High maneuverability unmanned aerial vehicle control method based on setting control parameters

Abstract

The application discloses a high-maneuverability unmanned aerial vehicle control method based on setting control parameters, wherein by setting a suitable fitness function, parameters of a PI controller are set, and therefore the fitness of the uncertainty of the model is improved; in the process of setting the parameters, the corresponding tool provided by matlab software is directly used to cooperate with the fitness function, so that the parameters of the PI controller can be conveniently and efficiently obtained, and the unmanned aerial vehicle is controlled according to the parameters, so that the response time is shortened, and the flight quality of the unmanned aerial vehicle is improved.

Description

Technical Field

[0001] This invention relates to the field of unmanned aerial vehicle (UAV) control, and more specifically to a control method for highly maneuverable UAVs based on tuned control parameters. Background Technology

[0002] The drone controls itself based on the desired roll angle obtained from its guidance module. Specifically, this is achieved by controlling the motors to control the drone's roll, i.e., controlling the roll angular velocity, and thus the roll angle. However, because the drone is subjected to various forces during flight, its actual flight trajectory can be disturbed. Therefore, adjustments to the desired roll angle are necessary to ensure that the final drone flight state closely approximates the expectation. The time required for this adjustment response, i.e., the response speed, directly affects the drone's final flight performance. Many traditional drone control schemes focus on improving response speed, but the designs for specific drones are often too complex and difficult to scale up.

[0003] PI control is one of the most widely used control strategies in engineering applications, and it can be applied to the regulation and response schemes of unmanned aerial vehicles (UAVs). Therefore, parameter optimization in PI control strategies is also a research direction that has attracted considerable attention.

[0004] In general, a PI controller can be designed and the corresponding parameters calculated for a typical UAV model. However, this requires the establishment of a relatively accurate UAV kinematic and dynamic model, so the reliability of the calculated results needs to be improved. In the specific field of attitude control models for quadrotor UAVs, there is currently no research on using genetic algorithms to tune PI control parameters for UAV attitude control.

[0005] Based on this, the inventors used a genetic algorithm to tune the PI control parameters of the quadcopter unmanned aerial vehicle (UAV) attitude control model, hoping to obtain a UAV control method that can solve the above problems. Summary of the Invention

[0006] To overcome the aforementioned problems, the inventors conducted intensive research and designed a high-maneuverability UAV control method based on tuned control parameters. In this method, by setting an appropriate fitness function, the parameters of the PI controller are tuned, thus improving the fitness for model uncertainties. During the parameter tuning process, MATLAB software provides corresponding tools that, in conjunction with the fitness function, can conveniently and efficiently obtain the PI controller parameters. Based on these parameters, the UAV can be controlled, shortening the response time and improving the flight quality of the UAV, thereby completing this invention.

[0007] Specifically, the purpose of this invention is to provide a high-maneuverability unmanned aerial vehicle (UAV) control method based on tuned control parameters. In this method, the desired roll angle is received in real time. Based on the deviation between the actual roll angle and the desired roll angle, a motor control command is obtained through the controller of the roll channel. The motor speed is controlled accordingly to respond to the desired roll angle, thereby enabling the actual roll angle of the UAV to quickly and accurately track the desired roll angle.

[0008] The motor control command is obtained through the following formula (I):

[0009]

[0010] Where thr represents the motor control command;

[0011] φ represents the actual roll angle;

[0012] φ d Indicates the expected roll angle;

[0013] K P and K I Each represents the controller gain coefficient independently;

[0014] p represents the roll angular velocity of the aircraft;

[0015] s represents the Laplace transform operator.

[0016] Wherein, the controller gain coefficient K P and K I The fitness function is obtained through a genetic algorithm, where the fitness function in the genetic algorithm is Equation (II).

[0017] t s =p 00 +p 10 ·K P +p 01 ·K I +p 20 ·K P 2 +p 11 ·K P ·K I +p 02 ·K I 2 (two)

[0018] Among them, t s Indicates the time required for adjustment response;

[0019] p 00 p 10 p 01 p 20 p 11 p02 Each parameter represents a design parameter independently.

[0020] In the genetic algorithm, when the number of iterations reaches a preset number or the fitness reaches a set threshold, the combination of individuals that minimizes the time required for adjustment response is the controller gain coefficient K output by the genetic algorithm. P and K I .

[0021] The design parameters are defined as follows:

[0022] p 00 =0.6849;

[0023] p 10 =0.06876;

[0024] p 01 = -0.01671;

[0025] p 20 = -0.02953;

[0026] p 11 =0.01431;

[0027] p 02 = -0.001737.

[0028] The beneficial effects of this invention include:

[0029] (1) According to the high-maneuverability UAV control method based on tuned control parameters provided by the present invention, the method is based on a genetic algorithm, and by setting an appropriate fitness function, the corresponding gain coefficient can be obtained quickly and accurately, thereby shortening the UAV response time and improving the flight control effect;

[0030] (2) According to the high-maneuverability UAV control method based on tuned control parameters provided by the present invention, the controller gain coefficient of the roll channel is set in a targeted manner, so that the UAV can quickly and accurately obtain the actual roll angle based on the desired roll angle, thereby controlling the UAV propeller to work and better control the UAV flight attitude. Attached Figure Description

[0031] Figure 1 This paper shows a schematic diagram of the attitude control loop corresponding to the control method in this application;

[0032] Figure 2 This illustration shows a schematic diagram of the process of obtaining the controller gain coefficient using a genetic algorithm in this embodiment.

[0033] Figure 3 The tuning results of the controller gain coefficient in the embodiment are shown;

[0034] Figure 4 The step response diagrams corresponding to different values ​​of the controller gain coefficient in the embodiments and comparative examples are shown. Detailed Implementation

[0035] The present invention will be further described in detail below with reference to the accompanying drawings and embodiments. Through these descriptions, the features and advantages of the present invention will become clearer and more apparent.

[0036] The term “exemplary” as used herein means “serving as an example, embodiment, or illustration.” Any embodiment illustrated herein as “exemplary” is not necessarily to be construed as superior to or better than other embodiments. Although various aspects of embodiments are shown in the accompanying drawings, the drawings are not necessarily drawn to scale unless specifically indicated otherwise.

[0037] This application provides a control method for highly maneuverable unmanned aerial vehicles (UAVs) based on tuned control parameters, wherein, in this method, such as Figure 1 As shown, the desired roll angle is received in real time. Based on the deviation between the actual roll angle and the desired roll angle, the motor control command is obtained through the controller of the roll channel. The motor speed is controlled accordingly to respond to the desired roll angle, so that the actual roll angle of the UAV can quickly and accurately track the desired roll angle. Figure 1 In this context, H represents a stage with a gain of 1, and its value is 1.

[0038] Preferably, in this application, the UAV is a rotary-wing UAV; the motor speed represents the propeller speed; the actual roll angle can be measured in real time by sensors on the aircraft. In a preferred embodiment, the motor control command is obtained by the following formula (a):

[0039]

[0040] Where thr represents the motor control command;

[0041] φ represents the actual roll angle, which is obtained in real time through sensors on the drone;

[0042] φ d Indicates the expected roll angle;

[0043] K P and K I Each represents the controller gain coefficient independently;

[0044] p represents the roll angular velocity of the aircraft, which is measured by sensors on the drone.

[0045] s represents the Laplace transform operator.

[0046] Preferably, the controller gain coefficient K P and K IThe fitness function is obtained through a genetic algorithm, where the fitness function in the genetic algorithm is Equation (II).

[0047] t s =p 00 +p 10 ·K P +p 01 ·K I +p 20 ·K P 2 +p 11 ·K P ·K I +p 02 ·K I 2 (two)

[0048] Among them, t s Indicates the time required for adjustment response;

[0049] p 00 p 10 p 01 p 20 p 11 p 02 Each parameter represents a design parameter independently.

[0050] Preferably, the genetic algorithm is executed using MATLAB software. Based on the fitness function mentioned above, the toolbox provided in the software can be directly called to perform a series of operations such as population selection, crossover, and mutation using the default parameters in the toolbox. In order to simplify the processing flow and save time, the value range of the controller gain coefficient can be initially limited. For example, the gain coefficients of both controllers can be limited to 1 to 2, or further limited to 1 to 1.5 and 1 to 1.2.

[0051] In genetic algorithms, when the number of iterations reaches a preset number, such as 100, or the fitness reaches a set threshold, such as 0.7, the combination of individuals that minimizes the time required to adjust the response is the controller gain coefficient K output by the genetic algorithm. P and K I .

[0052] Preferably, the design parameters in the fitness function are:

[0053] p 00 =0.6849;

[0054] p 10 =0.06876;

[0055] p 01 = -0.01671;

[0056] p 20 = -0.02953;

[0057] p 11 =0.01431;

[0058] p 02 = -0.001737.

[0059] The inventors have discovered that by setting the controller gain coefficient of the aforementioned roll channel, the response time of the UAV can be minimized, thereby improving the flight quality of the UAV and ensuring that the actual flight attitude meets expectations. In this application, the tracking response for the desired pitch angle and desired z-axis acceleration on the UAV can be performed using methods similar to those described above, or methods already existing in the art can be employed.

[0060] Example

[0061] The desired roll angle is continuously generated and provided by a computer. The following method is used to track the response to the desired roll angle: based on the deviation between the actual roll angle and the desired roll angle, motor control commands are obtained through the controller of the roll channel, and the motor speed is controlled accordingly to respond to the desired roll angle and obtain the actual roll angle of the UAV.

[0062] The UAV obtains motor control commands in real time through the following formula (1), and controls the UAV motor to work and control the UAV flight status according to the motor control commands.

[0063]

[0064] K P and K I Each represents the controller gain coefficient independently;

[0065] p represents the roll angular velocity of the aircraft;

[0066] s represents the Laplace transform operator.

[0067] The controller gain coefficient K P and K I The fitness function is obtained through a genetic algorithm, which is executed by MATLAB software. The fitness function in the algorithm is given by equation (ii).

[0068] t s =p 00 +p 10 ·K P +p 01 ·K I +p 20 ·K P 2 +p11 ·K P ·K I +p 02 ·K I 2 (two)

[0070] The design parameter is p. 00 =0.6849, p 10 =0.06876, p 01 = -0.01671, p 20 = -0.02953, p 11 =0.01431, p 02 = -0.001737.

[0071] This MATLAB software toolbox is used to perform population selection, crossover, and mutation operations using default parameters. K is pre-set. P The value range of K is [1, 1.1]. I The value range is [1, 1.1].

[0072] The individual data output by this genetic algorithm is shown below.

[0073]

[0074]

[0075]

[0076] The time t required for the output regulation response s Specifically:

[0077]

[0078] The iterative output curve and parameter tuning results, such as Figure 2 and Figure 3 As shown in the image. Figure 2 The horizontal axis of the multiple data points represents the number of iterations in the genetic algorithm, characterizing the convergence of optimization as the number of generations increases; the vertical axis is the shortest adjustment time in each iteration, i.e., the best fitness in each iteration.

[0079] The results above show that the optimal combination of parameters for the output is K. P =1,K I =1.1, corresponding to t s The minimum value within the interval is 0.719388s.

[0080] After receiving the motor control command, the actual roll angle is obtained through simulation using the following formula (iii):

[0081]

[0082] Among them, the drone model G0 is obtained by the following formula (iv):

[0083]

[0084] The final roll angle step response diagram is as follows Figure 4 As shown by the black lines in the middle.

[0085] Comparative Example

[0086] The computer continuously generates and provides a desired roll angle that is completely consistent with the embodiment. The desired roll angle is tracked in response using the following method: based on the deviation between the actual roll angle and the desired roll angle, the controller of the roll channel obtains motor control commands, thereby controlling the motor speed to respond to the desired roll angle and obtain the actual roll angle of the UAV.

[0087] The UAV obtains motor control commands in real time through the following formula (1) and controls the UAV motor to work according to the actual roll angle, thereby controlling the UAV's flight status.

[0088]

[0089] K P and K I Each represents the controller gain coefficient independently;

[0090] p represents the roll angular velocity of the aircraft;

[0091] s represents the Laplace transform operator.

[0092] The controller gain coefficient K P and K I The value of can be divided into the following four cases:

[0093] Scenario 1 1.2 1.2 Scenario 2 1.4 1.4 Scenario 3 1.6 1.6 Scenario 4 1.8 1.8

[0094] For the four scenarios described above, after obtaining the motor control command, the actual roll angle is then simulated using the following formula (iii):

[0095]

[0096] Among them, the drone model G0 is obtained by the following formula (iv):

[0097]

[0098] The roll angle step response diagrams for the four final cases are as follows: Figure 4 As shown.

[0099] Figure 4 The vertical axis amplitude in the figure refers to the amplitude of the input signal. Figure 4 The signal in the middle is a step signal with a constant amplitude of 1, so the expected input is 1. Take a line parallel to the X-axis that intersects these curves. The x-coordinate of the intersection point corresponds to the time required to reach this y-value. As shown in the figure, the black line requires the shortest time.

[0100] based on Figure 4 It can be seen that the controller gain coefficient obtained by the method in the embodiment can better track the desired roll angle; the high-maneuverability UAV control method based on tuned control parameters provided in the embodiment can improve the speed of UAV attitude response and improve the UAV's tracking effect on the desired trajectory.

[0101] The present invention has been described above with reference to preferred embodiments; however, these embodiments are merely exemplary and illustrative. Various substitutions and modifications can be made to the present invention based on these embodiments, all of which fall within the scope of protection of the present invention.

Claims

1. A control method for highly maneuverable unmanned aerial vehicles (UAVs) based on tuned control parameters, characterized in that, In this method, The system receives the desired roll angle in real time. Based on the deviation between the actual roll angle and the desired roll angle, it obtains motor control commands through the controller of the roll channel, and controls the motor speed accordingly. This allows the system to respond to the desired roll angle, enabling the UAV's actual roll angle to quickly and accurately track the desired roll angle. The motor control command is obtained through the following formula (I): (one), in, Indicates motor control commands; Indicates the actual roll angle; Indicates the expected roll angle; and Each represents the controller gain coefficient independently; This indicates the roll angular velocity of the aircraft; Represents the Laplace transform operator. The controller gain coefficient and The fitness function is obtained through a genetic algorithm, where the fitness function in the genetic algorithm is Equation (II). (two), in, Indicates the time required for adjustment response; Each represents a design parameter independently. The design parameters are defined as follows: ； ； ； ； ； 。 2. The high-maneuverability unmanned aerial vehicle control method based on tuned control parameters according to claim 1, characterized in that, In genetic algorithms, when the number of iterations reaches a preset number or the fitness reaches a set threshold, the combination of individuals that minimizes the time required for adjustment response is the controller gain coefficient output by the genetic algorithm. and .

Images