Multi-unmanned aerial vehicle double-layer collaborative navigation method based on event triggering

Abstract

The invention relates to a multi-unmanned aerial vehicle double-layer collaborative navigation method based on event triggering, which comprises the following steps: when the navigation performance of any unmanned aerial vehicle in an unmanned aerial vehicle cluster does not meet the navigation performance requirement, triggering two levels of information fusion collaborative navigation steps; according to the method, each slave in an unmanned aerial vehicle cluster sends local satellite navigation position data and ranging data to a host in the unmanned aerial vehicle cluster, and the host fuses the satellite navigation position data and the ranging data and outputs position estimation and covariance of each unmanned aerial vehicle as a preliminary collaborative navigation result; and then, performing state updating on each unmanned aerial vehicle in the unmanned aerial vehicle cluster by using a state recursion result of the local filter and the preliminary collaborative navigation result to obtain a final collaborative navigation result. According to the method, the fusion strategy of the unmanned aerial vehicle cluster can be dynamically adjusted, so that the navigation performance meets the navigation performance requirement, and meanwhile, the communication burden, the calculation complexity and the system power consumption of collaborative navigation can be reduced.

Description

Technical Field

[0001] The present invention relates to a technology in the field of collaborative navigation, and specifically to a multi-UAV double-layer collaborative navigation method based on event triggering. Background Art

[0002] Multi-UAV cooperative navigation is a technology that uses mutual communication and relative observation between UAVs to improve the absolute navigation performance of UAV clusters and the relative navigation performance between UAVs. For a UAV cluster, cooperative navigation has several different information fusion strategies, including fully centralized, fully distributed, and a variety of fusion strategies in between. Analysis of existing literature shows that domestic and foreign scholars mainly conduct research on cooperative navigation algorithms under specific mission scenarios and fixed information fusion strategies, and propose technologies such as centralized cooperative navigation based on filtering and distributed cooperative navigation based on filtering. The main problems of existing technologies include two points: First, the navigation performance of the cooperative navigation system does not change with the changes in navigation performance requirements, which may cause the actual navigation performance to not meet the navigation performance requirements, or lead to a waste of communication resources / computing resources; Second, the existing centralized cooperative navigation method has a high computational burden and communication load, resulting in a shortened flight time of the UAV cluster. Summary of the invention

[0003] In view of the problem that the existing collaborative navigation technology cannot dynamically adjust the fusion strategy according to the navigation performance requirements and the problem that the centralized collaborative navigation method has high computational complexity / communication volume, the present invention proposes a multi-UAV double-layer collaborative navigation method based on event triggering, which can not only dynamically adjust the fusion strategy of the UAV cluster so that the navigation performance meets the navigation performance requirements, but also reduce the communication burden, computational complexity and system power consumption of collaborative navigation.

[0004] The present invention is achieved through the following technical solutions:

[0005] The present invention relates to a multi-UAV double-layer cooperative navigation method based on event triggering. When the navigation performance of any UAV in a UAV cluster does not meet the navigation performance requirement, that is, the position error covariance output by a navigation filter is greater than or equal to the expected position error covariance, two-level information fusion cooperative navigation steps are triggered, that is, each slave in the UAV cluster sends local satellite navigation position data and ranging data to a host in the UAV cluster, and the host fuses the satellite navigation position and ranging data and outputs the position estimation of each UAV and its covariance as a preliminary cooperative navigation result; then, each UAV in the UAV cluster updates its state using the state recursion result of the local filter and the preliminary cooperative navigation result to obtain a final cooperative navigation result.

[0006] The two levels of information fusion include: the first level is to fuse the observation data of the satellite navigation receiver carried by each UAV with the ranging data between the UAVs to obtain a preliminary collaborative navigation result; the second level is to further fuse the preliminary collaborative navigation results obtained by each UAV using the first level of information fusion with the inertial measurement unit data carried by the UAV to obtain the final collaborative navigation result. Technical Effects

[0007] The present invention introduces an event trigger mechanism and a double-layer information fusion strategy to ensure that the navigation performance of the drone cluster meets the navigation performance requirements; it effectively reduces the computational complexity, communication burden and system power consumption of the collaborative navigation system. BRIEF DESCRIPTION OF THE DRAWINGS

[0008] Figure 1 It is a flow chart of the present invention;

[0009] Figure 2 It is a schematic diagram of an embodiment scenario;

[0010] Figure 3 It is a schematic diagram of the effect of the embodiment. DETAILED DESCRIPTION

[0011] like Figure 1 As shown, this embodiment relates to a multi-UAV dual-layer collaborative navigation method based on event triggering, including:

[0012] Step 1: Each drone in the drone cluster independently builds a local filter and initializes it, that is, sets the initial position, speed and attitude information of the drone.

[0013] The drone cluster includes: a designated drone as a host and the remaining drones as slaves, wherein the host can perform two-way communication with each slave drone.

[0014] like Figure 2 As shown, in this embodiment, a drone cluster consisting of three drones is considered, where drone No. 1 is the master, and drones No. 2 and No. 3 are slaves. Each drone is equipped with a satellite navigation receiver, an inertial measurement unit, and an ultra-wideband sensor.

[0015] In this embodiment, the local filter is implemented by using an error state Kalman filter, and its state vector Among them: δp, δv and δφ represent the inertial navigation position error, velocity error and attitude error in the northeast sky coordinate system respectively; and represent the random drift of the accelerometer and gyroscope respectively.

[0016] Step 2: When the inertial measurement unit observation data of each drone arrives, the state is recursively deduced through the local filter according to the local inertial measurement unit observation data, specifically: in: represents the status update result at time k-1, is the corresponding state error covariance; is the state recursion result at time k, is the corresponding state error covariance; F k|k-1 is the state recursion matrix; Q k is the process noise covariance matrix.

[0017] Step 3: When the satellite navigation receiver data and ranging data arrive, each drone performs local filtering update through a local filter according to the local satellite navigation position information to obtain the local combined navigation result of the drone and the corresponding position error covariance, specifically: Where: K k is the filter gain; I is the unit matrix; z k is the observation vector corresponding to the satellite navigation position information, H k is the corresponding observation matrix, R k is the corresponding observation error covariance; is the state update result at time k, is the corresponding state error covariance, position error covariance

[0018] Step 4: After the satellite navigation data and ranging data of each UAV arrive and the local filtering update in step 3 is completed, when the navigation performance of all UAVs meets the navigation performance requirements, collaborative navigation is not triggered, otherwise collaborative navigation is triggered and step 5 is executed.

[0019] The above-mentioned satisfaction of navigation performance requirements refers to: Where: C REQ,k represents the expected position error covariance.

[0020] Step 5: Each slave sends its local satellite navigation position data and ranging data to the host. The host fuses the satellite navigation position and ranging data and outputs the position estimate and covariance of each UAV as the preliminary collaborative navigation result.

[0021] The fusion is realized based on the weighted least square method. Taking a drone cluster consisting of three drones as an example, the state vector to be estimated is X = [ 1 p; 2 p; 3 p], where: ip represents the position of the i-th UAV; observation vector model The weighted least squares method is used to solve the estimated value of X to achieve fusion, where: represents the unit sight vector from drone i to drone j; represents the satellite navigation position data of drone i, is the corresponding error vector; represents the distance between UAV i and UAV j output by the ultra-wideband ranging module, is the corresponding ranging error, the error covariance Λ of the observation vector Z.

[0022] The weighted least squares method specifically includes:

[0023] a) After setting the initial estimate of the state vector X, iteratively solve Get the least squares estimate of X That is, the preliminary collaborative navigation results about the positions of each UAV, where: ΔZ is equal to is the equivalent observation vector based on the estimated value of X in the previous iteration, and the weight matrix W = Λ -1 ;

[0024] b) Calculate the state estimation error covariance P C =(G T WG) -1 ;

[0025] c) Based on and P C Get the estimated position of UAV i under cooperative navigation and its corresponding error covariance As a preliminary collaborative navigation result.

[0026] Step 6: After the host sends the corresponding preliminary collaborative navigation results to each slave, each drone updates its state using the state recursion result of the local filter and the preliminary collaborative navigation result to obtain the final collaborative navigation result, which is as follows: Where: observation vector c z k Based on the preliminary collaborative navigation results Build obtained.

[0027] Through specific experiments, taking a drone cluster consisting of 3 drones as an example, the data frequency of the satellite navigation receiver is 1Hz, and the standard deviation of the position error in the northeast direction is 5 meters, 5 meters, and 10 meters. The data frequency of the ultra-wideband ranging model is 1Hz, and the standard deviation of the ranging error is 0.1 meters. The data frequency of the inertial measurement unit is 100Hz, and the performance indicators refer to the ADIS16505 chip. Figure 3 As shown in the figure, the positioning error, positioning error covariance and navigation performance requirements of each drone. It can be seen from the figure that the present invention can dynamically adjust the fusion strategy so that the navigation performance meets the navigation performance requirements. At the same time, the results in the figure show that when the local combined navigation performance meets the performance requirements, there is no need for collaborative navigation, thereby reducing the communication volume and calculation amount of the drone cluster navigation system.

[0028] Compared with the prior art, the method triggers the collaborative navigation step only when necessary to ensure that the navigation performance meets the navigation performance requirements; the collaborative navigation step is not executed when it is not necessary, thereby reducing the computational complexity, communication burden and system power consumption of the navigation system.

[0029] The above-mentioned specific implementation can be partially adjusted in different ways by those skilled in the art without departing from the principle and purpose of the present invention. The protection scope of the present invention shall be based on the claims and shall not be limited by the above-mentioned specific implementation. Each implementation scheme within its scope shall be subject to the constraints of the present invention.

Claims

1. A multi-UAV dual-layer collaborative navigation method based on event triggering, characterized in that: When the navigation performance of any UAV in the UAV cluster does not meet the navigation performance requirements, that is, the position error covariance output by the navigation filter is greater than or equal to the expected position error covariance, the two-level information fusion collaborative navigation steps are triggered, that is, each slave in the UAV cluster sends local satellite navigation position data and ranging data to the host in the UAV cluster, and the host fuses the satellite navigation position and ranging data and outputs the position estimate and covariance of each UAV as the preliminary collaborative navigation result; Then, each UAV in the UAV cluster updates its state using the state recursion result of the local filter and the preliminary collaborative navigation result to obtain the final collaborative navigation result.

2. The event-triggered multi-UAV dual-layer collaborative navigation method according to claim 1 is characterized in that: The two levels of information fusion include: the first level is to fuse the observation data of the satellite navigation receiver carried by each UAV with the ranging data between the UAVs to obtain a preliminary collaborative navigation result; the second level is to further fuse the preliminary collaborative navigation results obtained by each UAV using the first level of information fusion with the inertial measurement unit data carried by the UAV to obtain the final collaborative navigation result.

3. The event-triggered multi-UAV dual-layer collaborative navigation method according to claim 1 is characterized in that: The navigation performance requirement means that the position error covariance output by the navigation filter is smaller than the expected position error covariance, that is, when the position error covariance output by the navigation filter is too large, it is considered that the actual navigation performance does not meet the navigation performance requirement.

4. The event-triggered multi-UAV dual-layer collaborative navigation method according to any one of claims 1 to 3, characterized in that: include: Step 1: Each drone in the drone cluster independently builds a local filter and initializes it, that is, sets the initial position, speed and attitude information of the drone; The drone cluster includes: a designated drone as a host, and the remaining drones as slaves, wherein the host communicates bidirectionally with each slave drone; Step 2: When the inertial measurement unit observation data of each drone arrives, the state is recursively deduced through the local filter according to the local inertial measurement unit observation data, specifically: in: represents the status update result at time k-1, is the corresponding state error covariance; is the state recursion result at time k, is the corresponding state error covariance; F k|k-1 is the state recursion matrix; Q k is the process noise covariance matrix; Step 3: When the satellite navigation receiver data and ranging data arrive, each drone performs local filtering update through a local filter according to the local satellite navigation position information to obtain the local combined navigation result of the drone and the corresponding position error covariance, specifically: Where: K k is the filter gain; I is the unit matrix; z k is the observation vector corresponding to the satellite navigation position information, H k is the corresponding observation matrix, R k is the corresponding observation error covariance; is the state update result at time k, is the corresponding state error covariance, position error covariance Step 4: After the satellite navigation data and ranging data of each UAV arrive and the local filtering update in step 3 is completed, when the navigation performance of all UAVs meets the navigation performance requirements, the collaborative navigation is not triggered, otherwise the collaborative navigation is triggered and step 5 is executed; Step 5: Each slave sends its local satellite navigation position data and ranging data to the host. The host fuses the satellite navigation position and ranging data and outputs the position estimate and covariance of each drone as the preliminary collaborative navigation result. Step 6: After the host sends the corresponding preliminary collaborative navigation results to each slave, each drone updates its state using the state recursion result of the local filter and the preliminary collaborative navigation result to obtain the final collaborative navigation result, which is as follows: Where: observation vector c z k Based on the preliminary collaborative navigation results Build obtained.

5. The event-triggered multi-UAV dual-layer collaborative navigation method according to claim 4 is characterized in that: The local filter is implemented by using error state Kalman filtering, and its state vector Among them: δp, δv and δφ represent the inertial navigation position error, velocity error and attitude error in the northeast sky coordinate system respectively; and represent the random drift of the accelerometer and gyroscope respectively.

6. The event-triggered multi-UAV dual-layer collaborative navigation method according to claim 4 is characterized in that: The above-mentioned satisfaction of navigation performance requirements refers to: Where: c REQ,k represents the expected position error covariance.

7. The event-triggered multi-UAV dual-layer collaborative navigation method according to claim 4 is characterized in that: The fusion is realized based on the weighted least square method. Taking a drone cluster consisting of three drones as an example, the state vector to be estimated is X = [ 1 p; 2 p; 3 p], where: i p represents the position of the i-th UAV; observation vector model The weighted least squares method is used to solve the estimated value of X to achieve fusion, where: represents the unit sight vector from drone i to drone j; represents the satellite navigation position data of drone i, is the corresponding error vector; represents the distance between UAV i and UAV j output by the ultra-wideband ranging module, is the corresponding ranging error, the error covariance Λ of the observation vector Z.

8. The event-triggered multi-UAV dual-layer collaborative navigation method according to claim 4 is characterized in that: The weighted least squares method specifically includes: a) After setting the initial estimate of the state vector X, iteratively solve Get the least squares estimate of X That is, the preliminary collaborative navigation results about the positions of each UAV, where: ΔZ is equal to is the equivalent observation vector based on the estimated value of X in the previous iteration, and the weight matrix W = Λ -1 ; b) Calculate the state estimation error covariance P C =(G T WG) -1 ; c) Based on and P C Get the estimated position of UAV i under cooperative navigation and its corresponding error covariance As a preliminary collaborative navigation result.

Images