Air cooperative positioning system based on unmanned aerial vehicle group

Abstract

The invention discloses an air cooperative positioning system based on an unmanned aerial vehicle group, and the system comprises a data collection module which collects the multi-modal data of each unmanned aerial vehicle in real time through sensors on a plurality of unmanned aerial vehicles, and the multi-modal data comprise the current position, speed, attitude and surrounding environment information; the cooperative positioning module calculates the accurate position of each unmanned aerial vehicle based on the acquired multi-modal data and a data fusion algorithm; and the group positioning optimization module optimizes the collective positioning precision of the unmanned aerial vehicle group in real time based on the calculated accurate position of each unmanned aerial vehicle in combination with the relative position among the multiple unmanned aerial vehicles. According to the invention, through multi-modal data acquisition, cooperative positioning calculation and relative position map optimization, high-precision, high-robustness and consistent collective positioning of an unmanned aerial vehicle group in a complex environment is realized.

Description

Technical Field

[0001] This invention relates to the field of cooperative positioning technology, and in particular to an aerial cooperative positioning system based on unmanned aerial vehicle (UAV) swarms. Background Technology

[0002] With the widespread application of UAV swarm technology in disaster monitoring, aerial inspection, military reconnaissance and other scenarios, the ability of swarms to work collaboratively has become a key factor in ensuring mission efficiency and safety. Among them, swarm aerial positioning, as the basic support for realizing collective path planning, target guidance and obstacle avoidance control, is directly related to the overall system's operational accuracy and stability. Traditional single-unit GNSS positioning can achieve high accuracy in open environments, but in complex terrain, urban canyons, forest cover or areas with severe electromagnetic interference, it often faces problems of signal blockage and accuracy degradation, making it difficult to meet the collaborative positioning needs of large-scale UAV swarms in dynamic environments.

[0003] Existing cooperative positioning technologies have attempted to enhance system robustness by incorporating multimodal data fusion, relative ranging, and IMU-assisted positioning. However, they generally suffer from the following shortcomings: First, most methods focus on improving the positioning accuracy of individual UAVs, neglecting the positioning consistency problem at the group level, which can easily lead to instability of the overall cluster structure. Second, the relative positional relationships between UAVs are not fully modeled during multi-source data fusion, making it difficult to suppress error accumulation and group dispersion effects. Third, positioning optimization relies solely on static filtering models and lacks real-time adjustment mechanisms in dynamic environments, limiting the system's adaptability to complex scenarios. Therefore, how to achieve high-precision, highly consistent positioning of UAV groups in complex environments without relying on high-precision external positioning equipment is a key issue that current technologies urgently need to overcome. Summary of the Invention

[0004] To address the above problems, this invention provides an aerial cooperative positioning system based on UAV swarms. It not only integrates GNSS, IMU, and environmental perception data to improve the reliability of single-unit positioning, but also actively suppresses group-level positioning errors through relative position constraint maps and position consistency enhancement mechanisms. This significantly improves the stability of the swarm structure and the overall task execution efficiency, and is suitable for UAV cooperative applications in various complex terrains and task scenarios.

[0005] This invention provides an aerial cooperative positioning system based on unmanned aerial vehicle (UAV) swarms, comprising: The data acquisition module is used to collect multimodal data from each UAV. The collaborative positioning module calculates the precise location of each UAV based on the collected multimodal data and data fusion algorithms. The group positioning optimization module optimizes the collective positioning accuracy of the drone group in real time based on the precise location of each drone and the relative position between drones.

[0006] Furthermore, the multimodal data includes current position, velocity, attitude, and surrounding environment information, and is collected through the following methods: The drone's current position and speed are collected in real time using GNSS sensors on the drone. The attitude data of the UAV, including roll angle, pitch angle and yaw angle, is obtained through the inertial measurement unit; Environmental sensors collect real-time information about the surrounding environment, including temperature, air pressure, and humidity.

[0007] By using GNSS sensors to acquire current position and velocity information, the dynamic trajectory of the UAV in the air can be accurately reflected. Combined with attitude data output by the IMU, changes in flight attitude can be perceived in real time, enhancing flight status monitoring capabilities. Furthermore, by collecting environmental information such as temperature, humidity, and air pressure, dynamic compensation of the external environment for flight safety and positioning accuracy can be achieved. The fusion of multi-dimensional data improves the comprehensiveness and accuracy of single-unit positioning and status perception.

[0008] Furthermore, the cooperative positioning module includes: The relative position calculation unit calculates the distance between drones using a ranging sensor and, combined with the current position and speed of each drone, infers their relative position. The data fusion unit uses a data fusion algorithm to fuse the current position, speed, attitude data, surrounding environment information, and relative position of each UAV, and calculates the position of each UAV. The position correction unit is used to correct the drone's position using an optimization algorithm by analyzing the positioning error in real time.

[0009] The range sensor combines position and velocity to calculate relative position, making the spatial relationship between UAVs more accurate. Then, by using data fusion algorithms, position, velocity, attitude and environmental information are integrated with relative position calculation, which can effectively overcome the limitations of a single data source and improve the positioning accuracy and stability of individual UAVs and groups. At the same time, by analyzing errors in real time and using optimization algorithms for correction, the accumulated deviation and external interference can be continuously suppressed, ensuring that the high-precision collaborative positioning capability of UAV groups can still be maintained in complex or signal-blocked environments.

[0010] Furthermore, the calculation of the distance between drones using ranging sensors, and the deduction of relative positions based on current positions and speeds, specifically includes: Using a distance sensor between drones, the straight-line distance between two drones can be measured in real time; The relative positions of the two drones are calculated based on their current location data. Based on the drone's current position and speed, the relative speed and future changes in relative position are calculated.

[0011] By acquiring the straight-line distance between drones in real time through ranging sensors, reliable basic data can be provided for subsequent relative position calculation. By combining the current position coordinates of the drones to construct position vectors, the spatial relationship between individual drones can be accurately represented. Furthermore, by combining speed information to calculate relative speed and future position changes, not only is the accuracy of relative positioning improved, but the ability to predict position evolution trends in dynamic scenarios is also enhanced, thus providing efficient support for the collaborative control and stable formation of drone swarms.

[0012] Furthermore, the data fusion algorithm employs the Kalman filter algorithm, including: Initialize the state vector and error covariance matrix of the Kalman filter; Based on the drone's status and control input, predict the drone's current state, including its position, speed, attitude, and surrounding environment information; The collected multimodal data is compared with the predicted state of the UAV at the current moment, and the state is updated by Kalman gain. The precise location of each drone is obtained through the updated status.

[0013] By initializing the state vector and error covariance matrix, a complete initial state model is established for the UAV. In the prediction phase, the position, velocity, attitude and environmental information are dynamically extrapolated using the state transition matrix and control input, which can estimate the UAV's motion state in advance. In the update phase, multimodal sensor data and predicted values ​​are combined and corrected by Kalman gain, which effectively suppresses the influence of process noise and observation noise. The final output accurate position not only integrates multi-source information, but also has real-time error correction capability, thereby significantly improving the positioning accuracy and robustness of the UAV in complex environments.

[0014] Furthermore, the position correction unit is specifically used for: The positioning error is analyzed by comparing the predicted position of the drone with the collected multimodal data in real time. Based on the error analysis results, an optimization algorithm is used to correct the drone's position, which is then used as the drone's precise location.

[0015] Real-time analysis of the difference between the predicted location and the actual collected data can quickly identify the source of deviation in the positioning process. Then, by using optimization algorithms to correct the UAV's position, the cumulative error caused by sensor noise, environmental interference, or model uncertainty can be effectively suppressed, thereby maintaining high accuracy and stability of the positioning results. This method not only improves the reliability of single-unit positioning, but also provides more accurate basic data support for group collaborative positioning.

[0016] Furthermore, the group positioning optimization module includes: The position constraint unit constructs a relative constraint graph based on the real-time relative positional relationships between multiple UAVs; The position enhancement unit takes the precise position calculated by each UAV as input, combines it with the position deviation information in the relative constraint map, and uses a weighted average method to output a collective positioning result with enhanced position consistency.

[0017] Introducing a relative position constraint mechanism can fully utilize the real-time spatial relationships between UAVs to construct a group-level constraint graph, maintaining the overall stability of the formation structure. Combined with a weighted average method to enhance the consistency of the precise positions of each UAV, it can not only effectively reduce the impact of single-UAV errors on the overall positioning results, but also output more coordinated and consistent group positioning results in complex or unstable signal environments, thereby significantly improving the collaborative operation accuracy and stability of UAV swarms.

[0018] Furthermore, the construction of a relative constraint graph based on the real-time relative positional relationships between multiple drones specifically includes: Determine the set of observable neighbors for each drone based on its communication or sensing radius; Assign edge weights to each edge in the relative constraint graph to represent the expected distance constraints between drones; Construct a relative constraint graph of relative positions. Where V is the set of drone identifiers, E is the set of connecting edges between drone pairs, and W is the set of weights for all edges.

[0019] By combining the adjacency relationship construction mechanism, the observable neighbor set of UAVs can be quickly determined within the communication or perception radius, thereby establishing the topological relationship between the group. Combined with the expected distance weight setting, it can not only quantify the spatial coupling strength between UAVs, but also highlight the role of key neighbors in position constraints. The final constructed relative constraint graph provides structured constraints for group positioning, which helps to suppress the spread of positioning errors of individual nodes and maintain the overall spatial stability and consistency of the cluster.

[0020] Furthermore, the position enhancement unit is specifically used for: Based on each pair of connected UAVs in the relative constraint diagram, calculate the deviation information between them and the expected distance; Each drone calculates its position update suggestion based on the deviation information from its neighbors, weighted by edge weights. The drone optimizes its position based on its current location and the proposed location update, and outputs the optimized drone position.

[0021] Constructing a relative position deviation vector can intuitively characterize the differences between UAVs and the desired constraints, ensuring that the sources of deviation are accurately quantified. Combined with a weighted residual propagation mechanism, each UAV can be assigned a weight based on the importance of its neighbors, thereby highlighting the dominant role of key constraints in position adjustment. Finally, through consistent position optimization, the current position and the updated suggestion are combined to output the optimized result. This not only effectively reduces the impact of single-machine errors on the whole, but also achieves spatial distribution coordination and consistency at the group level, thereby significantly improving the collaborative positioning accuracy and stability of UAV swarms in complex environments.

[0022] Compared with the prior art, the beneficial effects of the present invention are: 1. This invention integrates GNSS, IMU, and various environmental sensors into the data acquisition module, enabling multimodal synchronous acquisition of UAV position, speed, attitude, and environmental information. This effectively improves data integrity and scene perception capabilities, providing not only basic positioning parameters but also reliable support for subsequent fusion and error correction algorithms, ensuring highly reliable data input in complex and dynamic environments.

[0023] 2. This invention significantly improves the independent positioning capability and overall collaborative robustness of each UAV in weak signal scenarios by employing distance and relative position estimation, Kalman filter fusion, and multimodal error correction mechanism through the collaborative positioning module. Even under non-ideal conditions such as GNSS obstruction and sensor data interference, it can achieve accurate estimation of the position of a single UAV, avoiding the impact of individual node errors on the overall cluster positioning accuracy.

[0024] 3. This invention constructs spatial constraint relationships between UAVs in real time by introducing a relative position constraint graph and a position consistency enhancement mechanism in the group positioning optimization module, and performs weighted residual adjustment on the position of each node, and finally outputs the collectively optimized spatial positioning result. This mechanism effectively improves the spatial consistency and overall navigation accuracy of UAV groups in collaborative operations, and is especially suitable for multi-aircraft collaborative scenarios such as formation flight and collaborative reconnaissance. Attached Figure Description

[0025] To more clearly illustrate the technical solutions in the embodiments of this drawing or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this drawing. For those skilled in the art, other drawings can be obtained based on the structures shown in these drawings without creative effort.

[0026] Figure 1 This is a schematic diagram of the system functional modules of the present invention. Detailed Implementation

[0027] To make the objectives, technical solutions, and advantages of this invention clearer, the invention will be described and illustrated below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the invention. All other embodiments obtained by those skilled in the art based on the embodiments provided by this invention without inventive effort are within the scope of protection of this invention.

[0028] This invention provides an aerial cooperative positioning system based on unmanned aerial vehicle (UAV) swarms, such as... Figure 1 As shown, it specifically includes the following modules: The data acquisition module is used to collect multimodal data of each drone in real time through sensors on multiple drones, including current position, speed, attitude and surrounding environment information.

[0029] Specifically, the data acquisition module collects multimodal data in the following ways: The drone's current position (latitude, longitude, altitude) and velocity (horizontal velocity, vertical velocity) are collected in real time using GNSS sensors, specifically as follows: ; ; in, For drones The current location; Latitude; Longitude; For height; For drones speed; Horizontal velocity; This represents the vertical velocity.

[0030] The attitude data of the UAV, including roll angle, pitch angle, and yaw angle, is acquired through an inertial measurement unit (IMU), specifically expressed as follows: ; in, For drones The posture; For roll angle; The pitch angle; This is the yaw angle.

[0031] Environmental sensors (temperature sensors, humidity sensors, air pressure sensors, etc.) are used to collect ambient environmental information in real time, including temperature, air pressure, and humidity, specifically as follows: ; in, For drones Information about the surrounding environment; For temperature; It refers to air pressure; Humidity.

[0032] The collaborative positioning module calculates the precise location of each UAV based on the collected multimodal data and data fusion algorithms.

[0033] Furthermore, the cooperative positioning module includes: The relative position calculation unit calculates the distance between drones using a ranging sensor and, combined with the current position and speed of each drone, infers their relative position.

[0034] Specifically, using inter-drone ranging sensors, the straight-line distance between two drones is measured in real time, and expressed as: ; in, For drones and The measured distance between them; For drones The current position coordinates; For drones The current position coordinates.

[0035] Based on the current position data of the two drones, their relative positions are calculated. Assuming the two drones are in a known reference coordinate system, the relative positions are calculated using position vectors, expressed as: ; in, For drones relative to drones The position vector; , drones and The current position coordinates.

[0036] Based on the drone's current position and speed, the relative speed and future change in relative position are calculated and expressed as follows: ; in, For drones relative to drones The relative velocity.

[0037] ; ; in, , drones , Future location information; , drones , Position information at the initial moment; , drones , speed; t represents time.

[0038] The data fusion unit uses a data fusion algorithm to fuse the current position, speed, attitude data, surrounding environment information, and relative position of each UAV, and calculates the position of each UAV.

[0039] Specifically, the data fusion algorithm uses the Kalman filter algorithm, and the fusion steps are as follows: The initial state vector and error covariance matrix of the Kalman filter are expressed as follows: ; ; in, This is the initial state vector; The initial error covariance matrix; For drones The initial position; For drones The initial velocity; For drones The initial posture; For drones Initial surrounding environment data; , , , , , These are the corresponding position error covariances.

[0040] Based on the UAV's state and control inputs, the current state of the UAV is predicted, including its position, velocity, attitude, and surrounding environment information, as follows: ; in, This is the state transition matrix; To control the input matrix; To control the input (acceleration); This refers to the state at the previous moment; The predicted state.

[0041] ; in, The error covariance matrix is ​​the prediction result. Let be the error covariance matrix of the previous time step. Let be the process noise covariance matrix.

[0042] The collected multimodal data is compared with the predicted state of the UAV at the current moment, and the state is updated using Kalman gain, as shown below: ; ; ; in, The observation matrix; To observe the noise covariance matrix; Kalman gain; These are actual measured values; I is the identity matrix; This is the updated state; This is the updated state error covariance matrix.

[0043] The predicted position of each drone is obtained based on the updated state, and is represented as follows: .

[0044] The position correction unit corrects the drone's position using an optimization algorithm by analyzing the positioning error in real time.

[0045] Specifically, by comparing the predicted position of the UAV with the collected multimodal data in real time, the positioning error is analyzed and expressed as: ; in, The positioning error at the current time k; This refers to the actual measured location; To predict the location.

[0046] Based on the error analysis results, an optimization algorithm is used to correct the drone's position, which is then used as the drone's precise position, represented as follows: ; in, The corrected location of the drone; This is the Kalman gain submatrix.

[0047] The group positioning optimization module optimizes the collective positioning accuracy of the drone group in real time based on the precise location of each drone and the relative position between drones.

[0048] Furthermore, the group positioning optimization module includes: The position constraint unit constructs a relative constraint graph based on the real-time relative positional relationships between multiple UAVs.

[0049] Specifically, based on the communication or sensing radius, the set of observable neighbors for each drone is determined, represented as: ; in, For drones The set of neighbors; For drones The current three-dimensional position vector; drones for neighbors The current three-dimensional position vector; For communication or sensing radius.

[0050] For each edge in the relative constraint graph The edge weights are assigned to represent the expected distance constraints between drones, expressed by the following formula: ; ; in, For drones and The relative distance currently being observed; The edge weight represents the constraint strength. To prevent small constants with a denominator of zero.

[0051] Construct a relative constraint graph of relative positions. Where V is the set of drone identifiers, E is the set of connecting edges between drone pairs, and W is the set of weights for all edges.

[0052] The position enhancement unit takes the precise position calculated by each UAV as input, combines it with the position deviation information in the relative constraint map, and uses a weighted average method to output a collective positioning result with enhanced position consistency.

[0053] Specifically, based on relative constraint graphs For each pair of connected drones, calculate the deviation information between it and the expected distance, as shown below: ; in, This is deviation information; , drones , The precise location; , drones , The desired relative position in the relative position constraint diagram.

[0054] Each drone calculates its position update suggestion based on the deviation information from its neighbors, weighted by edge weights, as follows: ; in, Update the location suggestion for drone i.

[0055] The drone optimizes its position based on its current location and the proposed location update, and outputs the optimized drone position, represented as follows: ; in, The optimized drone position; This is the learning rate.

[0056] This system ensures that UAVs have complete and reliable basic information input in complex environments by collecting multimodal data in real time. It calculates accurate positions by using distance, relative position and fusion algorithms between multiple UAVs, enabling the system to maintain stable positioning even under weak signal or obstruction conditions. Furthermore, the group positioning optimization module uses relative position constraints to achieve group-level consistency correction, thereby improving the accuracy and robustness of overall collaborative positioning and ensuring the efficiency and stability of UAV swarm operations in diverse mission scenarios.

[0057] It should be noted that the present invention is not limited to the above-described embodiments. The above embodiments are merely examples, and any embodiments that have the same structure and perform the same effects as the technical concept within the scope of the present invention are included within the scope of the present invention. Furthermore, various modifications that can be conceived by those skilled in the art to the embodiments, and other ways of constructing by combining some of the constituent elements of the embodiments, without departing from the spirit of the present invention, are also included within the scope of the present invention.

Claims

1. An aerial cooperative positioning system based on a swarm of unmanned aerial vehicles, characterized in that, The application relates to a multi-robot cooperative positioning method and system. The application comprises: a data acquisition module for acquiring multi-modal data of each unmanned aerial vehicle (UAV); a cooperative positioning module for calculating the accurate position of each UAV based on the acquired multi-modal data and a data fusion algorithm; 2. The UAV swarm-based aerial cooperative positioning system of claim 1, wherein, a group positioning optimization module for optimizing the collective positioning accuracy of the UAV group in real time based on the accurate positions of the UAVs and the relative positions between the UAVs. The multi-modal data includes current position, speed, attitude and surrounding environment information, and the multi-modal data is acquired in the following manner: The current position and speed are acquired in real time through a GNSS sensor on the UAV; The attitude data of the UAV, including roll angle, pitch angle and yaw angle, are acquired through an inertial measurement unit; 3. The UAV swarm-based aerial cooperative positioning system of claim 2, wherein, The surrounding environment information, including temperature, air pressure and humidity, are acquired in real time through an environment sensor. The cooperative positioning module comprises: a relative position calculation unit for calculating the distance between the UAVs through a distance measuring sensor and calculating the relative position by combining the current position and speed of each UAV; a data fusion unit for fusing the current position, speed, attitude data, surrounding environment information and relative position of each UAV by using a data fusion algorithm to calculate the position of each UAV; 4. The UAV swarm-based aerial cooperative positioning system of claim 3, wherein, a position correction unit for correcting the position of the UAV by analyzing the positioning error in real time and using an optimization algorithm. The distance between the UAVs is calculated through a distance measuring sensor, and the relative position is calculated by combining the current position and speed, and the specific process comprises the following steps: The straight-line distance between two UAVs is measured in real time by using a distance measuring sensor between the UAVs; The relative position is calculated according to the current position data of the two UAVs; 5. The UAV swarm-based aerial cooperative positioning system of claim 3, wherein, The relative speed and future relative position change are calculated according to the current position and speed of the UAV. The data fusion algorithm adopts a Kalman filter algorithm, which comprises the following steps: The state vector and error covariance matrix of the Kalman filter are initialized; The state of the UAV at the current time, including position, speed, attitude and surrounding environment information, is predicted based on the state of the UAV and the control input; The collected multi-modal data and the predicted state of the UAV at the current time are compared, and the state is updated through Kalman gain; 6. The UAV swarm-based aerial cooperative positioning system of claim 3, wherein, The accurate position of each UAV is obtained through the updated state. The position correction unit is specifically used for: Analyzing the positioning error by comparing the predicted position of the UAV with the collected multi-modal data in real time; 7. The UAV swarm-based aerial cooperative positioning system of claim 1, wherein, Correcting the position of the UAV by using an optimization algorithm according to the error analysis result, as the accurate position of the UAV. The group positioning optimization module comprises: a position constraint unit for constructing a relative constraint graph based on the real-time relative position relationship between the multiple UAVs; 8. The UAV swarm-based aerial cooperative positioning system of claim 7, wherein, a position reinforcement unit for taking the accurate positions calculated by the UAVs as input, combining the position deviation information in the relative constraint graph, and outputting the collective positioning result with position consistency reinforced by using a weighted average method. The relative constraint graph is constructed based on the real-time relative position relationship between the multiple UAVs, and the specific process comprises the following steps: The observable neighbor set of each UAV is determined according to the communication or perception radius; Constructing relative constraint graph of relative position wherein V is a set of UAV identifiers, E is a set of connecting edges between pairs of UAVs, and W is a set of weights for all edges.

9. The UAV swarm-based aerial cooperative positioning system of claim 1, wherein, An edge weight is assigned to each edge in the relative constraint graph to represent the expected distance constraint between the UAVs; The position reinforcement unit is specifically used for: Based on each pair of connected UAVs in the relative constraint graph, deviation information corresponding to the expected distance is calculated; According to the deviation information with neighbors, each UAV calculates a position update suggestion weighted by edge weights; The UAVs optimize the current position according to the position update suggestion, and output the optimized UAV position.

Images