distance information; R D represents the distance measurement covariance matrix; Represents a weighted matrix R D The Euclidean distance of I Represents the IMU pre-integrated covariance matrix; Represents a weighted matrix R I The Euclidean distance of It represents the error of the position change of UAV i obtained from the inertial data information.

[0028] Preferably, the utility function for calculating the utility function value is expressed as:

[0029] U i,j (t) = ω LET ·f LET (x i ,x j )+ω d ·f d (x j ,x D )+ω θ ·f θ (x i ,x j ,x D )+ω c ·C ij (t)

[0030] Among them, U i,j (t) represents the utility function of drone j relative to drone i; ω LET ,ω d ,ω,ω c Respectively represent the weight coefficient, ω LET +ω d +ω θ +ω c =1;x i represents the position state of node i; x j represents the location status of neighbor node j; D represents the target node; f LET (x i ,x j ) represents the link duration between UAVs i and j; f d (x j ,x D ) represents the distance from UAV j to the target node D; f θ (x i ,x j ,x D ) represents the angle formed by UAV i as the vertex and j and D as the two sides; C ij Represents the credibility of drone j relative to drone i.

[0031] On the other hand, a method for fusion of UAV cluster communication positioning in a GNSS-denied environment is provided, wherein each UAV in the UAV cluster includes a UAV cluster communication positioning fusion device in a GNSS-denied environment; the method steps are as follows:

[0032] S1, initialize all drone parameters. All drones at the edge of the GNSS-denied environment area obtain GNSS information as their own location information through the sensor perception unit;

[0033] S2, each drone regularly broadcasts a Hello data packet to neighboring drones through the network construction unit, and the Hello data packet includes its own location information and node degree; each drone stores the information in the received Hello data packet in the neighbor table;

[0034] S3, each drone is a node; when there is a need for data transmission between drones, the routing decision unit of the data transmission source node drone initiates a RREQ request to obtain the RREP response of other drones and obtain the routing path; the data transmission unit of the data transmission source node drone transmits the data packet according to the routing path;

[0035] S4, the collaborative positioning iterative optimization unit of each UAV calculates the credibility of each neighbor node according to the node degree in the neighbor table, and marks the neighbor node with a credibility higher than a preset credibility threshold as a trusted neighbor node;

[0036] S5, the collaborative positioning iterative optimization unit of each UAV builds a local factor graph based on the trusted neighbor nodes, and iteratively updates the local factor graph by minimizing the objective function, and obtains its own position estimation from the final local factor graph;

[0037] S6, each UAV node performs collaborative positioning performance evaluation on its own position estimation. If the collaborative positioning performance evaluation result exceeds the preset performance threshold, its own position estimation is used as its own position information and used for the Hello data packet broadcast in the next cycle;

[0038] S7, when there is a need for data transmission between drones during the collaborative positioning process, the routing decision unit of the drone checks whether there is its own location information, the location information of the neighboring nodes, and the location information of the target node; if the location information is complete, the routing decision unit selects the neighboring node with the largest utility function value as the forwarding node as the optimal transmission path; if the location information is incomplete, the routing decision unit obtains the routing path as the optimal transmission path by initiating an RREQ request and obtaining an RREP response; the data transmission unit of the drone transmits the data packet according to the optimal transmission path.

[0039] Preferably, if the location information is incomplete, the routing decision unit obtains a RREP response by initiating an RREQ request, and obtains a routing path as the optimal transmission path, as follows:

[0040] The routing decision unit initiates a RREQ request; once the target node or the intermediate node that knows the path to the target node is found, the node returns a RREP response to inform the data transmission source node of the available routing path for the drone; the intermediate node can either provide a complete path through its routing table or adopt a greedy forwarding strategy to select a forwarding node for data packet forwarding using the utility function value of the neighboring node.

[0041] Compared with the prior art, the present invention has the following beneficial effects:

[0042] In the routing process of the present invention, the routing decision unit selects the optimal transmission path of the data packet based on the collaborative positioning result and the shared information of the network construction unit, and then the data transmission unit completes the transmission of the data packet along the path to realize the routing function; at the same time, the routing information will be fed back to the network construction unit, and the accuracy of the topology information will be further improved by optimizing the neighbor node table and eliminating routing holes, thereby realizing the collaborative optimization of the communication and positioning functions of the drone cluster. BRIEF DESCRIPTION OF THE DRAWINGS

[0043] The present invention is further described in detail below in conjunction with the accompanying drawings;

[0044] Figure 1 It is a schematic diagram of a device for UAV cluster communication positioning fusion in a GNSS-denied environment according to an embodiment of the present invention;

[0045] Figure 2 A system model diagram of a UAV cluster communication positioning fusion method in a GNSS-denied environment according to an embodiment of the present invention;

[0046] Figure 3 It is a system flow chart of the UAV cluster communication positioning fusion method in a GNSS denied environment according to an embodiment of the present invention;

[0047] Figure 4 This is a system timing diagram of a UAV cluster communication positioning fusion device in a GNSS denied environment according to an embodiment of the present invention. DETAILED DESCRIPTION

[0048] The present invention is further described below through specific implementation modes.

[0049] like Figure 1 As shown, a UAV cluster communication positioning fusion device in a GNSS denial environment includes:

[0050] The positioning module 100 includes a collaborative positioning iterative optimization unit 101 and a sensor sensing unit 102 to realize a collaborative positioning function.

[0051] The collaborative positioning iteration optimization unit 101, during the collaborative positioning process, the collaborative positioning iteration optimization unit 101 of the device uses the self-attitude, speed, proximity, historical location information collected by the sensor perception unit 102 and the attitude, speed, location, and connection status information of the neighboring nodes shared by the network construction unit to iteratively optimize the drone location information, thereby realizing the precise collaborative positioning of the drone cluster. Specifically, when the sensor perception unit 102 is within the GNSS denied environment area, the collaborative positioning iteration optimization unit 101 calculates the credibility of each neighboring node according to the node degree in the neighbor table, and marks the neighboring nodes with credibility higher than the preset credibility threshold as trusted neighboring nodes; according to the local factor graph of the trusted neighboring nodes, the local factor graph is iteratively updated by minimizing the objective function based on the inertial data information and the distance information, and the self-position estimate is obtained from the final local factor graph; the collaborative positioning performance evaluation is performed on the self-position estimate, and if the collaborative positioning performance evaluation result exceeds the preset performance threshold, the self-position estimate is used as the self-position information.

[0052] The sensor sensing unit 102 includes a distance measuring unit, an inertial measurement unit (IMU) and a global navigation satellite system (GNSS). The sensor sensing unit 102 collects information including its own attitude, speed, proximity (distance information), inertial data information, GNSS information and historical location information.

[0053] The routing module 200 includes a network construction unit 201, a routing decision unit 202 and a data transmission unit 203 to implement a routing function.

[0054] The network construction unit 201 periodically broadcasts Hello packets to neighboring drone nodes; the Hello packets include its own location information and node degree; and the data in the received Hello packets are stored in a neighbor table, including attitude, speed, location, and connection status information.

[0055] Routing decision unit 202, in the routing process, routing decision unit 202 selects the optimal transmission path of the data packet based on the collaborative positioning result and the shared information of network construction unit 201, and then data transmission unit 203 completes the transmission of the data packet along the path to realize the routing function. At the same time, the routing information will be fed back to the network construction unit 201, and the accuracy of the topology information will be further improved by optimizing the neighbor node table and eliminating routing holes. Specifically, before the collaborative positioning iteration optimization unit 101 performs collaborative positioning, an on-demand routing protocol is adopted, and the obtained routing path is the optimal transmission path; during the collaborative positioning iteration optimization unit 101 performs collaborative positioning, check whether the location information is complete; if the location information is incomplete, an on-demand routing protocol is adopted, and the obtained routing path is the optimal transmission path; if the location information is complete, select the neighbor node with the largest utility function value as the forwarding node as the optimal transmission path; in this embodiment, the location information includes its own location information, the own location information of the neighbor node in the neighbor table, and the own location information of the target node.

[0056] The data transmission unit 203 transmits the data packet along the optimal transmission path.

[0057] See also Figure 2 , Figure 3 and Figure 4 As shown in FIG. 1 , a method for fusion of UAV cluster communication positioning in a GNSS-denied environment is described as follows:

[0058] Step 1: System parameter definition and drone configuration. The behavior of the drone is described by the Gauss-Markov model. Each drone device has a positioning module and a routing module. The speed and position information of the drone obey the Gaussian distribution. In a GNSS-denied environment, the inner area has strong interference and the drone cannot obtain position information; while the edge area has less interference, and the drone in the edge area of ​​the GNSS-denied environment can still receive some GNSS signals and achieve accurate positioning.

[0059] In this embodiment, the system parameters include: the UAV flight space is 1000×1000 meters, the flight altitude H∈[10,500]m, the number of UAVs n∈[80,160], the transmission power is 20dBm, the channel transmission rate is 1Mbps, the IEEE802.11n MAC layer protocol is adopted, the signal bandwidth is 20MHz, and the transmission range is 200 meters.

[0060] Step 2: Network topology initialization: The network construction unit 201 of all drones regularly broadcasts Hello packets, sends its own position, speed and node degree information to neighboring nodes, and stores the received information in the neighbor table to prepare for subsequent collaborative positioning and routing. Among them, the edge area drones can broadcast their location information initially, while the location information of internal drones needs to be updated to a certain extent before it can be obtained and broadcast.

[0061] Step 3: Initial route establishment. When the drone performing the cluster mission needs to transmit the mission data to the designated target node, the routing decision unit 202 of the source node will initiate a routing request (RREQ, Router Request), which is forwarded by the neighboring nodes hop by hop until it reaches the target node or finds the routing path to the target. The target node then returns a routing reply (RREP, Router Reply), and the source node transmits the data packet according to the routing path determined in the routing reply.

[0062] Step 4: Neighbor node credibility evaluation: The collaborative positioning iterative optimization unit 101 of each UAV calculates the credibility C of each neighbor node based on the neighbor table, and marks the nodes with credibility higher than a preset threshold as credible nodes, in preparation for further factor graph optimization.

[0063] Specifically, for the current drone node i, the credibility C of its neighbor node j ij As follows:

[0064]

[0065] Among them, D j represents the node degree of drone j, delay ij represents the delay of sending hello data packets between drones i and j, BER ij represents the bit error rate. ω1, ω2, ω3, and ω4 represent weight coefficients, and the relationship between them satisfies ω1+ω2+ω3+ω4=1.

[0066] represents the Fisher Information (FI) value of UAV j, as follows:

[0067]

[0068] Among them, D j represents the node degree of drone j, delay ij represents the delay of sending hello data packets between drones i and j, BER ij represents the bit error rate. n(j) represents j’s neighbor, z jkrepresents the measured distance between UAV j and k in the distance information, h(·) represents the Euclidean distance between the two node coordinates, σ jk represents the standard deviation of the distance measurement between UAVs j and k, and Represent the expectation and standard deviation of the k position of the drone, respectively. According to the credibility of the neighbor nodes and the set credibility threshold, the node selects the neighbor nodes with higher credibility as factor nodes to participate in the factor graph algorithm optimization.

[0069] Step 5: Factor graph optimization: The collaborative positioning iterative optimization unit 101 of each UAV constructs a factor graph based on the trusted neighbor nodes, and iteratively updates its position estimation using a weighted least squares method based on a time window.

[0070] Specifically, based on the screened factor nodes, the collaborative positioning iterative optimization unit 101 uses a weighted least square method based on a time window to iteratively update the factor graph. The specific formula is as follows:

[0071]

[0072] in, represents the optimal solution of the UAV position state in the local factor graph, It means finding the parameters x, x that minimize the objective function. i represents the position state of UAV i, L is the sliding time window, N is the number of factor nodes of the local factor graph constructed by the current UAV, and C ij represents the credibility between drone i and drone j, z ij Represents the observed distance between node i and node j in the distance information, R D is the distance measure covariance matrix, Represents a weighted matrix R D The Euclidean distance, R I is the IMU pre-integrated covariance matrix, Represents a weighted matrix R I The Euclidean distance of It represents the error of the position change of UAV i obtained from the inertial data information.

[0073] Step 6: Performance evaluation of collaborative positioning results. The drone must calculate the Fisher Information (FI) between itself and its trusted neighbor nodes. If the FI value exceeds the preset threshold FI th , then the positioning result is considered reliable and broadcasts its own location information in the next cycle; otherwise, it does not broadcast.

[0074] Step 7: Data packet forwarding decision. During the collaborative positioning process, before the data transmission unit 203 of the drone forwards the data packet, the routing decision unit 202 will check whether it has the location information of itself, its neighbors, and the target node. If the conditions are met, the forwarding node is selected according to the utility function U of the neighbor node; during the request process, if the location information of the source node is insufficient, an RREQ request can be initiated. Once the target node or the intermediate node that knows the path to the target node is found, the node will return a RREP response to inform the source node of the available path. The intermediate node can either provide a complete path through its routing table, or adopt a greedy forwarding strategy, using the utility function U value of the neighbor node to select a forwarding node for data packet forwarding, and finally pass the data to the target node.

[0075] Specifically, when the drone has a data packet to send, the routing decision unit 202 calculates the utility function U between the drone and the neighbor link, and selects a forwarding node according to the maximum value of the utility function, as follows:

[0076] U i,j (t) = ω LET ·f LET (x i ,x j )+ω d ·f d (x j ,x D )+ω θ ·f θ (x i ,x j ,x D )+ω c ·C ij (t)

[0077] Among them, ω LET ,ω d ,ω θ ,ω c Represents the weight coefficient, satisfying ω LET +ω d +ω θ +ω c =1. x i and x j They represent the position status of node i and its neighbor node j respectively, D represents the target node, and f LET (x i ,x j ) represents the link duration between UAV i and UAV j, f d (x j ,x D ) represents the distance from UAV j to the target node D, f θ (x i ,x j ,xD ) represents the angle formed by the two sides with UAV i as the vertex, j and D as the angle between the two sides, C ij Represents the credibility of drone j relative to drone i.

[0078] The above is only a specific implementation of the present invention, but the design concept of the present invention is not limited to this. Any non-substantial changes to the present invention using this concept shall be deemed as an infringement of the protection scope of the present invention.

Claims

1. A UAV cluster communication positioning fusion device in a GNSS denial environment, characterized in that: include: Positioning module, including a sensor perception unit and a collaborative positioning iterative optimization unit; The sensor sensing unit acquires inertial data information in real time; When outside the GNSS-denied environment, the GNSS information is obtained as its own location information; The collaborative positioning iteration optimization unit, when the sensor perception unit is within the GNSS denied environment area, calculates the credibility of each neighbor node according to the node degree, and marks the neighbor nodes with credibility higher than the preset credibility threshold as trusted neighbor nodes; constructs a local factor graph according to the trusted neighbor nodes, and iteratively updates the local factor graph by minimizing the objective function based on the inertial data information and the distance information, and obtains its own position estimate from the local factor graph finally obtained; performs collaborative positioning performance evaluation on the own position estimate, and if the collaborative positioning performance evaluation result exceeds the preset performance threshold, uses the own position estimate as its own position information; Routing module, including network construction unit, routing decision unit and data transmission unit; The network construction unit receives a Hello data packet and regularly broadcasts the Hello data packet to neighboring drone nodes; the Hello data packet includes its own location information and node degree; when the network construction unit broadcasts the Hello data packet, the sensor perception unit obtains distance information; The routing decision unit selects a routing strategy to obtain an optimal transmission path: before the collaborative positioning iterative optimization unit performs collaborative positioning, an on-demand routing protocol is used, and the obtained routing path is the optimal transmission path; during the collaborative positioning iterative optimization unit performs collaborative positioning, it checks whether the location information is complete; if the location information is incomplete, an on-demand routing protocol is used, and the obtained routing path is the optimal transmission path; if the location information is complete, a neighbor node with the largest utility function value is selected as a forwarding node as the optimal transmission path; the location information includes its own location information, the neighbor node's own location information, and the target node's own location information; The data transmission unit performs data transmission according to the optimal transmission path of the routing decision unit.

2. According to the UAV cluster communication positioning fusion device in a GNSS denial environment according to claim 1, the distance information and the data in the received Hello data packet are stored in a neighbor table.

3. The UAV cluster communication positioning fusion device in a GNSS-denied environment according to claim 2 is characterized in that: The network construction unit receives the routing information from the routing decision unit, and uses the routing information to eliminate routing holes and optimize the neighbor table.

4. The UAV cluster communication positioning fusion device in a GNSS denial environment according to claim 1, characterized in that: The credibility of each neighbor node is calculated according to the node degree, as follows: For the current drone i, the credibility of its neighboring node drone j is expressed as: Among them, C ij represents the credibility of drone j relative to drone i; D j represents the node degree of drone j, delay ij represents the delay of sending Hello packets between UAV i and UAV j; BER ij represents the bit error rate; ω1, ω2, ω3, and ω4 represent weight coefficients, ω1+ω2+ω3+ω4=1; Represents the Fisher information value of UAV j.

5. The UAV cluster communication positioning fusion device in a GNSS-denied environment according to claim 4 is characterized in that: The Fisher information value of the UAV j is expressed as: Among them, x j represents the position state of UAV j; E represents the expected value; n(j) represents the neighbor of UAV j; z jk represents the measured distance between UAVs j and k in the distance information; h(·) represents the relative distance between two nodes; ||·|| 2 represents the Euclidean distance; σ jk represents the standard deviation of the measured distances of UAVs j and k; represents the expected position of UAV k; The standard deviation of the position of drone k.

6. The UAV cluster communication positioning fusion device in a GNSS-denied environment according to claim 1, characterized in that: The weighted least squares method based on the time window is used to iteratively update the local factor graph by minimizing the objective function, which is expressed as: in, Represents the optimal solution of the UAV position state in the local factor graph; It means finding the parameter x that makes the objective function reach the minimum value; x i represents the position state of UAV i; L represents the sliding time window; N represents the number of factor nodes of the local factor graph constructed by the current UAV; C ij represents the credibility of drone j relative to drone i; z ij Represents the distance observation value between UAV i and UAV j in the distance information; R D represents the distance measurement covariance matrix; Represents a weighted matrix R D The Euclidean distance of I Represents the IMU pre-integrated covariance matrix; Represents a weighted matrix R I The Euclidean distance of It represents the error of the position change of UAV i obtained from the inertial data information.

7. The UAV cluster communication positioning fusion device in a GNSS-denied environment according to claim 1, characterized in that: The utility function for calculating the utility function value is expressed as: U i,j (t)=ω LET ·f LET (x i ,x j )+ω d ·f d (x j ,x D )+ωθ·f θ (x i ,x j ,x D )+ω c ·C ij (t) Among them, U i,j (t) represents the utility function of drone j relative to drone i; ω LET ,ω d ,ω θ ,ω c Respectively represent the weight coefficient, ω LET +ω d +ω θ +ω c =1;x i represents the position state of node i; x j represents the location status of neighbor node j; D represents the target node; f LET (x i ,x j ) represents the link duration between UAVs i and j; f d (x j ,x D ) represents the distance from UAV j to the target node D; f θ (x i ,x j ,x D ) represents the angle formed by UAV i as the vertex and j and D as the two sides; C ij Represents the credibility of drone j relative to drone i.

8. The UAV cluster communication positioning fusion device in a GNSS-denied environment according to claim 1, characterized in that: The collaborative positioning performance evaluation specifically calculates the Fisher information between the drone itself and the trusted neighbor nodes.

9. A method for fusion of UAV cluster communication positioning in a GNSS-denied environment, characterized in that: Each drone in the drone cluster includes the drone cluster communication positioning fusion device in a GNSS denial environment as described in any one of claims 1 to 8; the method steps are as follows: S1, initialize all drone parameters. All drones at the edge of the GNSS-denied environment area obtain GNSS information as their own location information through the sensor perception unit; S2, each drone regularly broadcasts a Hello data packet to neighboring drones through the network construction unit, and the Hello data packet includes its own location information and node degree; each drone stores the information in the received Hello data packet in the neighbor table; S3, each drone is a node; when there is a need for data transmission between drones, the routing decision unit of the data transmission source node drone obtains the RREP response of other drones by initiating RREQ request and obtaining the routing path; The data transmission unit of the data transmission source node drone transmits the data packet according to the routing path; S4, the collaborative positioning iterative optimization unit of each UAV calculates the credibility of each neighbor node according to the node degree in the neighbor table, and marks the neighbor node with a credibility higher than a preset credibility threshold as a trusted neighbor node; S5, the collaborative positioning iterative optimization unit of each UAV builds a local factor graph based on the trusted neighbor nodes, and iteratively updates the local factor graph by minimizing the objective function, and obtains its own position estimation from the final local factor graph; S6, each UAV node performs collaborative positioning performance evaluation on its own position estimation. If the collaborative positioning performance evaluation result exceeds the preset performance threshold, its own position estimation is used as its own position information and used for the Hello data packet broadcast in the next cycle; S7, when there is a need for data transmission between UAVs during the collaborative positioning process, the UAV's routing decision unit checks whether there is its own location information, the location information of neighboring nodes, and the location information of the target node; If the location information is complete, the routing decision unit selects the neighbor node with the largest utility function value as the forwarding node as the optimal transmission path; if the location information is incomplete, the routing decision unit initiates an RREQ request to obtain an RREP response and obtains the routing path as the optimal transmission path; the data transmission unit of the drone transmits data packets according to the optimal transmission path.

10. The method for fusion of UAV cluster communication positioning in a GNSS-denied environment according to claim 9, characterized in that: If the location information is incomplete, the routing decision unit obtains a RREP response by initiating an RREQ request, and obtains a routing path as the optimal transmission path, as follows: The routing decision unit initiates a RREQ request; once the target node or the intermediate node that knows the path to the target node is found, the node returns a RREP response to inform the data transmission source node of the available routing path for the drone; the intermediate node can either provide a complete path through its routing table or adopt a greedy forwarding strategy to select a forwarding node for data packet forwarding using the utility function value of the neighboring node.

Images