Small unmanned aerial vehicle modeling and control law design method and system based on flight test

Abstract

The invention provides a small unmanned aerial vehicle modeling and control law design method and system based on a test flight test, and the method comprises the steps: S1, building a small unmanned aerial vehicle test flight test platform, enabling the small unmanned aerial vehicle to fly without obstacles, S2, carrying out the configuration before the test flight test, obtaining the preliminary control capability of the small unmanned aerial vehicle, and S3, enabling the small unmanned aerial vehicle to reach a balance state based on the preliminary control capability. Obtaining a transfer function and a linear state equation of each control channel of the small unmanned aerial vehicle; and S4, based on the transfer function and the linear state equation, generating a preliminary control law and carrying out flight test, judging whether the preliminary control law is reasonable or not, and optimizing PID control parameters. Compared with a traditional dynamic model modeling method, the dynamic model is established by determining the aerodynamic parameters and the dynamic characteristics of the small unmanned aerial vehicle based on the actual flight data, the method is closer to a real physical system, the model accuracy is high, the modeling process is simplified, and the cost is reduced.

Description

Technical Field

[0001] The present invention relates to the field of aviation technology, and particularly relates to a method and system for modeling and control law design of a small unmanned aerial vehicle based on flight test experiments. Background Art

[0002] In the existing aircraft design, especially the method for establishing the dynamic model of a low-cost small unmanned aerial vehicle, it mainly comes from the design method of manned aircraft (i.e., large aircraft), and it needs to go through a series of links such as parameter measurement and calculation, aerodynamic simulation or measurement, dynamic model modeling, control law design and parameter adjustment, simulation test, hardware-in-the-loop simulation, and actual flight verification. The links are complex and time-consuming. The dynamic modeling of existing small unmanned aerial vehicles usually needs to obtain the aerodynamic parameters of the aircraft through wind tunnel tests, and then establish the corresponding six-degree-of-freedom dynamic equations to obtain the dynamic model of the small unmanned aerial vehicle. The whole process is cumbersome and time-consuming; moreover, in the simulation test and hardware-in-the-loop simulation links, it is necessary to solve the errors caused by inaccurate dynamic model modeling, and at the same time, it is necessary to continuously modify and adjust the control law parameters. The existing design and production technology of small unmanned aerial vehicles has been relatively mature, but the rapid implementation and verification of low-cost small unmanned aerial vehicles urgently need to be solved. Because different configurations of small unmanned aerial vehicles will also put forward different or new requirements for dynamic model modeling, and the control law design also requires a long time of iteration and verification. The program is complex, and the time and economic costs are too high for current small aircraft design enterprises with rapid iteration and design and production. This has inhibited the research and development and testing of low-cost small unmanned aerial vehicles. Specifically, the following main technical problems exist in the establishment of the dynamic model of small unmanned aerial vehicles involved in the existing technology:

[0003] Complex and time-consuming modeling: The dynamic model of small unmanned aerial vehicles usually adopts the design method of manned aircraft, and through complex parameter measurement, aerodynamic simulation, wind tunnel test and other steps, the dynamic model is established, and the process is cumbersome and time-consuming.

[0004] Inaccurate model: The existing modeling method relies on wind tunnel tests to obtain aerodynamic parameters, but the model accuracy is not high, and errors often occur in the simulation and hardware-in-the-loop test stages, resulting in the need to continuously correct the model.

[0005] Difficult control law design: The control law design depends on a complex iterative process. Affected by the inaccurate model, the control law parameters need to be adjusted repeatedly, increasing the development difficulty and time cost.

[0006] Diverse design requirements for new small unmanned aerial vehicles: Different configurations and rapid implementation requirements of new small unmanned aerial vehicles require a faster and more flexible method for establishing the dynamic model, and the traditional method is difficult to adapt.

[0007] High costs and complex processes: For low-cost small drones, the complex modeling and control law design processes result in excessively high time and economic costs, suppressing the need for rapid research, development, and verification.

[0008] Therefore, during the research, development, and testing processes of low-cost small drones, the complex and time-consuming processes of dynamic model modeling and control law design have become technical problems that urgently need to be solved. Summary of the Invention

[0009] To solve the technical problems that the processes of dynamic model modeling and control law design for low-cost small drones require multiple design and test links, have a long development cycle, and high development costs, the present invention proposes a method and system for small drone modeling and control law design based on flight test trials, which solves the problems of complex processes, many design and test links, long development cycles, and high development costs for dynamic model modeling and control law design of small drones, and can perform dynamic model modeling and control law design quickly, accurately, and safely.

[0010] The specific solutions provided by the present invention are as follows:

[0011] In a first aspect, a method for small drone modeling and control law design based on flight test trials includes:

[0012] S1. Set up a test platform: Fix and arrange a triaxial force sensor in an obstacle-free test space using the fixing components of the auxiliary fixing device in the test space, bind the triaxial force sensor to the small drone. The test space and the auxiliary fixing device form a flight test platform for the small drone to fly under the support of the auxiliary fixing device in the hemispherical space. The force of support or traction given by the auxiliary fixing device to the small drone is set as the rope force. The triaxial force sensor acquires the rope force of the small drone during flight under the action of the engine thrust. The acting point of the resultant force received by the small drone approaches the center of gravity of the small drone. The test space and the auxiliary fixing device jointly limit the small drone from getting out of control and being higher than the ground, so that the small drone can fly without obstacles within the test space range;

[0013] S2. Pre - test configuration: Connect the tri - axial force sensor, the small unmanned aerial vehicle (UAV), and the flight control integrated with an attitude sensor on the small UAV to a computer respectively for controlling signal upload and obtaining signals from each sensor. Among them, the tri - axial force sensor transmits the measured rope force in real - time, and the flight control of the small UAV transmits sensor parameters, obtaining the measured attitude angles of pitch, roll, and yaw as [θ, φ, ψ], the three - axis velocity components of the body - axis system with the center of gravity of the small UAV as the coordinate origin as [u, v, w], and the three - axis angular velocity components as [p, q, r]. Then the state variable is denoted as x = [u, v, w, p, q, r, θ, φ, ψ]; Control the deflection of the control surface servo and the engine speed through PWM remote control, and determine the corresponding relationship between the three - channel control amounts of the PWM control signal and the state variables such as the rotation speed and deflection angle of the small UAV through measurement to obtain the preliminary control ability.

[0014] S3. Establishment of the flight - test model: Adjust the engine speed of the small UAV based on the preliminary control ability to enable the small UAV to achieve balanced flight of rope force, thrust, aerodynamic force, and gravity with human assistance; and record the force data and motion data of the small UAV in real - time to establish a linear state equation, that is, a linear dynamic model; Use methods such as frequency identification to perform a sweep - frequency operation on the three - channel control amounts in the balanced state of the small UAV to obtain the response characteristics of the small UAV in this balanced state, perform parameter identification operations on the small UAV in the effective frequency band to obtain the aerodynamic parameters of the UAV required in the linear dynamic model, and then obtain the state - space model matrix of the linear dynamic model, and obtain the transfer function of the control channel through Laplace transform.

[0015] S4. Control - law design and optimization: Based on the transfer function and the linear state equation, use PID control to generate a control - law design to obtain a preliminary control law. According to the preliminary control law, continue to conduct flight - test experiments on the small UAV on the test platform to obtain the transient response and steady - state response of the small UAV under the actual flight effect, obtain the time - domain analysis method and frequency - domain analysis method of the small UAV. Based on the transient response and steady - state response in the time - domain analysis method, and the amplitude - frequency characteristic and phase - frequency characteristic in the frequency - domain analysis method, judge whether the preliminary control law is reasonable, and then adjust and optimize the PID control parameters.

[0016] The auxiliary fixing device of the flight - test platform is a steel - frame link structure. The steel - frame link structure is a six - degree - of - freedom link structure fixed at the bottom on the ground, and its top is connected to a platform or frame as the fixed component through a ball - joint. The fixed connection point of the small UAV and the fixed component is such that the acting point of the resultant force received by the small UAV is at the connection point approaching the center of gravity of the small UAV.

[0017] The auxiliary fixing device is the ceiling, the fixing component is the fixing hanging point arranged on the ceiling, the three-axis force sensor is bound to the small unmanned aerial vehicle through an inelastic rope, the arrangement of the connection hanging points of the rope on the small unmanned aerial vehicle is such that the acting point of the resultant force of the rope on the small unmanned aerial vehicle is near the center of gravity of the small unmanned aerial vehicle, and the length of the rope is less than the vertical distance from the fixing hanging point to the ground. Preferably, in S2, the operations for controlling signal uploading and sensor signal acquisition include: uploading a PWM signal to the small unmanned aerial vehicle; acquiring the rope force data measured by the three-axis force sensor in real time; acquiring the current position, attitude, three-axis speed, three-axis acceleration, three-axis angular velocity, and three-axis angular acceleration of the small unmanned aerial vehicle measured by the attitude sensor. Preferably, in S2, the operations of controlling the deflection of the control surface servo and the engine speed through PWM remote control, and determining the corresponding relationship between the three-channel control amount of the PWM control signal and the state variables such as the rotation speed and deflection angle of the small unmanned aerial vehicle through measurement to obtain the preliminary control ability include: setting PWM signals with different duty cycles to control the rotation speed of the motor and the deflection angle of the servo of the small unmanned aerial vehicle; measuring the motor rotation speed and servo deflection angle corresponding to PWM signals with different duty cycles; plotting the corresponding curves of the PWM signal duty cycle with the rotation speed and the deflection angle to obtain the control effects of different PWM signals on the motor and servo of the small unmanned aerial vehicle, that is, obtaining the preliminary control ability.

[0018] Preferably, S3 specifically includes the following steps:

[0019] According to the preliminary control ability, using data communication pulse width modulation to remotely control the deflection of the control surface servo and the engine speed of the small unmanned aerial vehicle, and respectively denoting the control amounts of the three channels of roll, pitch, and yaw as [δ lat , δ lon , δ ped , and denoting the thrust throttle amount as [δ col , then in the dynamic model of the small unmanned aerial vehicle, the control amount is denoted as U = [δ lat , δ lon , δ ped , δ col ;

[0020] The flight control of the small unmanned aerial vehicle transmits back the parameters of the attitude sensor to obtain the pitch angle, roll angle, and yaw angle as [θ, φ, ψ] respectively, the three-axis velocity components of the body axis system with the center of gravity of the small unmanned aerial vehicle as the coordinate origin are [u, v, w], and the three-axis angular velocity components are [p, q, r], then the state variables of the small unmanned aerial vehicle are denoted as x = [u, v, w, p, q, r, θ, φ, ψ];

[0021] The three-axis force sensor transmits back the measured force The small unmanned aerial vehicle is under the rope force Thrust Aerodynamic force Gravity Under the action of, it reaches a stable flight equilibrium state. The resultant external force of the small unmanned aerial vehicle except gravity is denoted as The resultant external force The three-axis force components are X, Y, Z, and the resultant external torque components on the three axes are L, M, N;

[0022] Based on the above data, a linear dynamics model is established as follows, where is the derivative of x, g represents the gravitational acceleration, and the remaining variables represent the partial derivatives of X, Y, Z, L, M, N with respect to the state variable x and the control variable U. Taking X u as an example, X u represents the partial derivative of the variable X with respect to u:

[0023]

[0024] In the formula, x is the state variable of the small unmanned aerial vehicle, U is the control variable of the small unmanned aerial vehicle, A is the system matrix, and B is the control matrix;

[0025] Adopt the frequency identification method. At the equilibrium state of the small unmanned aerial vehicle, perform a sweep frequency operation on each control quantity [δ lat , δ lon , δ ped , δ col , obtain the response characteristics of the small unmanned aerial vehicle at the current equilibrium state, adopt the corresponding filter to reduce the interference of signals outside the frequency band, and extract the effective frequency band to perform parameter identification operations. The rope force F s that changes at different state points. The magnitude and direction of the rope force can be obtained through the three-axis force sensor. Therefore, the partial derivatives of the rope force with respect to the state variable and the control variable U are used as known quantities, and at the same time, the thrust is a known quantity. From this, the partial derivatives of the aerodynamic force F air with respect to the state variable and the control variable U can be separated from the parameters obtained by sweep frequency parameter identification. Thus, the aerodynamic parameters of the measured unmanned aerial vehicle in the normal flight state without the rope force are obtained, that is, the derivatives of lift and drag with respect to speed and attitude. Further, a linear dynamics model of the unmanned aerial vehicle under the action of only thrust and aerodynamic force can be obtained. Based on the model, the transfer function of the control channel can be further obtained through Laplace transform.

[0026] Preferably, the time domain analysis method and the frequency domain analysis method in S4 are specifically:

[0027] Time-domain analysis method: The transient response is the initial response of the small unmanned aerial vehicle after receiving a control signal, and the parameters include rise time, settling time, overshoot, and number of oscillations; the steady-state response is the difference between the actual state of the small unmanned aerial vehicle after flying to a balanced state under the current control signal and the desired state of the current control signal, and the parameters are steady-state error and steady-state gain;

[0028] Frequency-domain analysis method: The amplitude-frequency characteristic is the amplitude ratio of the system between the sinusoidal input signal and the output of the open-loop system, representing the amplification or attenuation ability of the system to input signals of different frequencies, and the parameter is gain margin; the phase-frequency characteristic is the phase difference between the sinusoidal input signal and the output of the open-loop system, representing the phase shift of the system to input signals of different frequencies, and the parameter is phase margin.

[0029] Preferably, in S4, judging whether the preliminary control law is reasonable according to the transient response and steady-state response in the time-domain analysis method and the amplitude-frequency characteristic and phase-frequency characteristic in the frequency-domain analysis method includes:

[0030] In the time-domain analysis method, a rise time threshold, a settling time threshold, an overshoot threshold, and a number of oscillations threshold are respectively set for the rise time, settling time, overshoot, and number of oscillations. When the rise time is less than the rise time threshold, the settling time is less than the settling time threshold, the overshoot is less than the overshoot threshold, and the number of oscillations is less than the number of oscillations threshold, it indicates that the preliminary control law is reasonable; otherwise, it is unreasonable. A steady-state gain threshold is set for the steady-state gain. When the steady-state error is 0 and the steady-state gain is equal to the steady-state gain threshold, it indicates that the preliminary control law is reasonable; otherwise, it is unreasonable;

[0031] In the frequency-domain analysis method, a gain margin threshold is set for the gain margin in the amplitude-frequency characteristic, and a phase margin threshold is set for the phase margin in the phase-frequency characteristic. When the gain margin is greater than the gain margin threshold and the phase margin is greater than the phase margin threshold, it indicates that the preliminary control law is reasonable; otherwise, it is unreasonable.

[0032] Preferably, adjusting and optimizing the PID control parameters in S4 includes: The control parameters include proportional gain, integral gain, and derivative gain; increasing or decreasing the proportional gain to improve the response speed of the small unmanned aerial vehicle or reduce overshoot and oscillation, and improve the gain margin and phase margin of the system; increasing or decreasing the integral gain to eliminate the steady-state error and reduce overshoot and oscillation, and improve the phase margin; increasing or decreasing the derivative gain to reduce overshoot or suppress the sensitivity to high-frequency noise, and improve the phase margin; adjusting each of the control parameters in turn, and gradually performing the flight test, and comprehensively adjusting the control parameters to make the various response performances of the system meet the requirements, then a controller with good control performance can be obtained.

[0033] In a second aspect, a small unmanned aerial vehicle (UAV) modeling and control law design system based on flight test includes: a test platform construction module, a test configuration module, a flight test model establishment module, and a control law design and optimization module;

[0034] The test platform construction module includes an auxiliary fixing device arranged in an unobstructed fixed space, a three-axis force sensor fixed by using a fixing component of the auxiliary fixing device, and a small UAV fixedly connected to the three-axis force sensor. The fixed space and the auxiliary fixing device form a flight test platform for the small UAV to fly under the support of the auxiliary fixing device in a hemispherical space. The force for the auxiliary fixing device to support or pull the small UAV is set as a rope force. The three-axis force sensor acquires the rope force of the small UAV during flight under the action of engine thrust. The acting point of the resultant force received by the small UAV approaches the center of gravity of the small UAV. The fixed space and the auxiliary fixing device jointly limit the small UAV from getting out of control and flying higher than the ground, so that it can fly without obstruction within the range of the test fixed space;

[0035] The test configuration module respectively connects the three-axis force sensor, the small UAV, and the flight control integrated with an attitude sensor of the small UAV to a computer for signal connection, which is used for uploading control signals and acquiring sensor signals; controls the deflection of the control surface servo and the engine speed through PWM remote control, measures and determines the corresponding relationship between the three-channel control quantity of the PWM control signal and state variables such as the rotation speed and deflection angle of the small UAV, and obtains the preliminary control ability;

[0036] The flight test model establishment module adjusts the engine speed of the small UAV based on the preliminary control ability, so that the small UAV realizes flight with the balance of rope force, thrust, aerodynamic force, and gravity under manual assistance; and records the force data and motion data of the small UAV in real time to establish a linear dynamics model; performs a frequency sweep operation on the three-channel control quantity in the balanced state of the small UAV by using a frequency identification method, obtains the response characteristics of the small UAV in this balanced state, extracts the effective frequency band to perform parameter identification operations, obtains the aerodynamic parameters of the UAV required in the linear dynamics model, and further obtains the state space model matrix of the linear dynamics model, and obtains the transfer function of the control channel through Laplace transform;

[0037] The control law design and optimization module, based on the transfer function and linear equation, uses PID control generation to design a control law to obtain a preliminary control law, and continues to perform flight tests on the small UAV on the test platform according to the preliminary control law, obtains the transient response and steady-state response of the small UAV under the actual flight effect, judges whether the preliminary control law is reasonable based on the transient response and steady-state response, and further adjusts and optimizes the PID control parameters.

[0038] Advantages of the present invention:

[0039] The present invention provides a method for modeling and control law design of a small unmanned aerial vehicle (UAV) based on flight test. By using the fixing components of the auxiliary fixing device to fixedly arrange a triaxial sensor connected to the small UAV in a fixed space without obstacles in S1, a flight test platform for the small UAV for flight test with measurable supporting force is built, enabling the small UAV to fly without obstacles. Then, through the pre-configuration of the flight test in S2, the preliminary control ability of the small UAV is obtained. In S3, a flight test model is established. Based on the preliminary control ability, the small UAV is brought to a balanced state, and the transfer function of each control channel of the small UAV and the linear state equation of the small UAV are obtained according to the data in the balanced state. In S4, the control law design is optimized. Based on the transfer function and the linear equation, a preliminary control law is generated by using PID control. The small UAV is flight-tested according to the preliminary control law, and it is judged whether the preliminary control law is reasonable according to the transient response and steady-state response of the small UAV, and then the PID control parameters are adjusted and optimized. Compared with the traditional dynamic model modeling method, the method provided by the present invention does not require expensive wind tunnel equipment or large ground test platforms to obtain aerodynamic parameters. The dynamic model is established based on actual flight data to determine the aerodynamic parameters and dynamic characteristics of the small UAV, which is closer to the real physical system and has high model accuracy. It does not require complex parameter measurement, aerodynamic simulation, wind tunnel test and other steps, nor multiple design test links, simplifies the modeling process, reduces costs and resource consumption, and greatly shortens the development cycle. For traditional control law optimization, PID is simulated and parameter-tuned on a computer and then used in practice. The process is cumbersome and there may be deviations. The present invention directly obtains signal data through test flight and adjusts parameters in real time, simplifies the optimization process, and improves accuracy.

[0040] Particularly, the present application uses the method of ropes or six-degree-of-freedom turntables to ensure the safety of the small UAV during flight in the initial stage of design. Even when the control law and flight dynamics model are imperfect, it can fly safely within the area to perform parameter measurement, model establishment and control law design optimization, ensuring that the experimental conditions can be better controlled and reproduced.

[0041] Particularly, the present application continuously verifies and iterates the dynamic model and control law of the small UAV during the flight test, with high design accuracy. At the same time, the control rate can be quickly adjusted according to the actual flight performance, accelerating the design speed and accuracy, reducing links such as simulation and hardware-in-the-loop simulation, reducing the cost of small UAV development and shortening the control system development cycle. Description of the Drawings

[0042] Figure 1Flow chart of a method for modeling and control law design of a small unmanned aerial vehicle based on flight test provided by the present invention.

[0043] Figure 2 Framework diagram of a system for modeling and control law design of a small unmanned aerial vehicle based on flight test provided by the present invention.

[0044] Figure 3 Structural schematic diagrams of two flight test platforms of the present invention. Detailed implementation manners

[0045] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0046] Figure 1 It is a flow chart of a method for modeling and control law design of a small unmanned aerial vehicle based on flight test provided by the present invention. As Figure 1 shown, a method for modeling and control law design of a small unmanned aerial vehicle based on flight test includes:

[0047] S1. Set up a test platform: In an unobstructed test space, use the auxiliary fixing device in the test space to connect the three-axis force sensor with the small unmanned aerial vehicle, ensuring that the small unmanned aerial vehicle can fly unobstructed within the test range of the space. This test platform is used to ensure the safety of the aircraft and personnel when the aircraft is unstable in the initial stage of the flight test. By using this method, expensive and complex wind tunnel tests are not required, thus reducing the cost of developing and testing small unmanned aerial vehicles.

[0048] Specifically, the above use of the auxiliary fixing device means using the fixing components of the auxiliary fixing device to fixedly arrange the three-axis force sensor, and then binding and connecting the three-axis force sensor with the small unmanned aerial vehicle. The test space and the auxiliary fixing device form a flight test platform for the small unmanned aerial vehicle to be supported and fly under the support of the auxiliary fixing device in the hemispherical space. The force of support or traction given by the auxiliary fixing device to the small unmanned aerial vehicle is set as the rope force. The three-axis force sensor obtains the rope force of the small unmanned aerial vehicle during flight under the action of the engine thrust, making the acting point of the resultant force received by the small unmanned aerial vehicle approach the center of gravity of the small unmanned aerial vehicle. The test space and the auxiliary fixing device jointly limit the small unmanned aerial vehicle from getting out of control and being higher than the ground, so that the small unmanned aerial vehicle can fly unobstructed within the test space range.

[0049] As Figure 3 shown, the auxiliary fixing device includes two types of auxiliary fixing devices: the steel frame link structure on the left and the bottom of the ceiling on the right. The auxiliary fixing device and the corresponding unmanned aerial vehicle can fly in the hemispherical range of space, and two test platforms on the left and right are respectively built.

[0050] Figure 3On the left side, the auxiliary fixing device for the flight test platform formed by the steel frame connecting rod structure and the test space thereon is the steel frame connecting rod structure. The steel frame connecting rod structure is a six-degree-of-freedom connecting rod structure (i.e., a six-degree-of-freedom turntable) with the bottom end fixed to the ground. Its top end is connected to a platform or frame as the fixing component through a ball joint. The fixed connection point between the small unmanned aerial vehicle and the fixing component of the steel frame connecting rod structure is the connection point where the acting point of the resultant force received by the small unmanned aerial vehicle approaches the center of gravity of the small unmanned aerial vehicle, so as to accurately measure the stress state of the small unmanned aerial vehicle. The installation point and connection method of the triaxial force sensor on the platform or frame should avoid loosening or displacement to ensure the accuracy of measurement.

[0051] Figure 3 On the right side, the auxiliary fixing device for the flight test platform formed by the space from the bottom of the ceiling to the ground to form the test space is the bottom of the ceiling. The fixing component of the auxiliary fixing device is a fixing piece fixedly connected to the bottom of the ceiling and the triaxial force sensor, such as fixing screws, etc., or a fixing hanging point for fixing the triaxial sensor. Then, use high-strength rope materials to connect the triaxial force sensor to the small unmanned aerial vehicle. When connecting, ensure that the acting point of the force on the sensor approaches or is located at the center of gravity of the small unmanned aerial vehicle, so as to accurately measure the stress state of the small unmanned aerial vehicle. Laser calibration tools or measuring tools can be used to repeatedly adjust the position to ensure accuracy. Then, adjust the height of the suspension point to ensure that the small unmanned aerial vehicle will not touch the ground in an out-of-control state, that is, the length of the rope material is less than the vertical distance from the fixed hanging point to the ground.

[0052] Both types of test platforms first measure the center of gravity, mass, moment of inertia, etc. of the aircraft through methods such as the traditional suspension method. Adjust the position of the unmanned aerial vehicle on the triaxial force sensor to ensure that the connection point or hanging point is the center of gravity. And through the adjustment of the hanging point position or the joint limit of the six-degree-of-freedom turntable, ensure that the out-of-control aircraft is higher than the ground, and at the same time ensure that it can fly without obstacles within the fixed space range. The triaxial force sensor is connected to the computer to real-time transmit the measured force, and the flight control of the unmanned aerial vehicle transmits the sensor parameters to obtain the measured heading angle, triaxial acceleration, and triaxial angular velocity.

[0053] S2. Configuration before the test: Connect the triaxial force sensor, the small unmanned aerial vehicle, and the flight control of the small unmanned aerial vehicle integrated with the attitude sensor to the computer respectively for controlling the upload of signals and obtaining the signals of each sensor; determine the corresponding relationship between the PWM signal and the rotation speed and deflection angle of the small unmanned aerial vehicle through the PWM remote control measurement experiment to obtain the preliminary control ability.

[0054] Specifically, the above for controlling signal uploading and sensor signal acquisition includes: uploading a PWM signal to the small unmanned aerial vehicle; acquiring the rope force data on the rope material or turntable measured in real time by the triaxial force sensor; acquiring the current position, attitude, triaxial velocity, triaxial acceleration, triaxial angular velocity, and triaxial angular acceleration of the small unmanned aerial vehicle measured by the attitude sensor.

[0055] Exemplarily, a signal transmission connection is established between a computer, a small unmanned aerial vehicle, a triaxial force sensor, and an attitude sensor to achieve the uploading of control signals, the data acquisition of the triaxial force sensor, and the data acquisition of the attitude sensor. The computer generates PWM (pulse width modulation) signals with different duty cycles to control the flight attitudes such as the rotation speed and direction of the motors of the small unmanned aerial vehicle. The PWM signals are transmitted from the computer to the control system of the small unmanned aerial vehicle through signal acquisition lines to adjust the actions of the small unmanned aerial vehicle; each PWM signal corresponds to a control instruction for a motor, such as the propulsion speed and direction, and the fine adjustment of the attitude of the small unmanned aerial vehicle is achieved by changing the duty cycle of the PWM. During the flight of the small unmanned aerial vehicle, the triaxial force sensor captures the force changes along the XYZ axes, reflecting the forces received by the small unmanned aerial vehicle during the test, including tensile force, thrust force, and reaction force, etc. The triaxial force sensor transmits the measured force data to the computer through a data acquisition device; the acquisition frequency can be set to a high-frequency mode to capture subtle force changes and ensure the real-time nature of the data. The attitude sensor is used to measure the flight attitude and position state of the small unmanned aerial vehicle and includes six key data, specifically,

[0056] Position: The position coordinates of the small unmanned aerial vehicle in three-dimensional space;

[0057] Attitude: The spatial attitude angles of the small unmanned aerial vehicle, including pitch angle, yaw angle, and roll angle;

[0058] Triaxial velocity: The linear velocity along the X, Y, and Z axes;

[0059] Triaxial acceleration: The acceleration along the X, Y, and Z axes;

[0060] Triaxial angular velocity: The angular velocity around the X, Y, and Z axes;

[0061] Triaxial angular acceleration: The angular acceleration around the X, Y, and Z axes.

[0062] The data of the attitude sensor is acquired through an attitude module and transmitted to the computer through signal acquisition lines to ensure the stability and accuracy of data transmission; a high-frequency sampling frequency is also set to accurately reflect the motion state of the small unmanned aerial vehicle.

[0063] Specifically, the above-mentioned correspondence between the PWM signal and the rotational speed and deflection angle of the small unmanned aerial vehicle is determined through the PWM remote control measurement experiment to obtain the preliminary control ability, including: setting PWM signals with different duty cycles to control the rotational speed of the motor and the deflection angle of the servo of the small unmanned aerial vehicle; measuring the rotational speed of the motor and the deflection angle of the servo corresponding to the PWM signals with different duty cycles; and plotting the correspondence curve between the duty cycle of the PWM signal and the rotational speed and the deflection angle to obtain the control effects of different PWM signals on the motor and the servo of the small unmanned aerial vehicle, that is, obtaining the preliminary control ability.

[0064] Exemplarily, PWM (pulse width modulation) signals with different duty cycles are set to control the rotational speed, direction and flight attitude of the motor of the small unmanned aerial vehicle. The PWM signals are connected and transmitted to the control system of the small unmanned aerial vehicle through signal acquisition lines to adjust the actions of the small unmanned aerial vehicle. While the flight state of the small unmanned aerial vehicle changes, the computer receives the data of the three-axis force sensor and the attitude sensor in real time, and extracts the key features of the force and attitude changes of the small unmanned aerial vehicle through the data analysis program. The computer compares and analyzes the correspondence between the force data and the attitude data and the uploaded PWM control signal, and establishes a mathematical model of the PWM signal and the attitude and position changes of the small unmanned aerial vehicle. According to the real-time collected data, the computer can adjust the duty cycle of the PWM signal when necessary to achieve the closed-loop control of the movement of the small unmanned aerial vehicle, so that the small unmanned aerial vehicle maintains stable flight and avoids losing control during the test. By changing the PWM signal, observing the reaction of the small unmanned aerial vehicle, and recording the corresponding force and attitude data, the data is analyzed and verified; verifying whether the relationship between the PWM signal and the movement state of the small unmanned aerial vehicle is accurate, and further optimizing the established control model so that the small unmanned aerial vehicle can accurately complete the flight actions according to the expected instructions, and finally form the preliminary ability of the small unmanned aerial vehicle control.

[0065] S3. Flight test model establishment: Based on the preliminary control ability, adjust the control signal of the small unmanned aerial vehicle (UAV), adjust the engine speed, and enable the small UAV to reach a balanced state of stable flight with the assistance of human, that is, achieve the balance of rope force, thrust, aerodynamic force, and gravity; and record the force data and motion data of the small UAV in real time, establish a linear state equation, namely a linear dynamics model, and perform a sweep frequency operation on the three-channel control quantities in the balanced state of the small UAV by using methods such as frequency identification, that is, perform single-channel sweep frequency on the pitch channel, roll channel, and yaw channel, obtain the response characteristics of the small UAV in this balanced state, perform parameter identification operation on the small UAV to extract the effective frequency band, obtain the aerodynamic parameters of the UAV required in the linear dynamics model, and then obtain the state space model matrix of the linear dynamics model, and obtain the transfer function of each channel for controlling the small UAV through Laplace transform. The human assistance means: For example, in some states, it hovers at 45° in the east direction. Due to the imperfect control law, continuous adjustment and fluctuations are required to reach this state. The aircraft can be held or clamped to reach the expected position first, and then the control signal is stabilized in this state to reduce the adjustment time.

[0066] Specifically, according to the preliminary control ability, use data communication pulse width modulation remote control to control the deflection of the rudder servo and the engine speed of the small UAV. Denote the control quantities of the roll, pitch, and yaw channels as [δ lat ,δ lon ,δ ped , and denote the thrust throttle quantity as [δ col . Then, in the dynamics model of the small UAV, the control quantity is denoted as U = [δ lat ,δ lon ,δ ped ,δ col ;

[0067] The small UAV flight control transmits back the attitude sensor parameters, and the pitch angle, roll angle, and yaw angle are obtained as [θ, φ, ψ] respectively. The three-axis velocity components of the body axis system with the center of gravity of the small UAV as the coordinate origin are [u, v, w], and the three-axis angular velocity components are [p, q, r]. Then, the state variables of the small UAV are denoted as x = [u, v, w, p, q, r, θ, φ, ψ];

[0068] The three-axis force sensor transmits back the measured force The small UAV is under the action of thrust aerodynamic force gravity to reach a balanced state of stable flight. The resultant external force of the small UAV except gravity is denoted as The resultant external force When the three-axis force components are X, Y, and Z, and the resultant external torque components in the three axes are L, M, and N;

[0069] Based on the above data, a linear dynamics model is established as follows, where is the derivative of x, g represents the acceleration due to gravity, and the remaining variables represent the partial derivatives of X, Y, Z, L, M, and N with respect to the state variable x and the control variable U. Taking X u as an example, X u represents the partial derivative of the variable X with respect to u:

[0070]

[0071] In the formula, x is the state variable of the small unmanned aerial vehicle, U is the control variable of the small unmanned aerial vehicle, A is the system matrix, and B is the control matrix;

[0072] Adopt the frequency identification method. When the small unmanned aerial vehicle is in the equilibrium state, perform a sweep frequency operation on each control quantity [δ lat , δ lon , δ ped , δ col , obtain the response characteristics of the small unmanned aerial vehicle in the current equilibrium state, use the corresponding filter to reduce the interference of signals outside the frequency band, and extract the effective frequency band to perform the parameter identification operation. The rope force F s that changes at different state points. The magnitude and direction of the rope force can be obtained through the three-axis force sensor. Therefore, the partial derivatives of the rope force with respect to the state variable x and the control variable U are used as known quantities. At the same time, the thrust is a known quantity. From this, the partial derivatives of the aerodynamic force F air with respect to the state variable x and the control variable U can be separated from the parameters obtained by the sweep frequency parameter identification. Thus, the aerodynamic parameters of the measured unmanned aerial vehicle in the normal flight state without the rope force are obtained, that is, the derivatives of the lift and drag with respect to the speed and attitude. Further, the linear dynamics model of the unmanned aerial vehicle under the action of only the thrust and the aerodynamic force can be obtained. Based on the model, the transfer function of the control channel can be further obtained through the Laplace transform.

[0073] S4. Control law design and optimization: Based on the transfer function and the linear state equation, use PID control generation to design the control law to obtain the preliminary control law. According to the preliminary control law, continue to conduct flight tests on the small unmanned aerial vehicle on the test platform to obtain the transient response and steady-state response of the small unmanned aerial vehicle under the actual flight effect, obtain the time-domain analysis method and frequency-domain analysis method of the small unmanned aerial vehicle. Based on the transient response and steady-state response in the time-domain analysis method, and the amplitude-frequency characteristic and phase-frequency characteristic in the frequency-domain analysis method, judge whether the preliminary control law is reasonable, and then adjust and optimize the PID control parameters.

[0074] Specifically, the above time-domain analysis method and frequency-domain analysis method include: In the time-domain analysis method, the transient response is the initial response of the small unmanned aerial vehicle after receiving a control signal, and the parameters include rise time, settling time, overshoot, and number of oscillations; the steady-state response is the difference between the actual state of the small unmanned aerial vehicle after flying to a balanced state under the current control signal and the desired state of the current control signal, and the parameters are steady-state error and steady-state gain. In the frequency-domain analysis method, the amplitude-frequency characteristic is the amplitude ratio of the system between the sine input signal and the output of the open-loop system, indicating the amplification or attenuation ability of the system to input signals of different frequencies, and the parameter is gain margin; the phase-frequency characteristic is the phase difference between the sine input signal and the output of the open-loop system, showing the phase shift of the system to input signals of different frequencies, and the parameter is phase margin.

[0075] Specifically, judging whether the preliminary control law is reasonable based on the transient response and steady-state response in the time-domain analysis method and the amplitude-frequency characteristic and phase-frequency characteristic in the frequency-domain analysis method includes:

[0076] In the time-domain analysis method, set a rise time threshold, a settling time threshold, an overshoot threshold, and a number of oscillations threshold for the rise time, settling time, overshoot, and number of oscillations respectively. When the rise time is less than the rise time threshold, the settling time is less than the settling time threshold, the overshoot is less than the overshoot threshold, and the number of oscillations is less than the number of oscillations threshold, it indicates that the preliminary control law is reasonable; otherwise, it is unreasonable. Set a steady-state gain threshold for the steady-state gain. When the steady-state error is 0 and the steady-state gain is equal to the steady-state gain threshold, it indicates that the preliminary control law is reasonable; otherwise, it is unreasonable. In the frequency-domain analysis method, set a gain margin threshold for the gain margin in the amplitude-frequency characteristic and a phase margin threshold for the phase margin in the phase-frequency characteristic. When the gain margin is greater than the gain margin threshold and the phase margin is greater than the phase margin threshold, it indicates that the preliminary control law is reasonable; otherwise, it is unreasonable.

[0077] Specifically, adjusting and optimizing the PID control parameters in S4 above includes: The control parameters include proportional gain, integral gain, and derivative gain; increase or decrease the proportional gain to improve the response speed of the small unmanned aerial vehicle or reduce overshoot and oscillations, and improve the gain margin and phase margin of the system; increase or decrease the integral gain to eliminate the steady-state error and reduce overshoot and oscillations, and improve the phase margin; increase or decrease the derivative gain to reduce overshoot or suppress the sensitivity to high-frequency noise, and improve the phase margin; adjust each of the control parameters in sequence and gradually conduct the flight test, and comprehensively adjust the control parameters to make each response performance of the system meet the requirements, then a controller with good control performance can be obtained.

[0078] Exemplarily, based on the transfer functions of each channel of the small unmanned aerial vehicle (UAV) obtained in S3 and the linear state equation of the small UAV, the proportional (P), integral (I), and derivative (D) parameters of the PID controller are initially set to generate a preliminary control law. Among them, proportional control determines the response speed, integral control eliminates the steady-state error, and derivative control is used to suppress system oscillation and overshoot. Apply the preliminarily designed control law to the flight control system of the small UAV, and enable the small UAV to perform basic flight operations such as hovering and moving on the test platform. Adjust the PWM signal output through the PID controller to make the small UAV reach the desired attitude and position, thereby verifying the basic feasibility of the control law.

[0079] During the flight of the small UAV, through time-domain analysis, observe the transient response and steady-state response of the small UAV. The transient response includes indicators such as rise time, overshoot, and settling time, which reflect the system response speed and stability; the steady-state response measures the error of the system at steady state to ensure that the small UAV can maintain stability when at the target position or attitude. At the same time, perform frequency-domain analysis using frequency-domain analysis method. Through the amplitude-frequency characteristic and phase-frequency characteristic, evaluate the frequency response performance of the system. The amplitude-frequency characteristic judges the gain response of the system to inputs of different frequencies to ensure the stability of the system within a specific frequency range; the phase-frequency characteristic judges the phase lag situation of the system to avoid instability caused by a large phase difference in the response.

[0080] Based on the above indicators of transient response, steady-state response, amplitude-frequency characteristic, and phase-frequency characteristic, judge the rationality of the control law. If there are problems such as excessive overshoot, slow response speed, or high steady-state error, it indicates that the preliminary control law is not reasonable enough and needs to be optimized. According to the index analysis results, adjust the parameters of the PID controller:

[0081] Proportional parameter (P): If the response speed is slow, the P value can be appropriately increased, but too large a value will cause overshoot;

[0082] Integral parameter (I): Used to eliminate the steady-state error, but too large an integral term will cause system oscillation;

[0083] Derivative parameter (D): Adjust the derivative term to suppress system oscillation and overshoot. Increasing the D value can improve the stability of the system.

[0084] After adjusting the parameters, flight tests and data collection are carried out repeatedly until the response of the small unmanned aerial vehicle (UAV) meets the expected requirements. After multiple tests and parameter adjustments, the PID control parameters are finally determined to make the control law meet the performance requirements of the system. The optimized control law is applied to the flight control of the small UAV, and a complete flight test is carried out to ensure that the system maintains good stability and response speed under different flight attitudes and position changes. The robustness of the control system is verified under different working conditions (such as different loads, wind speed environments) to ensure that the control law has stability and reliability in practical applications.

[0085] Figure 2 It is a framework diagram of a system for modeling and control law design of a small unmanned aerial vehicle based on flight test provided by the present invention. As Figure 2 shown, it includes: a test platform construction module, a test configuration module, a flight test model establishment module, and a control law design and optimization module;

[0086] The test platform construction module includes an auxiliary fixing device arranged in an unobstructed fixed space, a three-axis force sensor fixed by using the fixing part of the auxiliary fixing device, and a small unmanned aerial vehicle fixedly connected to the three-axis force sensor. The fixed space and the auxiliary fixing device form a flight test platform for the small unmanned aerial vehicle to fly under the support of the auxiliary fixing device in the hemispherical space. The force of the auxiliary fixing device supporting or pulling the small unmanned aerial vehicle is set as the rope force. The three-axis force sensor obtains the rope force of the small unmanned aerial vehicle under the action of the engine thrust. The acting point of the resultant force received by the small unmanned aerial vehicle approaches or is located at the center of gravity of the small unmanned aerial vehicle. The fixed space and the auxiliary fixing device jointly limit the small unmanned aerial vehicle from getting out of control and being higher than the ground, so that it can fly without obstruction within the test range of the fixed space;

[0087] The test configuration module connects the three-axis force sensor, the small unmanned aerial vehicle, and the flight control integrated with the attitude sensor of the small unmanned aerial vehicle to a computer for controlling signal uploading and sensor signal acquisition. The signal connection is a wireless signal connection or a signal acquisition line connection; the deflection of the control surface servo and the engine speed are controlled by PWM remote control, and the corresponding relationship between the three-channel control amount of the PWM signal and the state variables such as the rotation speed and deflection angle of the small unmanned aerial vehicle is determined by experimental measurement to obtain the preliminary control ability;

[0088] The flight test model establishment module adjusts the control signal of the small unmanned aerial vehicle (UAV) based on the preliminary control ability, adjusts the engine speed, and enables the small UAV to reach a balanced state of stable flight with the assistance of human, that is, to achieve flight with the balance of rope force, thrust, aerodynamic force and gravity; and records the force data and motion data of the small UAV in real time, establishes a linear dynamics model, performs a frequency sweep operation on the three-channel control quantities in the balanced state of the small UAV by using the frequency identification method, performs single-channel frequency sweep on the pitch channel, roll channel and yaw channel, obtains the response characteristics of the small UAV in this balanced state, extracts the effective frequency band to perform parameter identification operation, obtains the aerodynamic parameters of the UAV required in the linear dynamics model, and further obtains the state space model matrix of the linear dynamics model, and obtains the transfer function of each control channel through Laplace transform; the human assistance means that, for example, in some states, the UAV hovers at 45° in the east direction. Due to the imperfect control law, continuous adjustment and fluctuations are required to reach this state. The aircraft can be held or clamped to reach the expected position first, and then the control signal is stabilized in this state to reduce the adjustment time.

[0089] The control law design and optimization module generates a preliminary control law by using PID control based on the transfer function and linear equation, and continues to perform flight tests on the small UAV on the test platform according to the preliminary control law, obtains the transient response and steady-state response of the small UAV under the actual flight effect, judges whether the preliminary control law is reasonable based on the transient response and steady-state response, and further adjusts and optimizes the PID control parameters.

[0090] It should be noted that the above specific implementation manners can enable those skilled in the art to understand the present invention more comprehensively, but do not limit the present invention in any way. Therefore, although this specification has described the present invention in detail with reference to the drawings and embodiments, those skilled in the art should understand that the present invention can still be modified or equivalently replaced. In short, all technical solutions and their improvements that do not depart from the spirit and scope of the present invention should be covered by the protection scope of the patent of the present invention.

Claims

1. A small UAV modeling and control law design method based on flight test, including: S1. Build a test platform: In an obstacle-free test space, a three-axis force sensor is fixedly arranged using the fixing components of the auxiliary fixing device in the test space, and the three-axis force sensor is bound to the small UAV. The test space and the auxiliary fixing device form a flight test platform for the small UAV to be supported by the auxiliary fixing device for test flight in a hemispherical space. The supporting or pulling force given to the small UAV by the auxiliary fixing device is set as the rope force. The three-axis force sensor obtains the rope force of the small UAV under the action of the engine thrust. The point of action of the resultant force on the small UAV approaches the center of gravity of the small UAV. The test space and the auxiliary fixing device jointly limit the small UAV from losing control above the ground, so that the small UAV can fly without obstacles within the test space. S2. Configuration before the test: The three-axis force sensor, the small UAV and the flight control with an integrated attitude sensor on the small UAV are respectively connected to the computer for signal uploading and acquisition of each sensor signal, wherein the three-axis force sensor transmits back the measured rope force in real time, and the small UAV flight control transmits back the sensor parameters; The deflection of the rudder servo and the engine speed are controlled by PWM remote control, and the corresponding relationship between the three-channel control quantity of the PWM control signal and the state variables such as the speed and deflection angle of the small UAV is determined by measuring, so as to obtain the preliminary control capability; S3. Establishment of test flight model: Based on the preliminary control capability, the engine speed of the small UAV is adjusted to enable the small UAV to achieve flight with rope force, thrust, aerodynamic force and gravity balance under human assistance; and the force data and motion data of the small UAV are recorded in real time to establish a linear state equation, i.e., a linear dynamic model; frequency identification and other methods are used to perform frequency sweeping operations on the three-channel control quantities in the equilibrium state of the small UAV to obtain the response characteristics of the small UAV in the equilibrium state, and the effective frequency band is extracted to perform parameter identification operations on the small UAV to obtain the aerodynamic parameters of the UAV required in the linear dynamic model, and then the state space model matrix of the linear dynamic model is obtained, and the transfer function of the control channel is obtained by Laplace transformation; S4. Control law design and tuning: Based on the transfer function and linear state equation, PID control generation is used to design the control law to obtain a preliminary control law. According to the preliminary control law, the small UAV is continuously tested on the test platform to obtain the transient response and steady-state response of the small UAV under actual flight effects, and the time domain analysis method and frequency domain analysis method of the small UAV are obtained. Based on the transient response and steady-state response in the time domain analysis method, and the amplitude-frequency characteristic and phase-frequency characteristic in the frequency domain analysis method, it is determined whether the preliminary control law is reasonable, and then the PID control parameters are adjusted and optimized.

2. According to the small UAV modeling and control law design method based on flight test according to claim 1, it is characterized in that The auxiliary fixing device of the flight test platform is a steel frame connecting rod structure, which is a six-degree-of-freedom connecting rod structure with its bottom end fixed to the ground, and its top end is connected to a platform or frame serving as the fixing component through a ball joint. The fixed connection point between the small UAV and the fixing component is a connection point that makes the point of action of the resultant force on the small UAV approach the center of gravity of the small UAV.

3. The small UAV modeling and control law design method based on flight test according to claim 1 or 2 is characterized in that The auxiliary fixing device is a ceiling, the fixing component is a fixed hanging point arranged on the ceiling, the three-axis force sensor is bound to the small drone through an inelastic rope, and the connection hanging point on the small drone and the rope is arranged so that the point of action of the rope on the small drone is near the center of gravity of the small drone, and the length of the rope is less than the vertical distance from the fixed hanging point to the ground. Preferably, the control signal upload and sensor signal acquisition in S2 include: uploading a PWM signal to the small drone; acquiring the rope force data measured in real time by the three-axis force sensor; acquiring the current position, attitude, three-axis velocity, three-axis acceleration, three-axis angular velocity and three-axis angular acceleration of the small drone measured by the attitude sensor.

4. According to the small UAV modeling and control law design method based on flight test according to claim 3, It is characterized in that the deflection of the rudder servo and the engine speed are controlled by PWM remote control as described in S2, and the preliminary control capability is obtained by measuring and determining the correspondence between the three-channel control quantity of the PWM control signal and the state variables such as the speed and deflection angle of the small UAV, including: setting PWM signals with different duty cycles to control the speed of the motor of the small UAV and the deflection angle of the servo; measuring the motor speed and the deflection angle of the servo corresponding to the PWM signals with different duty cycles; drawing the corresponding curves of the PWM signal duty cycle and the speed and the deflection angle to obtain the control effects of different PWM signals on the motor and the servo of the small UAV, that is, obtaining the preliminary control capability.

5. The small UAV modeling and control law design method based on flight test according to claim 4 is characterized in that S3 specifically includes the following steps: According to the preliminary control capability, the rudder deflection and engine speed of the small UAV are remotely controlled by using data communication pulse width modulation. The control quantities of the three channels of roll, pitch and yaw are recorded as [δ lat ,δ lon ,δ ped ], the thrust throttle is recorded as [δ col ], then in the dynamic model of the small UAV, the control quantity is recorded as U = [δ lat ,δ lon ,δ ped ,δ col ]; The small UAV flight control returns the attitude sensor parameters, and the pitch angle, roll angle and yaw angle are obtained as [θ, φ, ψ] respectively. The three-axis velocity components of the body axis system with the center of gravity of the small UAV as the coordinate origin are [u, v, w], and the three-axis angular velocity components are [p, q, r]. Then the state variable of the small UAV is recorded as x = [u, v, w, p, q, r, θ, φ, ψ]; The three-axis force sensor transmits the measured force F s , the small UAV has a rope force F s Thrust T t 、 Aerodynamic force F air , the small UAV reaches a stable flight equilibrium state under the action of gravity G=mg, and the combined external force of the small UAV except gravity is recorded as the vector sum of the three forces F=F s +T t +F air , the three-axis force components of the total external force F are X, Y, and Z, and the three-axis torque components of the total external moment are L, M, and N; Based on the above data, a linear kinetic model is established as follows: is the derivative of x, g represents the acceleration of gravity, and the remaining variables represent the partial derivatives of X, Y, Z, L, M, and N on the state variable x and the control variable U. u For example, X u Represents the partial derivative of variable X on u: Wherein, x is the state variable of the small UAV, U is the control quantity of the small UAV, A is the system matrix, and B is the control matrix.

6. The small UAV modeling and control law design method based on flight test according to claim 5 is characterized in that S3 also includes taking a frequency identification method to identify each control variable [δ lat ,δ lon ,δ ped ,δ col ] performs a frequency sweep operation to obtain the response characteristics of the small UAV in the current equilibrium state, adopts a corresponding filter to reduce the interference of non-frequency band signals, extracts the effective frequency band to perform parameter identification operations; the rope force F that varies at different state points s The magnitude and direction of the rope force can be obtained through the three-axis force sensor, so that the partial derivative of the rope force on the state quantity x and the control quantity U is taken as a known quantity, and the thrust is also a known quantity, so that the individual aerodynamic force F can be separated from the parameters obtained by the sweep parameter identification. air The partial derivatives on the state quantity x and the control quantity U are used to obtain the aerodynamic parameters of the measured UAV in normal flight state without rope force, that is, the derivatives of lift and drag on speed and attitude; further, the linear dynamic model of the UAV under the action of thrust and aerodynamic force only can be obtained, and on the basis of the model, the transfer function of the control channel can be further obtained through Laplace transform.

7. The method for modeling and designing control laws of small unmanned aerial vehicles based on flight tests according to claim 6 is characterized in that The time domain analysis method and frequency domain analysis method described in S4 are specifically: Time domain analysis method: The transient response is the initial reaction of the small UAV after receiving the control signal, and the parameters include rise time, adjustment time, overshoot and number of oscillations; the steady-state response is the difference between the actual state of the small UAV after reaching the equilibrium state under the current control signal and the expected state of the current control signal, and the parameters are steady-state error and steady-state gain; Frequency domain analysis method: The amplitude-frequency characteristic is the amplitude ratio of the system's sinusoidal input signal and the open-loop system output, which represents the system's ability to amplify or attenuate input signals of different frequencies, and the parameter is the amplitude margin; the phase-frequency characteristic is the phase difference between the system's sinusoidal input signal and the open-loop system output, which represents the system's phase shift for input signals of different frequencies, and the parameter is the phase margin.

8. The method for modeling and designing control laws of small unmanned aerial vehicles based on flight tests according to claim 7 is characterized in that The method of judging whether the preliminary control law is reasonable according to the transient response and steady-state response in the time domain analysis method and the amplitude-frequency characteristic and phase-frequency characteristic in the frequency domain analysis method in S4 includes: In the time domain analysis method, a rise time threshold, an adjustment time threshold, an overshoot threshold and an oscillation number threshold are set for the rise time, the adjustment time, the overshoot and the oscillation number respectively. When the rise time is less than the rise time threshold, the adjustment time is less than the adjustment time threshold, the overshoot is less than the overshoot threshold and the oscillation number is less than the oscillation number threshold, it indicates that the preliminary control law is reasonable, otherwise, it is unreasonable; a steady-state gain threshold is set for the steady-state gain. When the steady-state error is 0 and the steady-state gain is equal to the steady-state gain threshold, it indicates that the preliminary control law is reasonable, otherwise, it is unreasonable; In the frequency domain analysis method, an amplitude margin threshold is set for the amplitude margin in the amplitude-frequency characteristic, and a phase margin threshold is set for the phase margin in the phase-frequency characteristic. When the amplitude margin is greater than the amplitude margin threshold and the phase margin is greater than the phase margin threshold, it indicates that the preliminary control law is reasonable, otherwise, it is unreasonable.

9. The method for modeling and designing control laws of small unmanned aerial vehicles based on flight tests according to claim 8 is characterized in that The adjustment and optimization of PID control parameters described in S4 includes: the control parameters include proportional gain, integral gain and differential gain; increasing or decreasing the proportional gain to improve the response speed of the small UAV or reduce overshoot and oscillation, and improve the amplitude margin and phase margin of the system; increasing or decreasing the integral gain to eliminate steady-state errors and reduce overshoot and oscillation, and improve the phase margin; increasing or decreasing the differential gain to reduce overshoot or suppress high-frequency noise sensitivity and improve the phase margin; adjusting each of the control parameters in turn, and gradually conducting the flight test, comprehensively adjusting the control parameters to achieve that each response performance of the system meets the requirements, and a controller with good control performance can be obtained.

10. A small UAV modeling and control law design system based on flight test, comprising: Test platform construction module, test configuration module, flight test model establishment module and control law design and tuning module; The test platform building module includes an auxiliary fixing device arranged in an obstacle-free fixed space, a three-axis force sensor fixed by a fixing component of the auxiliary fixing device, and a small UAV fixedly connected to the three-axis force sensor, the fixed space and the auxiliary fixing device form a flight test platform for the small UAV to be supported by the auxiliary fixing device for test flight in a hemispherical space, the force of the auxiliary fixing device to support or pull the small UAV is set as a rope force, the three-axis force sensor obtains the rope force of the small UAV under the action of engine thrust, the point of action of the resultant force on the small UAV approaches the center of gravity of the small UAV, and the fixed space and the auxiliary fixing device jointly limit the small UAV from being out of control above the ground, so that it can fly obstacle-free within the fixed space range of the test; The test configuration module connects the three-axis force sensor, the small UAV and the flight control of the small UAV with an integrated attitude sensor to the computer for signal connection, so as to upload control signals and obtain sensor signals; controls the deflection of the rudder servo and the engine speed through PWM remote control, measures and determines the corresponding relationship between the three-channel control quantity of the PWM control signal and the state variables such as the speed and deflection angle of the small UAV, and obtains preliminary control capability; The test flight model establishment module adjusts the engine speed of the small UAV based on the preliminary control capability, so that the small UAV can achieve flight with balanced rope force, thrust, aerodynamic force and gravity under human assistance; The force data and motion data of the small UAV are recorded in real time to establish a linear dynamics model; a frequency identification method is used to perform a frequency sweep operation on the three-channel control quantity in a balanced state of the small UAV to obtain the response characteristics of the small UAV in the balanced state, and an effective frequency band is extracted to perform a parameter identification operation to obtain the aerodynamic parameters of the UAV required in the linear dynamics model, thereby obtaining the state space model matrix of the linear dynamics model, and obtaining the transfer function of the control channel through Laplace transformation; The control law design and tuning module, based on the transfer function and the linear equation, uses PID control generation to design the control rate to obtain a preliminary control law, and continues to perform flight tests on the small UAV on the test platform according to the preliminary control law to obtain the transient response and steady-state response of the small UAV under actual flight effects. Whether the preliminary control law is reasonable is determined based on the transient response and steady-state response, and then the PID control parameters are adjusted and optimized.

Images