What are the key technologies for swarm drones
The overall trend of the future development of drones is Atonomy, which can be summarized by the 5S, the first 4 of which are: Small
In the future, robots will be widely used in search and rescue scenarios. If a drone is too large, it will be very difficult for it to explore the environment. In the face of an unknown environment, small drones have stronger autonomy. They are like a group of little bees. However, this also brings some negative challenges. When a drone's size is reduced to even a diameter of 11 centimeters and a weight of 20 grams, it is simply unable to move objects such as wood and stones. Safe
Small and safe drones that won't injure pedestrians even if they collide with them are easier to control in various environments of daily life. Moreover, since the machine's size is smaller, its inertia will also be reduced, allowing it to quickly adjust its balance in case of a collision. Smart & Speed
Drones can avoid obstacles during flight by using sensors, cloud-based control, cameras, and other closed-loop systems. Additionally, they can use computer vision to detect and analyze the features of their environment, enabling them to plan their own routes, just like humans who know how to avoid obstacles.
The fifth "S" is Swarm, which is one of the costs of miniaturization. With smaller payload, the drones can complete fewer tasks. To address this, they draw inspiration from the way honeybees work, allowing multiple drones to collaborate and complete tasks that individual drones cannot perform.
The organization of drone swarms is as follows:
1. Individual drones act independently, and their actions are local and independent.
2. They can act based on local information alone, even if they do not have global information, individual drones can act.
3. Actions are anonymous, independent of identity, and tasks can be completed without knowing individual information.
Advantages of Drone Swarms
Solving conflicts among multiple drones in limited space
For example, how to make future swarms of delivery drones work together like human delivery workers, i.e., keeping good orderly air traffic in a certain area while avoiding obstacles of the same type, requires mutual cooperation. Essentially, it is equivalent to operating a coordinated swarm system.
Meeting functional requirements at low cost and high dispersion Unmanned System Clusters can be implemented with a mix of different platforms, allowing for the achievement of various functions through a series of mechanisms involving a large number of low-cost, dispersed systems working together to complete tasks, which is entirely different from the strategy of investing in the development of expensive, technologically complex multi-task systems. For different types of work objectives, unmanned system clusters can utilize the heterogeneous advantage of mixed pairing to complete tasks efficiently and cost-effectively.
Dynamic Self-Healing Networks
Drones and autonomous systems can collaborate to form a proactive response network that performs information gathering and communication relay actions with self-healing capabilities. The cluster network of unmanned and autonomous systems works together, collecting information separately and adjusting the number of unmanned systems carrying communication payloads as needed to form relay stations with a certain degree of redundancy.
Distributed Cluster Intelligence
A large number of platforms can implement distributed voting to solve problems, such as the problem of determining targets in a cluster operation, by having each platform send signals indicating its judgment of the location of the same target. The distributed voting results often have a very high accuracy rate.
Distributed Detection
The ability of widely distributed sensors to detect actively and passively, as well as to improve positioning accuracy, is clearly beneficial. Multi-platforms can collaborate to achieve precise target location, and when active detection is needed, platforms can use radars with different frequencies and wavelengths for full-spectrum detection, greatly enhancing detection capability.
Reliability
The drone swarm data link network can support redundant backup mechanisms and have a certain self-healing ability to provide availability assurance. The network can monitor established connections and have automatic recovery ability to deal with unexpected interruptions. The swarm should have certain congestion handling and conflict resolution capabilities.
Decentralized self-organizing network enhances fault tolerance, self-healing, and efficient information sharing capability
At present, the communication mode of drones is still mainly based on single-machine communication with ground stations, and information transmission is still centralized. Decentralized drone swarms can utilize self-organizing network technology to realize high-speed information sharing among drones, while improving the swarm's fault tolerance and self-healing ability.
Clustered unmanned aerial vehicle key technologies
Cluster control algorithm In order for multiple unmanned aerial vehicle (UAV) systems to achieve mutual coordination, it is necessary to determine the logical and physical information and control relationships between the UAVs. Research on the system architecture aimed at addressing these issues can integrate the structure and control of the UAV system, ensuring the smooth flow of information and control within the system, and providing a framework for the interaction between UAVs. Cluster control algorithms not only ensure effective coordination between UAVs, but also do not depend on the number of UAVs, meaning that UAVs can leave or join the cluster at any time without affecting the overall structure of the control system.
Communication network design
In the self-organizing system of multi-UAV cooperative tasks, UAVs act as nodes in the communication network, and the spatial distribution of UAVs determines the topology of the network. Different network topologies have different communication performance. Given a communication topology and performance, allocating communication resources based on the tasks to be performed can improve communication quality, which is one of the challenges of cluster technology.
Coupling of control algorithms and communication technology
In order to improve the efficiency of coordinated task completion, multiple UAVs need to exchange information. In order to ensure timely and complete transmission of the interacted information, there are certain requirements for the performance of the communication network. The cooperative control method based on communication quality constraints is to design the multi-unmanned aerial vehicle (UAV) cooperative control method under the current communication service quality constraints, so that the motion of the multi-UAVs not only satisfies the task requirements but also can make the communication network performance constructed by the UAVs meet the demand for timely and complete transmission of information, thereby improving the efficiency of the multi-UAVs in completing tasks.
Task planning technology
In order to achieve effective task coordination among the multi-UAVs and ensure that the control structure is not dependent on the number of UAVs, a distributed hierarchical architecture of the multi-UAV cooperative task self-organizing system is constructed, in which the basic behaviors and simple tasks of the UAVs are autonomously completed by themselves, and when facing complex tasks and tasks requiring cooperation, the current UAV can publish the task information and resource requirements to the network composed of UAVs, and the UAVs can respond based on their current task and resource situation.
In this way, the exit or entry of any UAV will not affect the system organization structure.
Path planning technology If there are unexpected situations during the actual flight of a drone, it must be re-routed to avoid threats. In order to meet the time-effectiveness requirements for coordinated work, the algorithm used for re-routing must have real-time and efficient features. Therefore, it is possible to use the field search characteristics of swarm algorithms, with the sudden threats on the reference trajectory serving as the leader trajectory, and the follower drones only conducting field searches on the sudden threat segment of the reference trajectory, without needing to search the entire trajectory. This can quickly obtain the correction trajectory segment and replace the original sudden threat trajectory segment. Throughout the flight, the drone continuously adjusts the reference trajectory based on the obtained threat information and reaches the target node.
Drone Swarm Control Technology
In mathematics, a drone swarm with a certain spatial distance can be considered a high-order group system with a time-varying formation problem, which is a challenging control problem and the existence of communication delay further complicates the formation analysis.
