Tilt-rotor aircraft and its attitude control method

Abstract

The present invention provides a tilt-rotor aircraft and an attitude control method thereof. The tilt-rotor aircraft includes a hovering capsule (10), a flight control module (30), a remote control module (40), and two sets of tilt-rotor combined structures (20); the tilt-rotor combined structure includes a motor (21) and a servo (22) tiltably mounted on the hovering capsule and connected to the motor; a propeller is mounted on the motor; the flight control module is mounted on the hovering capsule and electrically connected to the tilt-rotor combined structure; the remote control module is communicatively connected to the flight control module. The tilt-rotor aircraft and the attitude control method thereof of the present invention creatively adopt the double-rotor technology with variable tilt angles, realize the movement and turning of multiple degrees of freedom with the least number of rotors, and adopt a hovering capsule to balance the reliability problems brought about by adopting two sets of tilt-rotor combined structures, avoiding the risk of crashing.

Description

Technical Field

[0001] The present invention relates to the field of unmanned aerial vehicles, and particularly to a tilt-rotor dual-rotor aircraft and an attitude control method thereof. Background Art

[0002] For an aircraft with determined motion characteristics, the more quantities that can participate in control, the easier it is to obtain good control effects. Therefore, the more rotors an aircraft has, the higher the tolerance of the aircraft to power system failures. On the contrary, the control of the vector structure of the aircraft is more troublesome, the reliability is not high, and it is prone to failures or even crashes. At the same time, the more rotors an aircraft has, the weaker the efficiency and endurance of the aircraft, the less flexible the flight is, and the slower the speed is. Summary of the Invention

[0003] The present invention aims at the above technical problems and provides a tilt-rotor dual-rotor aircraft and an attitude control method thereof.

[0004] The technical solutions proposed by the present invention are as follows:

[0005] The present invention provides a tilt-rotor dual-rotor aircraft, including a hovering capsule, a flight control module, a remote control module, and two sets of tilt-rotor combined structures;

[0006] The tilt-rotor combined structure includes a motor, a servo motor that is tiltably installed on the hovering capsule and connected to the motor for tilting the motor; a propeller is installed on the motor;

[0007] The flight control module is installed on the hovering capsule and electrically connected to the tilt-rotor combined structure for controlling the operation of the tilt-rotor combined structure; the remote control module is communicatively connected to the flight control module for sending control instructions to the flight control module.

[0008] In the above tilt-rotor dual-rotor aircraft of the present invention, the hovering capsule is filled with helium or hydrogen.

[0009] In the above tilt-rotor dual-rotor aircraft of the present invention, the hovering capsule is spherical, and the two sets of tilt-rotor combined structures are respectively mounted on the equator of the hovering capsule and are centrosymmetric with the center of the hovering capsule as the center of symmetry.

[0010] The present invention also provides an attitude control method for the tilt-rotor dual-rotor aircraft as described above, including the following steps:

[0011] Step 1: Construct a PID neural network structure, and the number of network layers of the PID neural network structure is 3; the input layer nodes of the PID neural network structure adopt the flight attitude information and attitude stability measurement values of the tilt-rotor dual-rotor aircraft, and the output layer nodes of the PID neural network structure adopt the flight control parameters of the tilt-rotor dual-rotor aircraft;

[0012] Step 2: Collect the historical flight attitude information of the tiltrotor aircraft, as well as the corresponding historical flight control parameters and historical attitude stability measurement values, and input them into the PID neural network structure for training to obtain a preliminarily trained neural network model;

[0013] Step 3: Use the preliminarily trained neural network model to control the flight of the tiltrotor aircraft. If the deviation value between the real-time feedback value of the attitude stability measurement value of the tiltrotor aircraft exceeds the preset PID control threshold, then use the real-time feedback value of the attitude stability measurement value of the tiltrotor aircraft and the corresponding flight attitude information to perform iterative learning training on the preliminarily trained neural network model to generate corresponding flight control parameters, and use these flight control parameters to control the flight of the tiltrotor aircraft.

[0014] In the above attitude control method of the present invention, the step of using the real-time feedback value of the attitude stability measurement value of the tiltrotor aircraft and the corresponding flight attitude information to perform iterative learning training on the preliminarily trained neural network model to generate corresponding flight control parameters, and using these flight control parameters to control the flight of the tiltrotor aircraft includes:

[0015] Step 3.1: Use the flight control parameters generated by performing iterative learning training on the preliminarily trained neural network model to control the flight of the tiltrotor aircraft, thereby generating a new real-time feedback value of the attitude stability measurement value;

[0016] If the deviation value of the new real-time feedback value of the attitude stability measurement value does not exceed the preset PID control threshold, then use the flight control parameters generated by performing iterative learning training on the preliminarily trained neural network model to perform iterative update on the preliminarily trained neural network model; otherwise, withdraw the flight control parameters generated by performing iterative learning training on the preliminarily trained neural network model.

[0017] In the above attitude control method of the present invention, the flight attitude information of the tiltrotor aircraft includes geomagnetic data, gyroscope data, GPS data, barometer data, rotor angle data, rotor speed data, and flight speed data.

[0018] In the above attitude control method of the present invention, the attitude stability measurement value of the tiltrotor aircraft includes the X-axis offset value, Y-axis offset value, and Z-axis offset value of the tiltrotor aircraft.

[0019] The tilt-rotor aircraft of the present invention and its attitude control method creatively adopt the technology of double rotors with variable tilt angles, reducing the need for rotors with a fixed orientation, achieving mobile steering with multiple degrees of freedom using the minimum number of rotors, and using a hovering bladder to enable the entire tilt-rotor aircraft to have the hovering ability with a static lift of 0, balancing the reliability issues brought about by the adoption of two sets of tilt-rotor combined structures and avoiding the risk of crashing. BRIEF DESCRIPTION OF THE DRAWINGS

[0020] Figure 1 FIG. shows a schematic structural diagram of a tilt-rotor aircraft according to a preferred embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0021] In order to make the technical solutions, technical objectives, and technical effects of the present invention clearer so that those skilled in the art can understand and implement the present invention, the present invention will be further described in detail below with reference to the accompanying drawings and specific embodiments.

[0022] As Figure 1 shown, Figure 1 FIG. shows a schematic structural diagram of a tilt-rotor aircraft according to a preferred embodiment of the present invention. The tilt-rotor aircraft includes a hovering bladder 10, a flight control module 30, a remote control module 40, and two sets of tilt-rotor combined structures 20;

[0023] The tilt-rotor combined structure 20 includes a motor 21 and a servo 22 that is tiltably mounted on the hovering bladder 10 and connected to the motor 21 for tilting the motor 21; a propeller is mounted on the motor 21.

[0024] The flight control module 30 is mounted on the hovering bladder 10 and electrically connected to the tilt-rotor combined structure 20 for controlling the operation of the tilt-rotor combined structure 20; the remote control module 40 is communicatively connected to the flight control module 30 for sending control instructions to the flight control module 30.

[0025] Here, the hovering bladder 10 is filled with helium or hydrogen, and through weight balancing, the entire tilt-rotor aircraft obtains the hovering ability with a static lift of 0, balancing the reliability issues brought about by the adoption of two sets of tilt-rotor combined structures 20 and avoiding the risk of crashing. At the same time, the hovering bladder 10 is made of a soft material, which can provide flexible protection for the tilt-rotor aircraft. In addition, the present invention creatively adopts the technology of double rotors with variable tilt angles, reducing the need for rotors with a fixed orientation and achieving mobile steering with multiple degrees of freedom using the minimum number of rotors.

[0026] Furthermore, the tilt-rotor combined structure 20 is the power source for the tilt-rotor aircraft of the present invention to achieve maneuverability. The servo 22 is used to achieve a 180-degree tilt of the motor 21.

[0027] When the gear of the steering gear 22 is centered, the motor 21 is in a horizontal position and generates a vertically downward thrust; when the gear of the steering gear 22 tilts within 90 degrees in the front and rear directions, the motor 21 tilts accordingly to generate a vector thrust at a corresponding angle.

[0028] In this embodiment, the hovering capsule 10 is spherical. Two sets of tilt-rotor combined structures 20 are respectively mounted on the equator of the hovering capsule 10 and are centrosymmetric with the center of the hovering capsule 10 as the center of symmetry. They are mirror-image placed at positions on the spherical surface of the hovering capsule 10 with a 180-degree interval in the middle. In this way, by controlling the thrust of the motors 21 of the two sets of tilt-rotor combined structures 20 and the control angles of the steering gears 22 of the two sets of tilt-rotor combined structures 20, corresponding vector composite thrusts can be generated to push the flying robot to move upward, forward, backward, and rotate.

[0029] Generally, the center of gravity of the hovering capsule 10 will tend to the bottom to keep the hovering capsule 10 vertical in a certain direction. However, lateral or roll operations require a thrust difference generated by the two motors 21 to complete, which will conflict with the vertical maintenance of the hovering capsule 10. Therefore, the tilt-dual-rotor aircraft of the present invention does not support lateral or roll movement. In scenarios where lateral or roll movement needs to be supported, the tilt-dual-rotor aircraft of the present invention will support lateral movement by adding one or two sets of auxiliary rotor structures, but this will increase the weight and cost.

[0030] Specifically, the specific positions of the tilt-rotor combined structures 20 on the hovering capsule 10 can be adjusted accordingly according to factors such as actual requirements, the shape of the hovering capsule 10, and the aerodynamic structure.

[0031] The flight control module 30 is a functional module for realizing flight and attitude control in the tilt-dual-rotor aircraft of the present invention. The flight control module 30 mainly includes a main control unit, a sensing unit, a communication unit, a power supply unit, etc. These units can be centrally or dispersedly distributed in one or more pods at the bottom of the hovering capsule 10, so that the overall center of gravity of the hovering capsule 10 is relatively low to keep the vertical orientation of the hovering capsule 10 unchanged.

[0032] The main control unit outputs signals through a control algorithm to control the motors and steering gears of the two sets of tilt-rotor combined structures 20, thereby controlling the attitude and movement of the tilt-dual-rotor aircraft itself; the sensing unit is used to monitor various flight states and parameters during the flight of the tilt-dual-rotor aircraft. In addition, additional sensing units can be added to achieve other functions, such as adding a positioning module or other modules as needed to achieve the ability of autonomous flight; the communication unit is used for communication between the main control unit and the remote control module; the size and capacity of the power supply unit can be configured according to the requirements of the entire tilt-dual-rotor aircraft.

[0033] The control logic of the main control unit mainly includes basic flight control based on general PID, combined with a flight attitude parameter adjustment algorithm based on machine learning.

[0034] The tilt-rotor aircraft of the present invention first sets the parameters obtained based on the general PID control theory as the baseline, which are the original parameters of the tilt-rotor aircraft. Since the tilt-rotor aircraft of the present invention introduces a large aerodynamic configuration of a hover bladder, the tilt-rotor aircraft of the present invention has undergone significant changes in specifications such as size and load. In addition, in the actual flight environment, the influence of wind speed and direction on the tilt-rotor aircraft is greater than that on traditional rotor UAVs. Therefore, when exceeding the general PID control boundary, the main control unit will start the parameter adjustment algorithm to further optimize the parameter settings.

[0035] The tilt-rotor aircraft of the present invention achieves the purpose of quickly converging and tuning PID parameters through a machine learning algorithm, realizing the hover self-stabilization of the hover bladder and the enhancement of the wind resistance ability. However, it is not limited to directly mapping the target state of the aircraft's motion attitude to the flight control quantity and directly controlling the flight with a machine learning algorithm.

[0036] The present invention proposes an attitude control method based on the above tilt-rotor aircraft, including the following steps:

[0037] Step 1: Construct a PID neural network structure, and the number of network layers of this PID neural network structure is 3; the input layer nodes of the PID neural network structure adopt the flight attitude information and attitude stability measurement values of the tilt-rotor aircraft, and the output layer nodes of the PID neural network structure adopt the flight control parameters of the tilt-rotor aircraft;

[0038] In this step, the selection of the hidden layer nodes of the PID neural network structure is related to the input layer nodes and the output layer nodes, and the weighting coefficients of each layer of the PID neural network structure are random values.

[0039] In this step, the flight attitude information of the tilt-rotor aircraft includes but is not limited to geomagnetic data, gyroscope data, GPS data, barometer data, rotor angle data, rotor speed data, and flight speed data;

[0040] The attitude stability measurement values of the tilt-rotor aircraft include but are not limited to the X-axis offset value, Y-axis offset value, and Z-axis offset value of the tilt-rotor aircraft;

[0041] The flight control parameters of the tilt-rotor aircraft include but are not limited to the action control parameters when the tilt-rotor aircraft performs pitch and yaw attitude actions.

[0042] Step 2: Collect the historical flight attitude information of the tiltrotor aircraft, as well as the corresponding historical flight control parameters and historical attitude stability measurement values, and input them into the PID neural network structure for training to obtain a preliminarily trained neural network model;

[0043] Step 3: Use the preliminarily trained neural network model to control the flight of the tiltrotor aircraft. If the deviation value between the real-time feedback value of the attitude stability measurement value of the tiltrotor aircraft exceeds the preset PID control threshold, then use the real-time feedback value of the attitude stability measurement value of the tiltrotor aircraft and the corresponding flight attitude information to perform iterative learning training on the preliminarily trained neural network model, generate the corresponding flight control parameters, and use these flight control parameters to control the flight of the tiltrotor aircraft.

[0044] In this step, the deviation value of the real-time feedback value of the attitude stability measurement value of the tiltrotor aircraft refers to the deviation amount of the real-time feedback value of the attitude stability measurement value of the tiltrotor aircraft relative to the predicted value of the attitude stability measurement value for flight control of the tiltrotor aircraft.

[0045] The predicted value of the attitude stability measurement value for flight control of the tiltrotor aircraft refers to the attitude stability measurement value corresponding to the flight attitude information of the tiltrotor aircraft when using the preliminarily trained neural network model to control the flight of the tiltrotor aircraft.

[0046] The using the real-time feedback value of the attitude stability measurement value of the tiltrotor aircraft and the corresponding flight attitude information to perform iterative learning training on the preliminarily trained neural network model, generate the corresponding flight control parameters, and use these flight control parameters to control the flight of the tiltrotor aircraft includes:

[0047] Step 3.1: Use the flight control parameters generated by performing iterative learning training on the preliminarily trained neural network model to control the flight of the tiltrotor aircraft, thereby generating a new real-time feedback value of the attitude stability measurement value;

[0048] If the deviation value of this new real-time feedback value of the attitude stability measurement value does not exceed the preset PID control threshold, then use the flight control parameters generated by performing iterative learning training on the preliminarily trained neural network model to perform iterative update on the preliminarily trained neural network model; otherwise, withdraw the flight control parameters generated by performing iterative learning training on the preliminarily trained neural network model.

[0049] The embodiments of the present invention have been described above in conjunction with the accompanying drawings. However, the present invention is not limited to the above specific embodiments. The above specific embodiments are merely illustrative and not restrictive. Under the inspiration of the present invention, those of ordinary skill in the art can also make many forms without departing from the spirit of the present invention and the scope protected by the claims. All of these are within the protection scope of the present invention.

Claims

1. An attitude control method for a tilt-rotor aircraft, characterized in that, The tilt-rotor aircraft includes a hovering capsule (10), a flight control module (30), a remote control module (40), and two tilt-rotor combined structures (20); The tilt-rotor combined structure (20) includes a motor (21) and a servo (22) that is tiltably mounted on the hovering capsule (10) and connected to the motor (21) for tilting the motor (21); a propeller is mounted on the motor (21); The flight control module (30) is mounted on the hovering capsule (10) and electrically connected to the tilt-rotor combined structure (20) for controlling the operation of the tilt-rotor combined structure (20); the remote control module (40) is communicatively connected to the flight control module (30) for sending control commands to the flight control module (30); the attitude control method includes the following steps: Step 1: Construct a PID neural network structure with 3 network layers; the input layer nodes of the PID neural network structure use the flight attitude information and attitude stability measurement values of the tilt-rotor aircraft, and the output layer nodes of the PID neural network structure use the flight control parameters of the tilt-rotor aircraft; among them, the attitude stability measurement values of the tilt-rotor aircraft include the X-axis offset value, Y-axis offset value, and Z-axis offset value of the tilt-rotor aircraft; Step 2: Collect the historical flight attitude information, corresponding historical flight control parameters, and historical attitude stability measurement values of the tilt-rotor aircraft, and input them into the PID neural network structure for training to obtain a preliminarily trained neural network model; Step 3: Use the preliminarily trained neural network model to control the flight of the tilt-rotor aircraft. If the deviation value of the real-time feedback value of the attitude stability measurement value of the tilt-rotor aircraft exceeds the preset PID control threshold, then use the real-time feedback value of the attitude stability measurement value of the tilt-rotor aircraft and the corresponding flight attitude information to perform iterative learning training on the preliminarily trained neural network model, generate corresponding flight control parameters, and use these flight control parameters to control the flight of the tilt-rotor aircraft.

2. The attitude control method according to claim 1, wherein The hovering capsule (10) is filled with helium or hydrogen.

3. The attitude control method according to claim 1, wherein The hovering capsule (10) is spherical, and the two tilt-rotor combined structures (20) are respectively mounted on the equator of the hovering capsule (10) and are centrosymmetric with the center of the hovering capsule (10) as the center of symmetry.

4. The attitude control method according to any one of claims 1-3, characterized in that, The using the real-time feedback value of the attitude stability measurement value of the tilt-rotor aircraft and the corresponding flight attitude information to perform iterative learning training on the preliminarily trained neural network model, generate corresponding flight control parameters, and use these flight control parameters to control the flight of the tilt-rotor aircraft includes: Step 3.1: Use the flight control parameters generated by performing iterative learning training on the preliminarily trained neural network model to control the flight of the tilt-rotor aircraft, thereby generating a new real-time feedback value of the attitude stability measurement value; If the deviation value of the real-time feedback value of the new attitude stability measurement value does not exceed the preset PID control threshold, the iteratively updated neural network model that has been preliminarily trained is updated iteratively using the flight control parameters generated by the iterative learning and training of the preliminarily trained neural network model; otherwise, the flight control parameters generated by the iterative learning and training of the preliminarily trained neural network model are withdrawn.

5. The attitude control method according to any one of claims 1-3, characterized in that The flight attitude information of the tilt-rotor aircraft includes geomagnetic data, gyroscope data, GPS data, barometer data, rotor angle data, rotor speed data, and flight speed data.

Images