Drone control system, drone control device, and drone control method

Abstract

This invention provides a drone control system, drone control device, and drone control method that automatically generate an optimal controller according to the aircraft configuration based on a description language capable of describing the unique physical characteristics of drones, thereby achieving high-precision and flexible flight control even for drones with special shapes or changes in center of gravity. [Solution] The system comprises a UADF description unit for describing the physical characteristics of a drone, a control logic generation unit that automatically generates control logic suitable for the drone based on the physical characteristics information described by the UADF description unit, and a flight control unit that executes flight control of the drone based on the control logic generated by the control logic generation unit. The UADF description unit describes the physical characteristics of the drone in UADF (Universal Aircraft Description Format) format, and the UADF format is a general-purpose aircraft description format that adds aircraft-specific joint types, including tilt rotors, propulsion units, and control surfaces, to the URDF robot description format.

Description

[Technical Field]

[0001] The present invention relates to a drone control system, a drone control device, and a drone control method. [Background Art]

[0002] In recent years, with the popularization of drones (unmanned aerial vehicles), demand for drones with various airframe shapes and configurations for industrial and research applications has been increasing. Conventional drone control technologies are mainly designed on the premise of a typical airframe configuration such as a four-rotor quadcopter, and flexible control for airframes with changes in center of gravity position, special shapes, asymmetric propeller arrangements, and the like remains a challenge.

[0003] Against this background, development of drone control systems using model-based control and optimization technologies is progressing. Patent Document 1 discloses a design technique that assists in solving multivariable feedback control problems, and discloses a system in which a user interactively designs a control system and performs performance optimization using H∞ constraints. Patent Document 2 discloses a method for generating an optimization parameter set based on a computer optimization method and improving system operation efficiency using a surrogate model for manufacturing or controlling a technical system. Patent Document 3 discloses a method for realizing accurate motion control by simulating with a 3D model and using dynamic coupling and PWM control in the motion control of an articulated bionic dolphin. [Prior Art Documents] [Patent Documents]

[0004] [Patent Document 1] Japanese Unexamined Patent Application Publication No. 2016-106300 [Patent Document 2] Japanese National Publication of International Patent Application No. 2023-515640 [Patent Document 3] Japanese Unexamined Patent Application Publication No. 2024-042639 [Summary of the Invention] [Problems that the invention aims to solve]

[0005] However, Patent Document 1 describes a multivariable control system design method that emphasizes an interactive user interface, requiring manual controller adjustments each time the drone's airframe configuration changes, and thus unable to automatically generate controllers for diverse airframe shapes.

[0006] Patent Document 2 describes a general parameter optimization method applicable to the entire technical system, and therefore cannot generate a controller that takes into account the physical characteristics specific to drones (tilt rotor, propulsion unit, steering surface, etc.), nor does it provide a description language specific to the aircraft.

[0007] Patent Document 3 describes a control method for multi-joint bionic dolphins in an underwater environment, and could not be applied to the automatic generation of controllers that can accommodate the special shapes of drones flying in the air or changes in the center of gravity.

[0008] Therefore, the present invention aims to provide a drone control system, a drone control device, and a drone control method that automatically generate an optimal controller according to the aircraft configuration based on a description language capable of describing the unique physical characteristics of drones, thereby achieving highly accurate and flexible flight control even for drones with special shapes or changes in center of gravity. [Means for solving the problem]

[0009] To solve the above problems, the drone control system of the present invention comprises a UADF description unit for describing the physical characteristics of a drone, a control logic generation unit that automatically generates control logic suitable for the drone based on the physical characteristic information described by the UADF description unit, and a flight control unit that performs flight control of the drone based on the control logic generated by the control logic generation unit. The UADF description unit describes the physical characteristics of the drone in UADF (Universal Aircraft Description Format) format, and the UADF format is a general-purpose aircraft description format that adds aircraft-specific joint types, including tilt rotors, propulsion units, and control surfaces, to URDF, which is a robot description format. [Effects of the Invention]

[0010] According to the present invention, it is possible to provide a drone control system, a drone control device, and a drone control method that automatically generate an optimal controller according to the aircraft configuration based on a description language capable of describing the physical characteristics specific to drones, thereby achieving highly accurate and flexible flight control even for drones with special shapes or changes in center of gravity. [Brief explanation of the drawing]

[0011] [Figure 1] This is a block diagram showing an example of the overall configuration of the drone system according to this embodiment. [Figure 2] This is a block diagram showing an example of the overall configuration of the drone control system according to this embodiment. [Figure 3] This figure shows an example of the detailed configuration of the UADF description unit according to this embodiment. [Figure 4] This flowchart shows an example of the processing flow of the control logic generation unit according to this embodiment. [Figure 5] This is a conceptual diagram showing an example of the process for analyzing aircraft characteristics and constructing dynamic models according to this embodiment. [Figure 6] This is a comparison diagram showing an example of the description content in UADF format and URDF format according to this embodiment. [Figure 7] This is a data flow diagram showing an example of the overall data flow of the system according to this embodiment. [Figure 8] This is a comparative conceptual diagram showing an example of the control performance between the model-based controller according to this embodiment and a conventional fixed-parameter controller. [Figure 9] (a) This figure shows the control tracking performance of a conventional flight controller. (b) This figure shows the control tracking performance of a flight controller according to this embodiment. [Modes for carrying out the invention]

[0012] The following description of this embodiment will be made with reference to the drawings. It should be noted that the present invention is not limited to the embodiments shown below, and can be modified, added, altered, or deleted to the extent that a person skilled in the art can conceive of it. Any embodiment that achieves the function and effect of the present invention is included within the scope of the present invention.

[0013] Figure 1 shows the overall configuration of the drone system 1. The drone system 1 includes a flight controller 7, a motor control device 2, a motor / propeller 3, a communication device 4, a camera etc. 5, a power management device 6, a battery 8, a GPS 9, an attitude sensor 11, a magnetic sensor 12, and a barometer 13. The flight controller 7 outputs control commands to the motor control device 2 based on information acquired from each sensor, and drives the motor / propeller 3 to perform flight control.

[0014] Figure 2 is a diagram showing the overall configuration of a drone control system 10 according to the present embodiment. The drone control system 10 according to the present embodiment includes: a UADF description unit 30 for describing the physical characteristics of a drone 80; a control logic generation unit 40 that automatically generates control logic 92 suitable for the drone 80 based on physical characteristic information 91 described by the UADF description unit 30; and a flight control unit 50 that executes flight control of the drone 80 based on the control logic 92 generated by the control logic generation unit 40. The UADF description unit 30 describes the physical characteristics of the drone 80 in a UADF (Universal Aircraft Description Format) 90 according to the present embodiment. The UADF format 90 is a general-purpose aircraft description format characterized in that it adds aircraft-specific joint types including tilt rotors 81, propulsion units 98, and control surfaces 82 to URDF 97, which is a robot description format. It should be noted that the UADF format 90 is a newly created description format developed after intensive studies by the inventor to enable comprehensive description of actuator characteristics and aerodynamic characteristics specific to aircraft, which were difficult to represent in the existing URDF format.

[0015] In addition, it is preferable that the control logic generation unit 40 generates control logic 92 specific to the airframe configuration 96 of the drone 80 based on physical characteristic information 91 including joint type information 94 described in the UADF format 90.

[0016] In addition, it is preferable that the system further comprises an airframe characteristic analysis unit 60 that constructs a dynamic model 93 of the drone 80 from the physical characteristic information 91 described in the UADF format 90, and the control logic generation unit 40 generates a model-based controller 71 based on the dynamic model 93. Figure 5 is a diagram showing the concept of a process in which the airframe characteristic analysis unit 60 constructs the dynamic model 93 from the physical characteristic information 91 and provides it to the control logic generation unit 40.

[0017] Further, said UADF description unit 30, said control logic generation unit 40 and said flight control unit 50 are implemented as an integrated model-based flight controller 70, and it is preferable that said model-based flight controller 70 automatically provides a controller optimized for each airframe configuration 96 for drones 80 having a plurality of different airframe configurations 96 described in the UADF format 90.

[0018] FIG. 3 is a diagram showing an example of the detailed configuration of the UADF description unit 30 and the content described by the UADF format 90. The UADF description unit 30 creates physical property information 91 and joint type information 94 using the UADF format 90, which is an extension of URDF 97. The joint type information 94 includes aircraft-specific joint types such as tilt rotors 81, propulsion units 98, and control surfaces 82.

[0019] FIG. 4 is a flowchart showing the processing procedure for automatic control logic generation in the control logic generation unit 40. The control logic generation unit 40 receives the physical property information 91 and joint type information 94 input from the UADF description unit 30, acquires the dynamic model 93 constructed by the airframe characteristic analysis unit 60, and synthesizes control logic 92 including the model-based controller 71.

[0020] FIG. 6 is a diagram comparing and showing the description contents of URDF 97 and UADF format 90. The UADF format 90 is extended to a general-purpose aircraft description format based on URDF 97, and can clearly express aircraft-specific elements as joint type information 94.

[0021] FIG. 7 is a data flow diagram showing a series of data flows from UADF description to control execution, and information exchange between each component. The UADF description unit 30 outputs physical property information 91 and joint type information 94, the control logic generation unit 40 and the airframe characteristic analysis unit 60 construct the dynamic model 93 to generate the control logic 92, and the flight control unit 50 generates a control signal 95 based on the control logic 92 and outputs it to the drone 80.

[0022] Figure 8 is a graph showing a conceptual comparison of control variables between a conventional fixed-parameter controller and the model-based controller 71 of the present invention. It can be seen that the control according to this embodiment approaches the set value faster than the conventional technology. By using the model-based controller 71 of the present invention for a drone 80 with a special shape that differs from the usual aircraft configuration 96, high control performance is achieved by automatically generating the control logic 92 based on the physical characteristics of the aircraft.

[0023] Furthermore, the drone control device 20 according to this embodiment includes a description means 110 that describes the physical characteristics of the drone 80 in UADF (Universal Aircraft Description Format) format 90, a control logic generation means 120 that automatically generates control logic 92 suitable for the drone 80 based on the physical characteristic information 91 described by the description means 110, and a control signal output means 130 that outputs a control signal 95 to the drone 80 based on the control logic 92 generated by the control logic generation means 120. Preferably, the UADF format 90 is a general-purpose aircraft description format that adds aircraft-specific joint types, including a tilt rotor 81, a propulsion unit 98, and a control surface 82, to the robot description format URDF 97.

[0024] Furthermore, the drone control method 100 according to this embodiment includes a description step 140 in which the physical characteristics of the drone 80 are described in UADF (Universal Aircraft Description Format) format 90, a control logic generation step 150 in which control logic 92 suitable for the drone 80 is automatically generated based on the physical characteristic information 91 described in the description step 140, and a control execution step 160 in which flight control of the drone 80 is performed based on the control logic 92 generated in the control logic generation step 150. Preferably, the UADF format 90 is a general-purpose aircraft description format that adds aircraft-specific joint types, including a tilt rotor 81, a propulsion unit 98, and a control surface 82, to the robot description format URDF 97.

[0025] The drone control system 10 of the present invention comprises a UADF description unit 30, a control logic generation unit 40, and a flight control unit 50. The drone control system 10 automatically generates a control system based on the physical characteristics of the drone 80.

[0026] The drone control system 10 automatically provides a controller optimized for each of the multiple different airframe configurations 96 of the drone 80, which are described in UADF format 90. The drone control system 10 enables highly accurate and flexible flight control even for drones 80 with special shapes or different centers of gravity, which were difficult for conventional flight controllers to handle.

[0027] The UADF description unit 30 is configured to describe the physical characteristics of the drone 80. The UADF description unit 30 describes the physical characteristics of the drone 80 in UADF format 90. The UADF description unit 30 describes in detail the aircraft's center of gravity, mass distribution, propeller arrangement, joint type, etc. The physical characteristics information 91 described by the UADF description unit 30 is provided to the control logic generation unit 40 and the aircraft characteristics analysis unit 60.

[0028] The control logic generation unit 40 automatically generates control logic 92 suitable for the drone 80 based on the physical characteristic information 91 described by the UADF description unit 30. The control logic generation unit 40 generates control logic 92 specific to the aircraft configuration 96 of the drone 80 based on the physical characteristic information 91, including joint type information 94 described in UADF format 90. The control logic generation unit 40 generates a model-based controller 71 based on the dynamics model 93. The control logic 92 generated by the control logic generation unit 40 is provided to the flight control unit 50.

[0029] The flight control unit 50 performs flight control of the drone 80 based on the control logic 92 generated by the control logic generation unit 40. The flight control unit 50 generates control commands for each actuator using the generated control logic 92 and outputs them as control signals 95. The flight control unit 50 is responsible for real-time flight control execution and controls the actual flight operation.

[0030] The aircraft characteristics analysis unit 60 constructs a dynamics model 93 of the drone 80 from physical characteristic information 91 described in UADF format 90. The aircraft characteristics analysis unit 60 converts the physical characteristics of the aircraft into a mathematical model and provides the design basis for the model-based controller 71 in the control logic generation unit 40. The aircraft characteristics analysis unit 60 performs a comprehensive dynamics analysis considering the center of gravity, moment of inertia, aerodynamic characteristics, etc. The dynamics model 93 constructed by the aircraft characteristics analysis unit 60 is provided to the control logic generation unit 40.

[0031] The model-based flight controller 70 has an integrated configuration comprising a UADF description unit 30, a control logic generation unit 40, and a flight control unit 50. The model-based flight controller 70 automatically provides a controller optimized for each of the multiple different aircraft configurations 96 described in UADF format 90 for a drone 80. The model-based flight controller 70 is an adaptive control system that differs from conventional fixed-parameter controllers.

[0032] The model-based controller 71 is a controller generated by the control logic generation unit 40 based on the dynamics model 93. The model-based controller 71 achieves high-precision control that takes into account the physical characteristics of the aircraft and can handle special shapes and changes in the center of gravity. Unlike conventional products, the model-based controller 71 is realized by an advanced control method that includes a detailed aircraft model internally.

[0033] The drone 80 is an unmanned aerial vehicle that is the subject of control according to the present invention. The drone 80 has a variety of airframe configurations 96, including special shapes, changes in the center of gravity, and asymmetrical propeller arrangements. The drone 80 is not limited to the typical airframe shapes supported by conventional products, but supports various shapes and configurations for industrial and research applications. The physical characteristics of the drone 80 are described in UADF format 90, and individually optimized control logic 92 is applied.

[0034] The Tiltrotor 81 is an aircraft-specific joint type defined in UADF Form 90. The Tiltrotor 81 represents a variable thrust vectoring mechanism capable of changing the rotor's direction. The Tiltrotor 81 is an advanced propulsion system capable of both vertical takeoff and landing and horizontal flight.

[0035] The control surface 82 is one of the aircraft-specific joint types defined in UADF format 90. The control surface 82 represents a control surface that changes the aerodynamic characteristics by controlling the rudder angle of the wing surface. The control surface 82 is a mechanism that applies conventional aircraft control surfaces such as elevators, ailerons, and rudders to the drone 80.

[0036] UADF format 90 is a general-purpose aircraft description format for describing the physical properties of a drone 80. UADF format 90 is an aircraft-specific joint type, including a tilt rotor 81, propulsion unit 98, and control surface 82, added to URDF format 97, which is a robot description format. UADF format 90 allows for detailed description of the aircraft's center of gravity, mass distribution, propeller arrangement, aerodynamic characteristics, etc., in a structured format. UADF format 90 is a core technical element of the present invention and enables support for diverse aircraft configurations 96.

[0037] Physical properties information 91 is information about the physical characteristics of the drone 80 described in UADF format 90. Physical properties information 91 is comprehensive aircraft information including the center of gravity, mass distribution, moment of inertia, propeller arrangement, joint type, etc. Physical properties information 91 is used as foundational data for control logic generation and dynamics model construction. Physical properties information 91 provides a detailed representation of the individual characteristics of the aircraft, enabling the generation of aircraft-specific control logic 92.

[0038] The control logic 92 is a control algorithm suitable for the drone 80, automatically generated by the control logic generation unit 40. The control logic 92 is a control law specific to the aircraft configuration 96 and can handle specially shaped aircraft that are difficult to handle with conventional general-purpose controllers. The control logic 92 is realized by a model-based control method based on the dynamics model 93. The control logic 92 serves as the basis for generating the actual control signals 95 in the flight control unit 50.

[0039] The dynamics model 93 is a mathematical representation of the physical behavior of the drone 80, constructed by the aircraft characteristics analysis unit 60. The dynamics model 93 is formulated as the equations of motion for the aircraft based on the physical characteristics information 91 described in UADF format 90. The dynamics model 93 provides the mathematical representation that forms the basis for the design of the model-based controller 71. The dynamics model 93 enables precise control that reflects the individual characteristics of each aircraft.

[0040] The joint type information 94 is information representing the type of movable part of the aircraft as described in UADF format 90. The joint type information 94 includes aircraft-specific joint types, including the tilt rotor 81, propulsion unit 98, and control surface 82. The joint type information 94 is directly used in the control logic generation unit 40 to generate the control logic 92 specific to the aircraft configuration 96.

[0041] The control signal 95 is a specific control command for each actuator of the drone 80, output from the flight control unit 50. The control signal 95 is a command value such as motor rotation speed and rudder angle, calculated based on the generated control logic 92. The control signal 95 is output as an electrical signal to control the flight state of the drone 80 in real time.

[0042] The aircraft configuration 96 is a concept representing the physical structure and arrangement of the drone 80. The aircraft configuration 96 includes special shapes, center of gravity position, propeller arrangement, joint type combinations, etc. The aircraft configuration 96 is described in UADF format 90, and the control logic generation unit 40 generates control logic 92 specific to the aircraft configuration 96.

[0043] URDF97 stands for Unified Robot Description Format and is an existing standard format for describing the physical structure of a robot. URDF97 is the underlying description language for UADF format 90, and this invention extends it for drones 80. URDF97 provides functions for describing basic robot components such as links, joints, and sensors.

[0044] The propulsion unit 98 is one of the aircraft-specific joint types defined in UADF format 90. The propulsion unit 98 represents a thrust generation mechanism using propellers or fans. The propulsion unit 98 is an actuator type capable of controlling the magnitude and direction of thrust.

[0045] The drone control device 20 is a device that implements the drone control system 10. The drone control device 20 includes a description means 110, a control logic generation means 120, and a control signal output means 130. The drone control device 20 has a hardware-software configuration that integrates a physical characteristics description function using UADF format 90 and an automatic control logic generation function 92. The drone control device 20 is provided in a form that can be mounted on various drones 80 as a standalone device.

[0046] The description means 110 is a means in the drone control device 20 for describing the physical characteristics of the drone 80 in UADF format 90. The description means 110 is an implementation of the UADF description unit 30 in the device and provides an input and description function for aircraft information. The description means 110 supplies physical characteristic information 91 to the control logic generation means 120.

[0047] The control logic generation means 120 is a means for automatically generating control logic 92 in the drone control device 20 based on physical characteristic information 91 described by the description means 110. The control logic generation means 120 is an implementation form of the control logic generation unit 40 in the device and provides a control algorithm generation function specific to the aircraft. The control logic generation means 120 provides the generated control logic 92 to the control signal output means 130.

[0048] The control signal output means 130 is a means in the drone control device 20 that outputs a control signal 95 to the drone 80 based on the control logic 92 generated by the control logic generation means 120. The control signal output means 130 is an implementation in the flight control unit 50 and is responsible for the generation and output functions of the actual control signal 95. The control signal output means 130 generates and outputs specific control commands for each actuator of the drone 80.

[0049] The drone control method 100 is a series of processing steps for realizing control of a drone 80 based on the UADF format 90. The drone control method 100 is executed in the following order: description step 140, control logic generation step 150, and control execution step 160. The drone control method 100 gradually realizes optimal control based on the physical characteristics of the aircraft.

[0050] Description step 140 is a process of describing the physical properties of the drone 80 in UADF format 90. In description step 140, the aircraft's center of gravity, mass distribution, propeller arrangement, joint type, etc. are described in detail to generate physical property information 91.

[0051] The control logic generation step 150 is a step in which control logic 92 suitable for the drone 80 is automatically generated based on the physical characteristic information 91 described in the description step 140. In the control logic generation step 150, a dynamics model 93 is constructed and control logic 92 including a model-based controller 71 is generated.

[0052] The control execution step 160 is a step in which flight control of the drone 80 is performed based on the control logic 92 generated in the control logic generation step 150. In the control execution step 160, control signals 95 are generated based on the control logic 92 and output to each actuator of the drone 80. [Examples]

[0053] The present invention will be described in more detail below with reference to examples, but the present invention is not limited to these examples.

[0054] The drone system 1 shown in Figure 1 will be described below. The drone system 1 comprises a flight controller 7, a motor control device 2, a motor / propeller 3, a communication device 4, a camera etc. 5, a power management device 6, a battery 8, a GPS 9, an attitude sensor 11, a magnetic sensor 12, and a barometer 13, all of which are mounted inside the drone 80. The flight controller 7 includes a processor and memory device, and acquires position from the GPS 9, attitude and acceleration from the attitude sensor 11, bearing from the magnetic sensor 12, and altitude from the barometer 13, and outputs control commands to the motor / propeller 3 via the motor control device 2 based on these. The communication device 4 is responsible for communication with the control device 14, receives control signals from the control device 14, and transmits them to the flight controller 7. The communication device 4 is, for example, an RC receiver or a WiFi communication module, and receives control signals from the control device 14 (for example, an RC transmitter or ground station). The communication device 4 may also have a function to transmit aircraft status and video data to the control device 14. The power management device 6 distributes power from the battery 8 to each component.

[0055] The overall configuration of the drone control system 10 shown in Figure 2 is described below. The drone control system 10 comprises a UADF description unit 30, a control logic generation unit 40, a flight control unit 50, and an aircraft characteristic analysis unit 60, which are integrated as a model-based flight controller 70. The UADF description unit 30 generates physical characteristic information 91 and joint type information 94 in UADF format 90. The aircraft characteristic analysis unit 60 constructs a dynamics model 93 from the physical characteristic information 91, and the control logic generation unit 40 synthesizes control logic 92 (including the model-based controller 71) based on the dynamics model 93 and joint type information 94, and outputs it to the flight control unit 50. The flight control unit 50 provides control signals 95 based on the control logic 92 to the drone 80.

[0056] The details of the UADF description unit 30 shown in Figure 3 will be explained. The UADF description unit 30 generates physical characteristic information 91, including the aircraft configuration 96, center of gravity, mass distribution, inertia tensor, aerodynamic characteristic parameters, actuator limitations, etc., in UADF format 90 based on URDF 97. The joint type information 94 includes aircraft-specific tilt rotors 81, propulsion units 98, and control surfaces 82, and attributes such as rotation axis direction, operating range, response bandwidth, effective control surface area, and control surface angle limitations are described for each. The generated physical characteristic information 91 and joint type information 94 are provided to the aircraft characteristic analysis unit 60 and the control logic generation unit 40.

[0057] The processing flow of the control logic generation unit 40 shown in Figure 4 will be explained. The processing is carried out in the following order: ST1: reading physical characteristic information 91, ST2: interpreting joint type information 94, ST3: obtaining a dynamics model 93 from the aircraft characteristic analysis unit 60, ST4: controller design based on the dynamics model 93 (model-based controller 71, e.g., LQR, MPC, gain scheduling, observation instrument design, etc.), ST5: generating control logic 92 specific to the aircraft configuration 96, and ST6: outputting the generated control logic 92 to the flight control unit 50. The inputs and outputs of each stage are organized as shown in Figure 4, and exception handling in case of failure and requests for re-identification of physical parameters are also handled within the flow. Note that, as processing before ST4, the control logic generation unit 40 may estimate the aerodynamic characteristics of the propeller (which may include attributes such as thrust / aerodynamic coefficient, operating limit, lag / hysteresis characteristics, etc.) and the electrical characteristics of the motor, and the model-based controller 71 may perform control considering these characteristics and the dynamics model obtained from the UADF.

[0058] The concepts of the aircraft characteristics analysis unit 60 and dynamics model construction shown in Figure 5 are described in detail. The aircraft characteristics analysis unit 60 constructs a dynamics model 93 by combining the rigid body 6-degree-of-freedom equations of motion and actuator equations, using the mass characteristics (center of gravity position, mass, inertia), aerodynamic and thrust models, and joint geometry given in UADF format 90. Specifically, it is preferable to include translational and rotational motion equations around the aircraft's center of gravity, thrust vector transformations dependent on the angle of the tilt rotor 81, saturation and lag models of the propulsion unit 98, and small disturbance equations including fixed wings and control surfaces. The model may be linearized for each operational point (hovering, forward flight, transition) and used for controller design.

[0059] Figure 6 describes the comparison between URDF97 and UADF format 90. URDF97 provides descriptions of links, basic joints, sensors, and mass characteristics, while UADF format 90 adds aircraft-specific extensions, allowing for the direct representation of additional joint types for the tilt rotor 81, propulsion unit 98, and control surfaces 82. This enables integrated modeling of the drone's specific drive system and aerodynamics, allowing the control logic generation unit 40 to automate aircraft-specific control synthesis.

[0060] The data flow of the entire system shown in Figure 7 will be explained. In response to the description step 140, the UADF description unit 30 outputs physical characteristic information 91 and joint type information 94. In response to the control logic generation step 150, the aircraft characteristic analysis unit 60 constructs a dynamic model 93, and the control logic generation unit 40 generates control logic 92. In response to the control execution step 160, the flight control unit 50 generates a control signal 95 and distributes it to each actuator of the drone 80. This entire process is contained within the model-based flight controller 70 and is linked in real time with the external control device 14 and sensor group (Figure 1).

[0061] The control performance comparison with the conventional technology shown in Figure 8 is described below. The graph in Figure 8 is a conceptual diagram comparing the response of a conventional fixed-parameter controller and the model-based controller 71 of this embodiment to a drone 80 with a disrupted propeller arrangement and mass distribution symmetry, under the same conditions. The solid line (this embodiment) shows a fast rise time, and the target value is achieved early. On the other hand, the dashed line (conventional technology) shows a large modeling error, so the control gain must be reduced to prevent vibration of the aircraft, resulting in a longer settling time and lower stability performance compared to this embodiment. The control logic 92 of the present invention is superior in both responsiveness and stability due to the accuracy of the dynamics model 93 based on the UADF format 90 and the control assignment specific to the aircraft configuration 96.

[0062] Figure 9 shows experimental results comparing the control tracking performance when a weight is attached to the drone 80 to intentionally shift its center of gravity, using a conventional fixed-parameter flight controller (Figure 9(a)) and the flight controller according to this embodiment (Figure 9(b)). The vertical axis represents the pitch angle of the aircraft from the horizontal plane (in degrees), and the horizontal axis represents time. In Figures 9(a) and 9(b), the dashed line shows the target value of the pitch angle of the drone 80, and the solid line shows the actual pitch angle estimated by the Kalman filter. In Figure 9(a), which is the control result using the fixed-parameter flight controller, it can be seen that an offset occurs in the vertical direction. In contrast, in Figure 9(b), which is the control result using the model-based controller 71 of this embodiment, the offset is smaller, and the control tracking performance in the time direction is also improved compared to the case in Figure 9(a).

[0063] An example of a specific UADF description in this embodiment will be explained. The aircraft configuration 96 has two tilt rotors 81 positioned forward and two propulsion units 98 positioned aft, and is equipped with a control surface 82 (rudder and elevator) at the tail. The UADF description unit 30 describes the rotation axis vector, tilt range (e.g., ±90 degrees), angular velocity limit, and upper limit of the control surface angle and operating speed of each rotor. Furthermore, it generates physical property information 91 including the coordinates of the center of gravity of the entire aircraft, mass, inertia tensor, thrust-rotation speed model, and linear approximation parameters of the control surface-lift coefficient.

[0064] An example of a model-based controller 71 synthesized by the control logic generation unit 40 is described below. At the operating point near hovering, the dynamics model 93 is linearized and the feedback gain of the attitude / position loop is calculated using LQR, and the tilt angle, rudder angle, and thrust are assigned using control assignment based on the least squares method. An extended state observer is used for disturbance estimation to ensure robustness against wind disturbances. Actuator saturation and rate limits are considered using iterative assignment, and redundant control is achieved using only available actuators.

[0065] In another modified embodiment, the control of the multicopter does not linearize the equations of motion, regardless of whether it is close to the equilibrium point, and the equations of motion are used directly in the mixer after the normal PID processing. In the mixer, the problem of minimizing the sum of squared thrusts while satisfying the equations of motion and actuator constraints is reduced to a quadratic programming problem and solved.

[0066] An example of the flight control unit 50 implementation is described below. The flight control unit 50 has a high-speed loop with a control period of 1 to 5 ms and operates the attitude estimator (IMU fusion), altitude estimator (barometer 13 and acceleration compensation), and position estimator (GPS 9 and inertial navigation). Commands calculated by the control logic 92 are output to the motor control device 2 as digital protocols such as PWM or DShot, and to the control surface 82 as servo signals, using a function equivalent to the control signal output means 130. The flight log is stored in the memory device.

[0067] The configuration of the drone control device 20 is described below. The description means 110 edits and verifies the UADF format 90 via a design tool and saves the generated physical characteristic information 91 and joint type information 94. The control logic generation means 120 automatically synthesizes the control logic 92 using the dynamics model 93 obtained by the aircraft characteristic analysis unit 60 as input and includes a controller as needed. The control signal output means 130 outputs a control signal 95 via a bus and, in the event of failsafe, executes safety operations such as RTL (Return To Launch), landing, emergency stop, and parachute activation.

[0068] The procedure for drone control method 100 is specified. In the description step 140, the UADF description unit 30 inputs the aircraft configuration 96, center of gravity, inertia, and actuator attributes during the maintenance and manufacturing stage. In the control logic generation step 150, the aircraft characteristic analysis unit 60 constructs a dynamics model 93, and the control logic generation unit 40 synthesizes a model-based controller 71. In the control execution step 160, the flight control unit 50 performs real-time control based on the control logic 92 and provides feedback to the description step 140 as needed based on data obtained from the flight log.

[0069] As described above, according to this embodiment, the UADF description unit 30 generates physical characteristic information 91 including aircraft-specific joint type information 94 in the UADF format 90 which is an extension of URDF 97, the aircraft characteristic analysis unit 60 constructs a dynamics model 93, the control logic generation unit 40 automatically generates control logic 92 with a model-based controller 71 at its core, and the flight control unit 50 outputs it as a control signal 95, thereby enabling highly accurate and flexible flight control even for drones 80 having multiple and different aircraft configurations 96. The components of this embodiment can be implemented in either software or hardware, and the distributed arrangement or integration of each block can be selected as appropriate.

[0070] <Aspect 1> A drone control system 10, A UADF description unit 30 for describing the physical characteristics of the drone 80, A control logic generation unit 40 automatically generates control logic 92 suitable for the drone 80 based on physical characteristic information 91 described by the UADF description unit 30, The system includes a flight control unit 50 that performs flight control of the drone 80 based on the control logic 92 generated by the control logic generation unit 40, The UADF description unit 30 describes the physical characteristics of the drone 80 in UADF (Universal Aircraft Description Format) format 90. The drone control system 10 is characterized in that the UADF format 90 is a general-purpose aircraft description format that adds aircraft-specific joint types 94, including a tilt rotor 81, a propulsion unit 98, and a control surface 82, to the URDF format 97, which is a robot description format.

[0071] According to this embodiment, describing the physical characteristics of a drone in UADF format enables more flexible and comprehensive aircraft description than conventional methods. This improves the accuracy of control logic generation and allows for optimal flight control even for drones with different aircraft configurations. Furthermore, the extensibility of the UADF format makes it easy to support new drone models in the future. In addition, its high compatibility with model-based development is expected to significantly improve development efficiency from design to operation.

[0072] <Aspect 2> The drone control system 10 described in Embodiment 1, The drone control system 10 is characterized in that the control logic generation unit 40 generates control logic 92 specific to the aircraft configuration 96 of the drone 80 based on physical characteristic information 91 including joint type information 94 described in UADF format 90.

[0073] According to this embodiment, by utilizing detailed joint type information obtained in UADF format, it becomes possible to generate highly customizable control logic tailored to the individual configuration of the drone. This enables stable flight control for drones with various configurations and characteristics, increasing design flexibility and reducing the need for redesign. By providing logic optimized for the aircraft configuration, it becomes possible to improve control performance and enhance reliability.

[0074] <Aspect 3> The drone control system 10 described in Embodiment 1, The aircraft further includes an aircraft characteristics analysis unit 60 that constructs a dynamics model 93 of the drone 80 from physical characteristics information 91 described in UADF format 90, The drone control system 10 is characterized in that the control logic generation unit 40 generates a model-based controller 71 based on a dynamics model 93.

[0075] According to this embodiment, by constructing a dynamics model from physical property information and generating a model-based controller based on it, more accurate control with less response delay becomes possible. Because dynamic and optimal control can be performed in response to environmental changes and aircraft behavior, flight stability and responsiveness are greatly improved. It can demonstrate particularly excellent performance in applications requiring highly difficult flight missions or complex operations.

[0076] <Aspect 4> The drone control system 10 described in embodiment 2, The aircraft further includes an aircraft characteristics analysis unit 60 that constructs a dynamics model 93 of the drone 80 from physical characteristics information 91 described in UADF format 90, The drone control system 10 is characterized in that the control logic generation unit 40 generates a model-based controller 71 based on a dynamics model 93.

[0077] According to this embodiment, by constructing a dynamics model specific to the aircraft configuration based on physical characteristics including joint type information, the accuracy of the model-based controller is improved and the control logic is optimized. As a result, it is possible to flexibly respond to differences in aircraft characteristics and ensure consistent control quality even with drones of different configurations. This also enhances the reliability of the entire system and contributes to stable control in the event of anomalies.

[0078] <Aspect 5> The drone control system 10 described in Embodiment 1, The UADF description unit 30, the control logic generation unit 40, and the flight control unit 50 are implemented as an integrated model-based flight controller 70. A drone control system 10 is characterized in that the model-based flight controller 70 automatically provides a controller optimized for each aircraft configuration to a drone 80 having multiple different aircraft configurations 96 described in UADF format 90.

[0079] According to this embodiment, it is possible to automatically generate and apply controllers according to differences in aircraft configuration through a model-based flight controller that integrates all the functions necessary for control. This enables rapid adaptation even when new drones are introduced, and significantly increases the system's scalability. Furthermore, it offers advantages such as reduced development and operational costs and ensured consistency with simulations.

[0080] <Aspect 6> A drone control device 20, A description means 110 for describing the physical characteristics of the drone 80 in UADF (Universal Aircraft Description Format) format 90, A control logic generation means 120 automatically generates a control logic 92 suitable for the drone 80 based on the physical characteristic information 91 described by the description means 110, The system includes a control signal output means 130 that outputs a control signal 95 to the drone 80 based on the control logic 92 generated by the control logic generation means 120, The drone control device 20 is characterized in that the UADF format 90 is a general-purpose aircraft description format that adds an aircraft-specific joint type 94, including a tilt rotor 81, a propulsion unit 98, and a control surface 82, to the URDF format 97, which is a robot description format.

[0081] According to this embodiment, the drone control device provided as hardware can consistently perform tasks from describing aircraft characteristics and generating control logic to outputting control signals. This promotes automation of control, increases the degree of freedom in drone design and operation, and reduces the effort required for initial setup and reconfiguration. The adoption of the UADF format also provides expandability that can flexibly accommodate various aircraft configurations.

[0082] <Aspect 7> A drone control method 100, The description process 140 describes the physical properties of the drone 80 in UADF (Universal Aircraft Description Format) format 90, A control logic generation step 150 automatically generates control logic 92 suitable for the drone 80 based on the physical characteristic information 91 described in the description step 140, The control execution step 160 includes a control logic generation step 150 which performs flight control of the drone 80 based on the control logic 92 generated in the control logic generation step 150, The drone control method 100 is characterized in that the UADF format 90 is a general-purpose aircraft description format that adds an aircraft-specific joint type 94, including a tilt rotor 81, a propulsion unit 98, and a control surface 82, to the URDF format 97, which is a robot description format.

[0083] According to this embodiment, by defining a series of processes as steps using a software method, the development and operation of drone control systems are systematized and standardized. By utilizing the UADF format, it becomes possible to automatically generate optimal control logic based on physical characteristics, reducing complex configuration work and improving development efficiency and reusability. [Explanation of Symbols]

[0084] 1. Drone System 2. Motor control device 3 Motors / Propellers 4. Communication equipment 5. Cameras, etc. 6 Power management device 7 Flight Controllers 8 batteries 9 GPS 10 Drone control systems 11. Attitude sensor 12 Magnetic Sensors 13 Barometer 14. Control devices 20 Drone control devices 30 UADF Description Section 40 Control Logic Generation Unit 50 Flight Control Unit 60 Aircraft Characteristics Analysis Department 70 Model-Based Flight Controllers 71 Model-Based Controllers 80 Drones 81 Tilt Rotor 82 Steering surface 90 UADF format 91 Physical property information 92 Control Logic 93 Dynamics Models 94 Joint Type Information 95 Control signals 96. Aircraft Configuration 97 URDF 98 Propulsion Unit 100 Drone Control Methods 110 Description means 120 Control logic generation means 130 Control signal output means 140 Description process 150 Control logic generation process 160 Control execution process

Claims

1. A drone control system, A UADF description section for describing the physical characteristics of the drone, A control logic generation unit that automatically generates control logic suitable for the drone based on the physical characteristics information described by the UADF description unit, The system includes a flight control unit that performs flight control of the drone based on the control logic generated by the control logic generation unit, The UADF description unit describes the physical characteristics of the drone in UADF (Universal Aircraft Description Format) format. The aforementioned UADF format is a general-purpose aircraft description format that adds aircraft-specific joint types, including tilt rotors, propulsion units, and control surfaces, to the URDF format, which is a robot description format, and is characterized as a drone control system.

2. A drone control system according to claim 1, The drone control system is characterized in that the control logic generation unit generates control logic specific to the drone's airframe configuration based on physical characteristic information including joint type information described in UADF format.

3. A drone control system according to claim 1, The system further includes an aircraft characteristics analysis unit that constructs a dynamic model of the drone from the physical characteristics information described in the UADF format, The drone control system is characterized in that the control logic generation unit generates a model-based controller based on the dynamics model.

4. A drone control system according to claim 2, The system further includes an aircraft characteristics analysis unit that constructs a dynamic model of the drone from the physical characteristics information described in the UADF format, The drone control system is characterized in that the control logic generation unit generates a model-based controller based on the dynamics model.

5. A drone control system according to claim 1, The UADF description unit, the control logic generation unit, and the flight control unit are implemented as an integrated model-based flight controller. The aforementioned model-based flight controller is a drone control system characterized by automatically providing a controller optimized for each aircraft configuration for drones having multiple different aircraft configurations described in AUDF format.

6. A drone control device, A description means for describing the physical characteristics of a drone in UADF (Universal Aircraft Description Format), A control logic generation means that automatically generates a control logic suitable for the drone based on the physical characteristic information described by the description means, The system includes a control signal output means that outputs a control signal to the drone based on the control logic generated by the control logic generation means, The aforementioned UADF format is a general-purpose aircraft description format that adds aircraft-specific joint types, including tilt rotors, propulsion units, and control surfaces, to the URDF format, which is a robot description format, and is a drone control device.

7. A drone control method, The process involves describing the physical characteristics of the drone using the UADF (Universal Aircraft Description Format) format, and A control logic generation step that automatically generates a control logic suitable for the drone based on the physical characteristic information described in the description step, The control execution step includes executing flight control of the drone based on the control logic generated in the control logic generation step, The aforementioned UADF format is a general-purpose aircraft description format that adds aircraft-specific joint types, including tilt rotors, propulsion units, and control surfaces, to the URDF format, which is a robot description format, and is characterized by being a drone control method.

Images