The essence and connotation of UAVs are thoroughly analyzed. This paper systematically elaborates on the core characteristics and fundamental nature of UAV autonomy and intelligent control, as well as their interrelationship. It proposes the design concept and engineering implementation elements of UAV autonomous control, rationally planning the level of autonomy. Furthermore, it presents engineering methods and schemes for autonomous intelligent control of drones, constructing the architecture of an autonomous intelligent control system.
From ancient myths and fables, we know that since the dawn of humanity, people have dreamed of flying like birds. Through generations of imagination, effort, and innovation, humans have explored ways to achieve free flight in the sky. Various attempts were made, such as kites, early rockets, fast cars, hot air balloons, and gliders, but none succeeded in achieving true freedom of flight.
In 1903, the Wright brothers successfully completed the first manned flight with their “flight machine-aircraft,†realizing the dream of human flight. As humans dreamed of flying, they also began to imagine a future where aircraft could be controlled from the ground without a pilot onboard—this idea laid the foundation for the concept of drones. In 1916, American engineers Sperry and Lawrence conducted the first drone flight, marking the beginning of drone research. After World War I and II, especially during the Cold War, drones advanced significantly. Over the past century, drones have evolved into a vast family with numerous branches, classified based on fuselage structure, size, weight, flight altitude, range, and application. For example, by fuselage structure, they can be categorized into unmanned helicopters, fixed-wing aircraft, multi-rotor drones, airships, parachute drones, and flapping-wing drones. By size and weight, they are divided into large, medium, small, and micro UAVs. By function, they include military drones (for reconnaissance, surveillance, electronic warfare, etc.), civil drones (used in police, fire, weather), and consumer drones (for recreational purposes). They can also be classified as high-altitude, long-range, or remote-controlled drones.
The development of any technology is always evolving with time. The definition and meaning of drones must be continuously updated. From the early era of manually controlled aircraft to the current phase of autonomous flight, the understanding of drones has deepened. Without clear definitions and improved comprehension, the development of drone technology may face obstacles. This paper studies and analyzes the evolution of UAVs, summarizes their connotations and definitions, explains the principles of UAV control, and discusses the relationship between autonomy and intelligence. It introduces the design concepts and engineering elements of autonomous intelligent control and proposes practical methods and strategies for its implementation.
The term "drone" has evolved over time. Initially, it referred to aircraft without a pilot onboard, known as "pilotless aircraft." With the advancement of radio remote control technology, the term "Remotely Piloted Aerial Vehicle (RPAV)" and "Remotely Piloted Aircraft System (RPAS)" emerged. Later, "Unmanned Aerial Vehicle (UAV)" became the standard term. According to the U.S. Department of Defense and FAA, a UAV is defined as an aircraft that does not carry a pilot, uses aerodynamics for lift, can be either autonomous or remotely piloted, and is reusable with either lethal or non-lethal payloads. This definition highlights three key points: no pilot onboard, the ability to perform missions, and reusability.
Ballistic or semi-ballistic aircraft, cruise missiles, and artillery shells cannot be considered drones because they are not recoverable. There is still debate about whether remote-controlled model aircraft qualify as drones. Some argue that modern model aircraft use advanced control systems and have changing functions, making them difficult to distinguish legally or operationally. However, if these models are used within visual line-of-sight for entertainment, they are generally not considered drones.
Today, the term "UAV" is widely used in technical contexts, while in civilian society, the term "drone" is more common. The distinction between "unmanned aerial vehicle" and "uninhabited aerial vehicle" is subtle. "Unmanned" implies both the absence of a pilot and the lack of human control, whereas "uninhabited" only refers to the absence of a pilot. Therefore, "unmanned aerial vehicle" better captures the true essence of a drone, as it implies the aircraft can operate independently without human intervention.
Over the past century, drone technology has advanced significantly, and its connotation has changed greatly. The most fundamental change lies in the shift in flight control modes. Drones have gone through several developmental stages, including remote piloting, semi-autonomous flight, fully automatic flight, and now, fully autonomous flight. Currently, the highest level of international drones is the combination of full autonomy and local autonomy.
At a narrow definition, drone flight is directly related to human involvement, with humans being isolated from the drone. At a broader sense, drones serve as tools or weapons, and humans must use them. This raises the issue of "human-machine authority": humans are the masters of drones, and drones must obey human commands. However, due to limitations in human control, drones must possess independent capabilities to operate autonomously.
A standard drone should have three working modes: autonomous (automatic) mode, manual intervention mode, and manual maneuvering mode. These modes are set and selected by the operator, considering environmental conditions. The selection follows the principle of "will be outside, the life is not allowed." In autonomous mode, the drone operates according to predefined rules. In manual intervention mode, the operator corrects deviations. In manual maneuvering mode, the operator directly controls the drone in emergencies.
Autonomy and intelligence are two distinct concepts. Autonomy refers to self-determined behavior, while intelligence involves the ability to complete tasks using appropriate strategies. The relationship between the two is complementary: autonomy is the foundation, and intelligence enhances it. Intelligence depends on autonomy, and the level of intelligence is determined by the level of autonomy. Intelligent systems require autonomous decision-making, information processing, and knowledge accumulation.
Intelligence is relative, and different individuals have varying levels of intelligence. These differences come from innate abilities and learned experiences. Nature and human society are filled with contradictions, and what is considered "smart" or "stupid" can change over time. When designing intelligent systems, it's important to understand and leverage these differences to gain strategic advantages.
It is crucial to distinguish between intelligence and intelligence. Intelligence refers to the ability to acquire and apply knowledge, while intelligence involves how well this knowledge aligns with natural laws. High intelligence requires high-level knowledge and reasoning. Drones must be designed with sufficient autonomy and intelligence to meet user needs and follow established relationships between autonomy and intelligence.
Drones operate under human authorization, and while they are empowered by humans, they must be carefully managed to prevent unintended consequences. These issues, though temporary, highlight the importance of balancing autonomy with human oversight. While machines may replace humans in certain tasks, they will not control them. Instead, they may bring challenges or opportunities, depending on the perspective.
Drone autonomy is limited by various factors, including human control, technical capabilities, and environmental conditions. To achieve true autonomy, drones must have independent information acquisition, processing, and execution capabilities. These three layers interact and depend on each other, forming a unified system. Autonomous drones must be able to perceive their environment, make decisions, and act independently without constant external input.
The implementation of intelligent control requires basic functions at each level of the system. These functions must be coordinated and capable of self-learning and improvement. Autonomous information acquisition is critical, and sensors such as inertial, visual, terrain-matching, optical, astronomical, and electromagnetic sensors play a key role. GPS and Beidou data, while useful, are not truly autonomous.
To illustrate, consider airport perception and feature extraction. Artificial markers are not suitable for autonomous systems, as they can be altered or destroyed. Natural features like mountains and rivers provide more reliable and unique information. Drones should use these features for navigation and decision-making.
Humans are the most complex intelligent beings, and drone design should follow similar principles. Complexity management should involve "divide and conquer," prioritizing critical tasks. UAV intelligence should be hierarchical, with three levels: individual safety, group task completion, and fleet cooperation. Each level builds on the previous one, ensuring efficient and safe operation.
To achieve autonomous intelligent control, drones must integrate all available information while ensuring the bottom line of autonomy and intelligence. They should have three information loops: autonomous, non-autonomous, and authoritative. Information acquisition, processing, and application must be done independently. An "agent" on the ground and one on the drone are essential for effective control.
Drones must be integrated into the natural world, closely related to their masters and opponents. True autonomy and intelligence require a balance between independence and interaction. Drones should operate in harmony with nature and humans, avoiding conflicts and ensuring sustainable development.
In conclusion, this paper provides a comprehensive analysis of UAVs, focusing on autonomy and intelligent control. It outlines the architectural framework for autonomous intelligent control systems and proposes practical implementation methods. Although significant progress has been made, further research is needed to advance the engineering applications of autonomous intelligent control. Future work should focus on improving specific methods, decision strategies, and algorithms for autonomous intelligent systems.
Single Phase Online UPS,Tower UPS,Rack Mount UPS,Online Dual Conversion
Shenzhen Unitronic Power System Co., Ltd , https://www.unitronicpower.com