CN111580550A - Unmanned aerial vehicle human-simulated intelligent control method - Google Patents

Unmanned aerial vehicle human-simulated intelligent control method Download PDF

Info

Publication number
CN111580550A
CN111580550A CN202010355519.7A CN202010355519A CN111580550A CN 111580550 A CN111580550 A CN 111580550A CN 202010355519 A CN202010355519 A CN 202010355519A CN 111580550 A CN111580550 A CN 111580550A
Authority
CN
China
Prior art keywords
control
layer
particle
follows
unmanned aerial
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202010355519.7A
Other languages
Chinese (zh)
Inventor
颜成钢
万斌
王廷宇
孙垚棋
张继勇
张勇东
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hangzhou Dianzi University
Original Assignee
Hangzhou Dianzi University
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hangzhou Dianzi University filed Critical Hangzhou Dianzi University
Priority to CN202010355519.7A priority Critical patent/CN111580550A/en
Publication of CN111580550A publication Critical patent/CN111580550A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
    • G05D1/10—Simultaneous control of position or course in three dimensions
    • G05D1/101—Simultaneous control of position or course in three dimensions specially adapted for aircraft
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B11/00—Automatic controllers
    • G05B11/01—Automatic controllers electric
    • G05B11/36—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential
    • G05B11/42—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential for obtaining a characteristic which is both proportional and time-dependent, e.g. P.I., P.I.D.
    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05D—SYSTEMS FOR CONTROLLING OR REGULATING NON-ELECTRIC VARIABLES
    • G05D1/00—Control of position, course or altitude of land, water, air, or space vehicles, e.g. automatic pilot
    • G05D1/08—Control of attitude, i.e. control of roll, pitch, or yaw
    • G05D1/0808—Control of attitude, i.e. control of roll, pitch, or yaw specially adapted for aircraft
    • G05D1/0816—Control of attitude, i.e. control of roll, pitch, or yaw specially adapted for aircraft to ensure stability
    • G05D1/0825—Control of attitude, i.e. control of roll, pitch, or yaw specially adapted for aircraft to ensure stability using mathematical models

Abstract

The invention provides an unmanned aerial vehicle humanoid intelligent control method. The method determines a flight mathematical model of the unmanned aerial vehicle; then, designing a human-simulated intelligent controller; the humanoid intelligent control comprises a three-layer structure: a control layer, a parameter correction layer and a task adaptation layer are operated; and finally, optimizing the differential coefficient of the comparative example by particle swarm optimization. The humanoid intelligent controller can determine a change strategy according to the deviation of the regulated quantity and the change trend of the deviation, and can effectively solve the problem of PID control parameter self-adaption of the quadrotor. However, the proportion, the differential coefficient and the attenuation coefficient in the humanoid intelligent control influence the final control effect, so the particle swarm optimization is adopted to carry out parameter optimization on the control effect. Therefore, the humanoid intelligent controller based on the particle swarm algorithm can realize the rapid response of the angle of the unmanned aerial vehicle, and the stability and the anti-interference performance are improved.

Description

Unmanned aerial vehicle human-simulated intelligent control method
Technical Field
The invention belongs to the field of flight control of unmanned aerial vehicles, and particularly relates to a human-simulated intelligent control system of an unmanned aerial vehicle, aiming at flight control of a single unmanned aerial vehicle.
Background
Nowadays, with the development of various technologies such as flight control, micro-electro-mechanical systems and the like, the four-rotor aircraft is more and more applied to the daily life of people. In the traditional four-rotor aircraft, the pitch angle (theta), the yaw angle (psi) and the roll angle (phi) of the four-rotor aircraft are controlled by adjusting the error between the input quantity and the feedback quantity through PID (proportion integration differentiation), so that the motion of an organism in the horizontal and vertical directions is completed. However, the traditional cascade PID adjusting process is too complicated and does not have self-parameter setting adaptive capacity, and the problems of large error and the like often occur in the four-rotor control, so that a novel four-rotor controller based on the combination of a particle swarm algorithm and humanoid intelligent control, namely the four-rotor humanoid intelligent controller based on the particle swarm algorithm, is designed.
Human-simulated Intelligent Control (HSIC) is a typical Intelligent Control mode, which is proposed by people of all ages in the 80 th century, and in recent years, a complete basic theoretical system and a perfect design method have been formed through the richness and development of numerous scholars such as professor of li zu jiangshu. The main idea of the human-simulated intelligent control is to further research and simulate the control behaviors of a human on the basis of macroscopic simulation of the control structure of the human, and identify and utilize characteristic information provided by the dynamic process of the system to the maximum extent. However, the proportion, the differential coefficient and the attenuation coefficient in the humanoid intelligent control influence the final control effect, so the particle swarm optimization is adopted to carry out parameter optimization on the control effect.
Particle Swarm Optimization (PSO) is an adaptive evolutionary computing technique based on population search proposed by Kennedy et al, which is similar to GA algorithm in computing method, but different from GA algorithm, PSO algorithm does not use factors such as hybridization and variation, but searches by simulating the population behavior of animal kingdoms such as bird foraging and herd. By combining the particle swarm algorithm with humanoid intelligent control, the problems of large steady-state error and flight attitude parameter self-adaption which often occur in a four-rotor aircraft can be effectively solved.
Disclosure of Invention
The invention aims to provide a human-simulated intelligent control method for an unmanned aerial vehicle, aiming at the problem of flight control of a single unmanned aerial vehicle. The system of the invention divides the controller into three layers: a direct control layer, a parameter correction layer and a task adaptation layer.
The technical scheme adopted by the invention for solving the technical problem specifically comprises the following steps:
step 1, determining a flight mathematical model of an unmanned aerial vehicle;
unmanned aerial vehicle organism motion model
The four-rotor aircraft adjusts a pitch angle (theta), a yaw angle (psi) and a roll angle (phi) by controlling the rotating speed of the four rotors so as to complete forward and backward movement, ascending and descending, left and right flying and a series of flying movements of the aircraft body.
First, two coordinate systems are established: a body coordinate system and a conventional coordinate system.
The conventional coordinate system e (oexyz) is stationary with respect to the earth's surface.
The machine body coordinate system A (oxyz) is coincident with the center of mass of the machine body, the horizontal axis ox points to the first motor, the longitudinal axis oy points to the fourth motor, and the oz is perpendicular to the oxyy surface.
Euler angles of four-rotor aircraft:
(1) roll angle (Φ): the machine body rotates around the ox shaft by an angle.
(2) Pitch angle (θ): the body rotates around the oy axis.
(3) Yaw angle (ψ): and the body rotates the included angle between the projection of the longitudinal axis of the aircraft around the oz axis in the horizontal plane and the axis of the inertial coordinate system OX.
Unmanned aerial vehicle dynamics model
According to attitude vectors [ phi, psi, theta ] (roll angle, yaw angle, pitch angle) and displacement vectors [ x, y, z ] of the four rotors, a kinetic equation is as follows:
Figure BDA0002473350140000021
Figure BDA0002473350140000022
Figure BDA0002473350140000023
Figure BDA0002473350140000024
Figure BDA0002473350140000031
Figure BDA0002473350140000032
wherein:
Figure BDA0002473350140000033
Figure BDA0002473350140000034
Figure BDA0002473350140000035
Figure BDA0002473350140000036
u1representing the total lift of four rotors, u2Representing roll moment, u3Representing the pitching moment, u4Representing yaw moment, w1、w2、w3、w4Respectively representing the rotational speeds of four rotors, Ix、Iy、IzRepresenting the moment of inertia of the body in the xyz axis.
Step 2, designing a human-simulated intelligent controller;
the humanoid intelligent control comprises a three-layer structure: the system comprises a running control layer, a parameter correction layer and a task adaptation layer.
The method is characterized in that:
(1) a hierarchical information processing and decision mechanism;
(2) online feature identification and feature memory;
(3) open-closed loop control is combined;
the prototype algorithm of the humanoid intelligent control is as follows:
Figure BDA0002473350140000037
μ — output of the controller;
e，
Figure BDA0002473350140000038
-representing the error and the rate of change of the error in the complex process, respectively;
em，i-the extreme value of the error i;
Kp-proportional gain of the controller;
k is the suppression coefficient;
the human-simulated intelligent controller adjusts control strategies by detecting errors between expected values and measured values of attitude angles of the four-rotor aircraft in the flight process and the change rate of the errors, and the basic control strategies are respectively as follows: bang-bang control, proportional derivative control. And the controller is divided into seven areas through threshold correction.
Operation control layer
The run control layer faces the real-time control problem, e,
Figure BDA0002473350140000042
Respectively representing the errors of attitude angles during flight andthe rate of change of the error.
The specific control strategy is as follows:
(1) when the deviation is overlarge, the corresponding region (i) adopts the control action as large as possible, such as bang-bang control.
(2) And under the condition of small deviation and deviation change rate (meeting the error requirement), the corresponding region II adopts hold mode control.
(3) Except the region (c), the other regions (c), (c) and (c) are all controlled by a proportional differential mode, but the comparative examples and differential parameters are required to be adjusted in different regions.
(4) And in a situation area with overlarge deviation change rate, introducing bang-bang control based on the deviation change rate.
Thus, let the feature set element set of the run control layer:
Q1＝{q1，q2，q3，q4，q5，q6，q7，q8} (14)
wherein:
q1＝{|en|≥e11} q2＝{|en|≥e14}
q3＝{|en|≥e12} q4＝{|en|≥e13}
Figure BDA0002473350140000041
the model of the run control layer is:
Φ1＝{Φ11，Φ12，Φ13，Φ14} (15)
Φ2＝{Φ131，Φ132，Φ133}
Φ11＝{q2}
Figure BDA0002473350140000051
Figure BDA0002473350140000052
Figure BDA0002473350140000053
Figure BDA0002473350140000054
Φ13＝{Φ131∪Φ132∪Φ133}
Figure BDA0002473350140000055
the control mode set of the operation control layer is as follows:
Ψ1＝{Ψ11，Ψ12，Ψ13，Ψ14} (16)
wherein:
Ψ11：{un＝sign(en)·Umax}
Ψ12：{un＝un-1}
Figure BDA0002473350140000056
Figure BDA0002473350140000057
wherein u isnIs output by the controller; u shapemaxIs the maximum value output by the controller; e, the number of the first and second groups,
Figure BDA0002473350140000058
respectively representing errors and the change rate of the errors; omegapωdProportional coefficient and differential coefficient respectively; k is the attenuation coefficient.
The inference rule set of the operation control layer is as follows:
Ω1＝{ω11，ω12，ω13，ω14} (17)
wherein:
ω11：Φ11→Ψ11
ω12：Φ12→Ψ12
ω13：Φ13→Ψ13
ω14：Φ14→Ψ14
parameter correcting layer
After simplification, the feature model of the parameter correction layer is substantially the same as the feature model of the run control layer.
Let the feature set element set of the parameter correction layer be:
Q2＝{q1，q2，q3，q4，q5，q6，q7，q8} (18)
wherein:
q1＝{|en|≥e21} q2＝{|en|≥e24}
q3＝{|en|≥e22} q4＝{|en|≥e23}
Figure BDA0002473350140000061
Figure BDA0002473350140000062
the model of the parameter correction layer is:
Φ1＝{Φ11，Φ12，Φ13，Φ14} (19)
Φ2＝{Φ131，Φ132，Φ133}
Φ11＝{q2}
Figure BDA0002473350140000063
Figure BDA0002473350140000064
Figure BDA0002473350140000065
Figure BDA0002473350140000066
Φ13＝{Φ131∪Φ132∪Φ133}
Figure BDA0002473350140000071
the decision mode set of the parameter correction layer is as follows:
Ψ2＝{Ψ21，Ψ22，Ψ23} (20)
Ψ21＝{NULL}
Ψ22＝{ωp＝ωp1，ωd＝ωd1，k＝k1}
Ψ23＝{ωp＝ωp2，ωd＝ωd2，k＝k2}
the inference rule set of the parameter correction layer is as follows:
Ω2＝{ω21，ω22，ω23，ω24} (21)
wherein:
ω21：Φ21→Ψ21
ω22：Φ22→Ψ21
ω23：Φ23→Ψ22
ω24：Φ24→Ψ23
task adaptation layer
Let the feature set element set of the task adaptation layer be:
Q3＝{q1，q2，q3，q4，q5，q6，q7，q8} (22)
wherein:
q1＝{|en|≥e31} q2＝{|en|≥e34}
q3＝{|en|≥e32} q4＝{|en|≥e33}
Figure BDA0002473350140000072
Figure BDA0002473350140000073
the model of the task adaptation layer is:
Φ1＝{Φ11，Φ12，Φ13，Φ14} (23)
Φ2＝{Φ131，Φ132，Φ133}
Φ11＝{q2}
Figure BDA0002473350140000081
Figure BDA0002473350140000082
Figure BDA0002473350140000083
Figure BDA0002473350140000084
Φ13＝{Φ131∪Φ132∪Φ133}
Figure BDA0002473350140000085
the decision mode set of the task adaptation layer is as follows:
Ψ3＝{Ψ31，Ψ32，Ψ33} (24)
Ψ31＝{NULL}
Ψ32＝{ωp＝ωp1，ωd＝ωd1，k＝k1}
Ψ33＝{ωp＝ωp2，ωd＝ωd2，k＝k2}
the inference rule set of the task adaptation layer is as follows:
Ω3＝{ω31，ω32，ω33，ω34} (25)
wherein:
ω31：Φ31→Ψ31
ω32：Φ32→Ψ31
ω33：Φ33→Ψ32
ω34：Φ34→Ψ33
step 3, optimizing differential coefficients of comparative examples by particle swarm optimization;
the particle swarm algorithm is an intelligent algorithm based on a population, each member in the population is called a particle and represents a feasible solution, and the position of food is a globally optimal solution. The group searches for an optimal solution in a D-dimensional space, each particle has an adaptive value and a speed to adjust the flight direction of the particle, and the algorithm implementation process of the position of the particle close to food is as follows through continuously updating a global optimal position (gbest) and an individual optimal position (pbest):
(1) initializing particlesGroup, including the size of the group and the position x of each particleiAnd velocity vi。
(2) The fitness value F [ i ] of each particle is calculated.
(3) For each particle, the fitness value F [ i ] of the particle is compared with the individual extremum pbest [ i ], and when F [ i ] is greater than pbest [ i ], gbest is replaced by F [ i ].
(4) For each particle, its fitness value F [ i ] is compared to the global extremum gbest, and if F [ i ] is > pbest [ i ], gbest is replaced by F [ i ].
(5) According to the formula:
Vi(t+1)＝w·Vi(t)+c1·r1(Pi(t)-Xi(t))+c2·r2tPg(t)-Xi(t)) (26)
Xi(t+1)＝Xi(t)+Vi(t+1) (27)
w is the inertial weight, r1And r2Is [0, 1 ]]The random number of the interval, t is the iteration number. Updating the position x of the particle by equations (26) (27)iAnd velocity vi。
The invention has the following beneficial effects:
advantages of the invention
The humanoid intelligent controller can determine a change strategy according to the deviation of the regulated quantity and the change trend of the deviation, and can effectively solve the problem of PID control parameter self-adaption of the quadrotor. However, the proportion, the differential coefficient and the attenuation coefficient in the humanoid intelligent control influence the final control effect, so the particle swarm optimization is adopted to carry out parameter optimization on the control effect. Therefore, the humanoid intelligent controller based on the particle swarm algorithm can realize the rapid response of the angle of the unmanned aerial vehicle, and the stability and the anti-interference performance are improved.
Drawings
Figure 1 is a quad-rotor drone coordinate system;
fig. 2 is a quad-rotor drone euler angle;
FIG. 3 is a human-simulated controller feature model;
FIG. 4 is a chart of attitude angle response;
FIG. 5 is a stability graph;
FIG. 6 is a graph of interference rejection curves;
detailed description of the invention
The present invention will be described in detail with reference to specific embodiments.
The human-simulated controller method provided by the invention is implemented according to the following steps:
step 1, determining a flight mathematic model of an unmanned aerial vehicle
Unmanned aerial vehicle organism motion model
The four-rotor aircraft adjusts a pitch angle (theta), a yaw angle (psi) and a roll angle (phi) by controlling the rotating speed of the four rotors so as to complete a series of flight motions of the aircraft body, such as forward and backward movement, ascending and descending, left and right flight and the like.
First, two coordinate systems are established: a body coordinate system and a conventional coordinate system. As shown in figure 1
The conventional coordinate system e (oexyz) is stationary with respect to the earth's surface.
The machine body coordinate system A (oxyz) is coincident with the center of mass of the machine body, the horizontal axis ox points to the first motor, the longitudinal axis oy points to the fourth motor, and the oz is perpendicular to the oxyy surface.
The Euler angles of the four-rotor aircraft are as shown in the attached figure 2:
(1) roll angle (Φ): the machine body rotates around the ox shaft by an angle.
(2) Pitch angle (θ): the body rotates around the oy axis.
(3) Yaw angle (ψ): and the body rotates the included angle between the projection of the longitudinal axis of the aircraft around the oz axis in the horizontal plane and the axis of the inertial coordinate system OX.
Unmanned aerial vehicle dynamics model
The four-rotor aircraft dynamic model is the basis for system design and implementation. According to attitude vectors [ phi, psi, theta ] (roll angle, yaw angle, pitch angle) and displacement vectors [ x, y, z ] of the four rotors, a kinetic equation is as follows:
Figure BDA0002473350140000111
Figure BDA0002473350140000112
Figure BDA0002473350140000113
Figure BDA0002473350140000114
Figure BDA0002473350140000115
Figure BDA0002473350140000116
wherein:
Figure BDA0002473350140000117
Figure BDA0002473350140000118
Figure BDA0002473350140000119
Figure BDA00024733501400001110
u1representing the total lift of four rotors, u2Representing roll moment, u3Representing the pitching moment, u4Representing yaw moment, w1、w2、w3、w4Respectively representing the rotational speeds of four rotors, Ix、Iy、IzRepresenting the moment of inertia of the body in the xyz axis.
Step 2, designing a human-simulated intelligent controller
Human-simulated Intelligent Control (HSIC) is a typical Intelligent Control mode, which is proposed by people of all ages in the 80 th century, and in recent years, a complete basic theoretical system and a perfect design method have been formed through the richness and development of numerous scholars such as professor of li zu jiangshu. The main idea of the human-simulated intelligent control is to further research and simulate the control behavior of a human on the basis of macroscopic simulation of the control structure of the human, and to identify and utilize characteristic information provided by the dynamic process of the system to the maximum extent. The control algorithm is based on the modeling of intelligent behaviors such as observation, memory, decision and the like of a person on a control object, and determines a change strategy according to the deviation of the regulated quantity and the change trend of the deviation.
The humanoid intelligent control comprises a three-layer structure: the system comprises a running control layer, a parameter correction layer and a task adaptation layer.
The method is characterized in that:
(1) a hierarchical information processing and decision mechanism;
(2) online feature identification and feature memory;
(3) open-closed loop control is combined;
the prototype algorithm of the humanoid intelligent control is as follows:
Figure BDA0002473350140000121
μ — output of the controller;
e，
Figure BDA0002473350140000122
-representing the error and the rate of change of the error in the complex process, respectively;
em，i-the extreme value of the error i;
Kp-proportional gain of the controller;
k is the suppression coefficient;
the human-simulated intelligent controller adjusts control strategies by detecting errors between expected values and measured values of attitude angles of the four-rotor aircraft in the flight process and the change rate of the errors, and the basic control strategies are respectively as follows: bang-bang control, proportional derivative control. And the controller is divided into seven areas through threshold correction.
Operation control layer
The run control layer faced the real-time control problem and was characterized by the model shown in FIG. 3, e,
Figure BDA0002473350140000123
Respectively representing the error of the attitude angle and the change rate of the error in the flight process.
The specific control strategy is as follows:
(1) when the deviation is overlarge, the corresponding region (i) adopts the control action as large as possible, such as bang-bang control.
(2) And under the condition of small deviation and deviation change rate (meeting the error requirement), the corresponding region II adopts hold mode control.
(3) Except the region (c), the other regions (c), (c) and (c) are all controlled by a proportional differential mode, but the comparative examples and differential parameters are required to be adjusted in different regions.
(4) And in a situation area with overlarge deviation change rate, introducing bang-bang control based on the deviation change rate.
Thus, let the feature set element set of the run control layer:
Q1＝{q1，q2，q3，q4，q5，q6，q7，q8} (14)
wherein:
q1＝{|en|≥e11} q2＝{|qn|≥e14}
q3＝{|en|≥e12} q4＝{|en|≥e13}
Figure BDA0002473350140000131
Figure BDA0002473350140000132
the model of the run control layer is:
Φ1＝{Φ11，Φ12，Φ13，Φ14} (15)
Φ2＝{Φ131，Φ132，Φ133}
Φ11＝{q2}
Figure BDA0002473350140000133
Figure BDA0002473350140000134
Figure BDA0002473350140000135
Figure BDA0002473350140000136
Φ13＝{Φ131∪Φ132∪Φ133}
Figure BDA0002473350140000137
the control mode set of the operation control layer is as follows:
Ψ1＝{Ψ11，Ψ12，Ψ13，Ψ14} (16)
wherein:
Ψ11：{un＝sign(en)·Umax}
Ψ12：{un＝un-1}
Figure BDA0002473350140000141
Figure BDA0002473350140000142
wherein u isnIs output by the controller; u shapemaxIs the maximum value output by the controller; e, the number of the first and second groups,
Figure BDA0002473350140000143
respectively representing errors and the change rate of the errors; omegapωdProportional coefficient and differential coefficient respectively; k is the attenuation coefficient.
The inference rule set of the operation control layer is as follows:
Ω1＝{ω11，ω12，ω13，ω14}(17)
wherein:
ω11：Φ11→Ψ11
ω12：Φ12→Ψ12
ω13：Φ13→Ψ13
ω14：Φ14→Ψ14
parameter correcting layer
After simplification, the feature model of the parameter correction layer is substantially the same as the feature model of the run control layer.
Let the feature set element set of the parameter correction layer be:
Q2＝{q1，q2，q3，q4，q5，q6，q7，q8} (18)
wherein:
q1＝{|en|≥e21} q2＝{|en|≥e24}
q3＝{|en|≥e22} q4＝{|en|≥e23}
Figure BDA0002473350140000151
Figure BDA0002473350140000152
the model of the parameter correction layer is:
Φ1＝{Φ11，Φ12，Φ13，Φ14} (19)
Φ2＝{Φ131，Φ132，Φ133}
Φ11＝{q2}
Figure BDA0002473350140000153
Figure BDA0002473350140000154
Figure BDA0002473350140000155
Figure BDA0002473350140000156
Φ13＝{Φ131∪Φ132∪Φ133}
Figure BDA0002473350140000157
the decision mode set of the parameter correction layer is as follows:
Ψ2＝{Ψ21，Ψ22，Ψ23} (20)
Ψ21＝{NULL}
Ψ22＝{ωp＝ωp1，ωd＝ωd1，k＝k1}
Ψ23＝{ωp＝ωp2，ωd＝ωd2，k＝k2}
the inference rule set of the parameter correction layer is as follows:
Ω2＝{ω21，ω22，ω23，ω24} (21)
wherein:
ω21：Φ21→Ψ21
ω22：Φ22→Ψ21
ω23：Φ23→Ψ22
ω24：Φ24→Ψ23
task adaptation layer
Let the feature set element set of the task adaptation layer be:
Q3＝{q1，q2，q3，q4，q5，q6，q7，q8} (22)
wherein:
q1＝{|en|≥e31} q2＝{|en|≥e34}
q3＝{|en|≥e32} q4＝{|en|≥e33}
Figure BDA0002473350140000161
Figure BDA0002473350140000162
the model of the task adaptation layer is:
Φ1＝{Φ11，Φ12，Φ13，Φ14} (23)
Φ2＝{Φ131，Φ132，Φ133}
Φ11＝{q2}
Figure BDA0002473350140000163
Figure BDA0002473350140000164
Figure BDA0002473350140000165
Figure BDA0002473350140000166
Φ13＝{Φ131∪Φ132∪Φ133}
Figure BDA0002473350140000167
the decision mode set of the task adaptation layer is as follows:
Ψ3＝{Ψ31，Ψ32，Ψ33} (24)
Ψ31＝{NULL}
Ψ32＝{ωp＝ωp1，ωd＝ωd1，k＝k1}
Ψ33＝{ωp＝ωp2，ωd＝ωd2，k＝k2}
the inference rule set of the task adaptation layer is as follows:
Ω3＝{ω31，ω32，ω33，ω34} (25)
wherein:
ω31：Φ31→Ψ31
ω32：Φ32→Ψ31
ω33：Φ33→Ψ32
ω34：Φ34→Ψ33
step 3, optimizing differential coefficients of comparative examples by particle swarm optimization;
when attitude angle simulation is carried out on the four-rotor aircraft, the fact that when errors and error change rates are large, proportion of proportion and differentiation in the process of control is automatically distributed by using a weight formula, the control effect is good, and when the errors and the error change rates are small, control parameters need to be adjusted. Therefore, Particle Swarm Optimization (PSO) was used for optimization.
The particle swarm algorithm is an intelligent algorithm based on a population, each member in the population is called a particle and represents a feasible solution, and the position of food is a globally optimal solution. The group searches for an optimal solution in a D-dimensional space, each particle has an adaptive value and a speed to adjust the flight direction of the particle, and the algorithm implementation process of the position of the particle close to food is as follows through continuously updating a global optimal position (gbest) and an individual optimal position (pbest):
(1) initializing a population of particles, including the size of the population and the location x of each particleiAnd velocity vi。
(2) The fitness value F [ i ] of each particle is calculated.
(3) For each particle, the fitness value F [ i ] of the particle is compared with the individual extremum pbest [ i ], and when F [ i ] is greater than pbest [ i ], gbest is replaced by F [ i ].
(4) For each particle, its fitness value F [ i ] is compared to the global extremum gbest, and if F [ i ] is > pbest [ i ], gbest is replaced by F [ i ].
(5) According to the formula:
Vi(t+1)＝w·Vi(t)+c1·r1(Pi(t)-Xi(t))+c2·r2(Pg(t)-Xi(t)) (26)
Xi(t+1)＝Xi(t)+Vi(t+1) (27)
w is the inertial weight, r1And r2Is [0, 1 ]]The random number of the interval, t is the iteration number. Updating the position x of the particle by equations (26) (27)iSpeed of mixingDegree vi。
In order to verify the effectiveness of the particle swarm algorithm, an initial population is set to be 30, a search space is a 3-dimensional space with 3 parameters to be optimized, and the iteration number is 100.
The PSO tuning parameters after 100 operations are shown in Table 2.
TABLE 2 Final optimization parameters
Figure BDA0002473350140000181
The optimization program comprises the following steps:
VStep(j,:)＝w*VStep(j,:)+c1*rand*(pbest(j,:)-Swarm(j,:))+c2*rand*(gbest-Swarm(j,:))；
If VStep(j,:)>Vmax,VStep(j,:)＝Vmax；end
IfVStep(j,:)<Vmin,VStep(j,:)＝Vmin；end
Swarm(j,:)＝Swarm(j,:)+VStep(j,:)；
for k＝1:Dim
if Swarm(j,k)>Ub(k),Swarm(j,k)＝Ub(k)；end
if Swarm(j,k)<Lb(k),Swarm(j,k)＝Lb(k)；end
end
wherein Ub (k), Lb (k) are upper and lower limits of the particle group. Then, calculating a fitness value fSwarm (j) of the particle according to the updated particle, and then comparing:
if fSwarm(j)<fpbest(j)
pbest(j,:)＝Swarm(j,:)；
fpbest(j)＝fSwarm(j)；
end
if fSwarm(j)<fgbest
gbest＝Swarm(j,:)；
fgbest＝fSwarm(j)；
end
and finally, obtaining the optimal control parameters by continuously updating the particles.
Examples
The modeling of the four-rotor fuselage is completed through a four-rotor dynamic model and a mathematical model, and a simulation model is built under the simulink environment
Angular response
The roll angle is taken as an example in the experiment. Assuming that the initial angle is 0 degree, the 30 degrees are respectively used as the input quantities of the particle swarm quadrotor humanoid controller and the cascade PID controller to observe whether the quadrotor controller designed in the text can quickly return to the balance state. The simulation results are shown in figure 4.
Stability test
And verifying whether the four rotors can complete stable control or not again, and observing whether the four rotors can return to a stable state or not by taking a 30-degree signal as an input quantity of an attitude angle as shown in the figure 5.
Anti-interference experiment
In the four-rotor flight process, the anti-interference performance is extremely important, and the four-rotor aircraft can stably fly after encountering interference in the air. The step signal with the input amount of 20 is taken as the interference signal at 0.1 second. The simulation results are shown in FIG. 6.

Claims (4)

1. An unmanned aerial vehicle humanoid intelligent control method is characterized by comprising the following steps:
step 1, determining a flight mathematical model of an unmanned aerial vehicle;
step 2, designing a human-simulated intelligent controller;
and 3, optimizing the differential coefficient of the comparative example by particle swarm optimization.
2. The unmanned aerial vehicle humanoid intelligent control method according to claim 1, wherein step 1 determines a flight mathematical model of the unmanned aerial vehicle, and specifically operates as follows;
unmanned aerial vehicle organism motion model
The four-rotor aircraft adjusts a pitch angle (theta), a yaw angle (psi) and a roll angle (phi) by controlling the rotating speed of four rotors so as to complete forward and backward movement, ascending and descending, left and right flying and a series of flying movements of the aircraft body;
first, two coordinate systems are established: a body coordinate system, a conventional coordinate system;
the conventional coordinate system e (oexyz) is stationary with respect to the earth's surface;
a machine body coordinate system A (oxyz) is superposed with the mass center of the machine body, a transverse axis ox points to a first motor, a longitudinal axis oy points to a fourth motor, and the oz is vertical to an oxy surface;
euler angles of four-rotor aircraft:
(1) roll angle (Φ): the machine body rotates around the ox shaft by an angle;
(2) pitch angle (θ): the machine body rotates around the oy axis by an angle;
(3) yaw angle (ψ): the aircraft body rotates around the oz axis to form an included angle between the projection of the longitudinal axis of the aircraft in the horizontal plane and the axis of an inertial coordinate system OX;
unmanned aerial vehicle dynamics model
According to attitude vectors [ phi, psi, theta ] (roll angle, yaw angle, pitch angle) and displacement vectors [ x, y, z ] of the four rotors, a kinetic equation is as follows:
Figure FDA0002473350130000011
Figure FDA0002473350130000012
Figure FDA0002473350130000013
Figure FDA0002473350130000014
Figure FDA0002473350130000015
Figure FDA0002473350130000021
wherein:
Figure FDA0002473350130000022
Figure FDA0002473350130000023
Figure FDA0002473350130000024
Figure FDA0002473350130000025
u1representing the total lift of four rotors, u2Representing roll moment, u3Representing the pitching moment, u4Representing yaw moment, w1、w2、w3、w4Respectively representing the rotational speeds of four rotors, Ix、Iy、IzRepresenting the moment of inertia of the body in the xyz axis.
3. The unmanned aerial vehicle humanoid intelligent control method according to claim 2, wherein the step 2 humanoid intelligent controller is designed and specifically operated as follows;
the humanoid intelligent control comprises a three-layer structure: a control layer, a parameter correction layer and a task adaptation layer are operated;
the method is characterized in that:
(1) a hierarchical information processing and decision mechanism;
(2) online feature identification and feature memory;
(3) open-closed loop control is combined;
the prototype algorithm of the humanoid intelligent control is as follows:
Figure FDA0002473350130000026
μ — output of the controller;
e，
Figure FDA0002473350130000027
-representing the error and the rate of change of the error in the complex process, respectively;
em，i-the extreme value of the error i;
Kp-proportional gain of the controller;
k is the suppression coefficient;
the human-simulated intelligent controller adjusts control strategies by detecting errors between expected values and measured values of attitude angles of the four-rotor aircraft in the flight process and the change rate of the errors, and the basic control strategies are respectively as follows: bang-bang control, proportional differential control; dividing the controller into a fifth area, a sixth area and a seventh area through threshold correction;
operation control layer
The run control layer faces the real-time control problem, e,
Figure FDA0002473350130000031
Respectively representing the error of the attitude angle and the change rate of the error in the flight process;
the specific control strategy is as follows:
(1) when the deviation is overlarge, the corresponding region (i) adopts the control action as large as possible, such as bang-bang control;
(2) under the condition of small deviation and deviation change rate (meeting the error requirement), the corresponding region II adopts the mode-keeping control;
(3) except the region I, (-) and the other regions III, (-) and III are all controlled by adopting a proportional differential mode, but the comparative example and the differential parameter are required to be adjusted in different regions;
(4) introducing bang-bang control based on the deviation change rate in a condition area with the excessive deviation change rate;
thus, let the feature set element set of the run control layer:
Q1＝{q1，q2，q3，q4，q5，q6，q7，q8} (14)
wherein:
q1＝{|en|≥e11} q2＝{|en|≥e14}
q3＝{|en|≥e12} q4＝{|en|≥e13}
Figure FDA0002473350130000032
Figure FDA0002473350130000033
the model of the run control layer is:
Φ1＝{Φ11，Φ12，Φ13，Φ14} (15)
Φ2＝{Φ131，Φ132，Φ133}
Φ11＝{q2}
Figure FDA0002473350130000041
Figure FDA0002473350130000042
Figure FDA0002473350130000043
Figure FDA0002473350130000044
Φ13＝{Φ131∪Φ132∪Φ133}
Figure FDA0002473350130000045
the control mode set of the operation control layer is as follows:
Ψ1＝{Ψ11，Ψ12，Ψ13，Ψ14} (16)
wherein:
Ψ11：{un＝sign(en)·Umax}
Ψ12：{un＝un-1}
Ψ13：
Figure FDA0002473350130000046
Ψ14：
Figure FDA0002473350130000047
wherein u isnIs output by the controller; u shapemaxIs the maximum value output by the controller; e, the number of the first and second groups,
Figure FDA0002473350130000048
respectively representing errors and the change rate of the errors; omegapωdProportional coefficient and differential coefficient respectively; k is an attenuation coefficient;
the inference rule set of the operation control layer is as follows:
Ω1＝{ω11，ω12，ω13，ω14} (17)
wherein:
ω11：Φ11→Ψ11
ω12：Φ12→Ψ12
ω13：Φ13→Ψ13
ω14：Φ14→Ψ14
parameter correcting layer
After simplification, the characteristic model of the parameter correction layer is basically the same as that of the operation control layer;
let the feature set element set of the parameter correction layer be:
Q2＝{q1，q2，q3，q4，q5，q6，q7，q8} (18)
wherein:
q1＝{|en|≥e21} q2＝{|en|≥e24}
q3＝{|en|≥e22} q4＝{|en|≥e23}
Figure FDA0002473350130000051
Figure FDA0002473350130000052
the model of the parameter correction layer is:
Φ1＝{Φ11，Φ12，Φ13，Φ14} (19)
Φ2＝{Φ131，Φ132，Φ133}
Φ11＝{q2}
Figure FDA0002473350130000053
Figure FDA0002473350130000054
Figure FDA0002473350130000055
Figure FDA0002473350130000056
Φ13＝{Φ131∪Φ132∪Φ133}
Figure FDA0002473350130000061
the decision mode set of the parameter correction layer is as follows:
Ψ2＝{Ψ21，Ψ22，Ψ23} (20)
Ψ21＝{NULL}
Ψ22＝{ωp＝ωp1，ωd＝ωd1，k＝k1}
Ψ23＝{ωp＝ωp2，ωd＝ωd2，k＝k2}
the inference rule set of the parameter correction layer is as follows:
Ω2＝{ω21，ω22，ω23，ω24} (21)
wherein:
ω21：Φ21→Ψ21
ω22：Φ22→Ψ21
ω23：Φ23→Ψ22
ω24：Φ24→Ψ23
task adaptation layer
Let the feature set element set of the task adaptation layer be:
Q3＝{q1，q2，q3，q4，q5，q6，q7，q8} (22)
wherein:
q1＝{|en|≥e31} q2＝{|en|≥e34}
q3＝{|en|≥e32} q4＝{|en|≥e33}
Figure FDA0002473350130000062
Figure FDA0002473350130000063
the model of the task adaptation layer is:
Φ1＝{Φ11，Φ12，Φ13，Φ14} (23)
Φ2＝{Φ131，Φ132，Φ133}
Φ11＝{q2}
Figure FDA0002473350130000071
Figure FDA0002473350130000072
Figure FDA0002473350130000073
Figure FDA0002473350130000074
Φ13＝{Φ131∪Φ132∪Φ133}
Figure FDA0002473350130000075
the decision mode set of the task adaptation layer is as follows:
Ψ3＝{Ψ31，Ψ32，Ψ33} (24)
Ψ31＝{NULL}
Ψ32＝{ωp＝ωp1，ωd＝ωd1，k＝k1}
Ψ33＝{ωp＝ωp2，ωd＝ωd2，k＝k2}
the inference rule set of the task adaptation layer is as follows:
Ω3＝{ω31，ω32，ω33，ω34} (25)
wherein:
ω31：Φ31→Ψ31
ω32：Φ32→Ψ31
ω33：Φ33→Ψ32
ω34：Φ34→Ψ33。
4. the unmanned aerial vehicle humanoid intelligent control method according to claim 3, wherein step 3 is optimized by adopting particle swarm optimization to compare differential coefficients, and the specific operation is as follows;
the particle swarm algorithm is an intelligent algorithm based on a population, each member in the population is called a particle and represents a feasible solution, and the position of food is a global optimal solution; the group searches for an optimal solution in a D-dimensional space, each particle has an adaptive value and a speed to adjust the flight direction of the particle, and the algorithm implementation process of the position of the particle close to food is as follows through continuously updating a global optimal position (gbest) and an individual optimal position (pbest):
(1) initializing a population of particles, including the size of the population and the location x of each particleiAnd velocity vi；
(2) Calculating a fitness value F [ i ] of each particle;
(3) for each particle, comparing the fitness value F [ i ] of the particle with the individual extreme value pbest [ i ], and replacing gbest with F [ i ] when F [ i ] is greater than pbest [ i ];
(4) for each particle, comparing its fitness value F [ i ] with the global extremum gbest, and if F [ i ] is > pbest [ i ], replacing gbest with F [ i ];
(5) according to the formula:
Vi(t+1)＝w·Vi(t)+c1·r1(Pi(t)-Xi(t))+c2·r2(Pg(t)-Xi(t)) (26)
Xi(t+1)＝Xi(t)+Vi(t+1) (27)
w is the inertial weight, r1And r2Is [0, 1 ]]Random number of interval, t is iteration number; updating the position x of the particle by equations (26) (27)iAnd velocity vi。
CN202010355519.7A 2020-04-29 2020-04-29 Unmanned aerial vehicle human-simulated intelligent control method Pending CN111580550A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202010355519.7A CN111580550A (en) 2020-04-29 2020-04-29 Unmanned aerial vehicle human-simulated intelligent control method

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202010355519.7A CN111580550A (en) 2020-04-29 2020-04-29 Unmanned aerial vehicle human-simulated intelligent control method

Publications (1)

Publication Number Publication Date
CN111580550A true CN111580550A (en) 2020-08-25

Family

ID=72113216

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202010355519.7A Pending CN111580550A (en) 2020-04-29 2020-04-29 Unmanned aerial vehicle human-simulated intelligent control method

Country Status (1)

Country Link
CN (1) CN111580550A (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112016162A (en) * 2020-09-05 2020-12-01 江西理工大学 Four-rotor unmanned aerial vehicle PID controller parameter optimization method

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1901153A1 (en) * 2006-09-12 2008-03-19 OFFIS e.V. Control system for unmanned 4-rotor-helicopter
CN101551642A (en) * 2009-04-08 2009-10-07 南京航空航天大学 Improved particle swarm algorithm for automatic optimization of control law parameters of unmanned aircraft
CN103853050A (en) * 2014-03-19 2014-06-11 湖北蔚蓝国际航空学校有限公司 PID optimization control method of four-rotor aircraft
CN104765272A (en) * 2014-03-05 2015-07-08 北京航空航天大学 Four-rotor aircraft control method based on PID neural network (PIDNN) control
WO2017037354A1 (en) * 2015-09-01 2017-03-09 Renault S.A.S Device and method for controlling a driver assistance system for correcting the direction of travel of a motor vehicle
CN110531779A (en) * 2019-08-02 2019-12-03 杭州电子科技大学 A kind of unmanned plane apery collaboration flight controller and its implementation

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP1901153A1 (en) * 2006-09-12 2008-03-19 OFFIS e.V. Control system for unmanned 4-rotor-helicopter
CN101551642A (en) * 2009-04-08 2009-10-07 南京航空航天大学 Improved particle swarm algorithm for automatic optimization of control law parameters of unmanned aircraft
CN104765272A (en) * 2014-03-05 2015-07-08 北京航空航天大学 Four-rotor aircraft control method based on PID neural network (PIDNN) control
CN103853050A (en) * 2014-03-19 2014-06-11 湖北蔚蓝国际航空学校有限公司 PID optimization control method of four-rotor aircraft
WO2017037354A1 (en) * 2015-09-01 2017-03-09 Renault S.A.S Device and method for controlling a driver assistance system for correcting the direction of travel of a motor vehicle
CN110531779A (en) * 2019-08-02 2019-12-03 杭州电子科技大学 A kind of unmanned plane apery collaboration flight controller and its implementation

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
涂亚庆等: "一类对象的仿人智能控制算法", 《重庆大学学报(自然科学版)》 *
窦立谦等: "基于状态观测器的多无人机编队跟踪控制", 《天津大学学报(自然科学与工程技术版)》 *
雷建和等: "基于粒子群算法的四旋翼仿人智能控制器设计", 《计算机仿真》 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN112016162A (en) * 2020-09-05 2020-12-01 江西理工大学 Four-rotor unmanned aerial vehicle PID controller parameter optimization method

Similar Documents

Publication Publication Date Title
CN106094860B (en) Quadrotor and its control method
CN111185907B (en) Pose stability control method for operation type flying robot after grabbing
CN108638068B (en) Design method of flying robot control system with redundant mechanical arm
CN110531786B (en) Unmanned aerial vehicle maneuvering strategy autonomous generation method based on DQN
CN112346470A (en) Four-rotor attitude control method based on improved active disturbance rejection control
CN109946971B (en) Smooth switching control method for transition section of tilt rotor unmanned aerial vehicle
CN110908281A (en) Finite-time convergence reinforcement learning control method for attitude motion of unmanned helicopter
CN111506099B (en) Intelligent control system and method for height of unmanned aerial vehicle
CN107085435A (en) Hypersonic aircraft attitude harmony control method based on coupling analysis
CN111459188B (en) Quaternion-based multi-rotor nonlinear flight control method
CN112327926B (en) Self-adaptive sliding mode control method for unmanned aerial vehicle formation
CN110333733A (en) A kind of the tandem variable universe fuzzy attitude control system and method for quadrotor
CN112578805A (en) Attitude control method of rotor craft
CN111240204B (en) Model reference sliding mode variable structure control-based flying projectile patrol control method
CN114721266B (en) Self-adaptive reconstruction control method under condition of structural failure of control surface of airplane
CN107479382A (en) Hypersonic aircraft neural network control method based on online data study
Akbar et al. Adaptive modified super-twisting control for a quadrotor helicopter with a nonlinear sliding surface
CN111580550A (en) Unmanned aerial vehicle human-simulated intelligent control method
CN116820134A (en) Unmanned aerial vehicle formation maintaining control method based on deep reinforcement learning
CN115685764B (en) Task self-adaptive anti-interference tracking control method and system for variable-span aircraft
CN113342023A (en) Unmanned aerial vehicle humanoid control method based on improved fuzzy PID
CN116540780A (en) Unmanned aerial vehicle decision control method based on game guidance
CN109857146B (en) Layered unmanned aerial vehicle tracking control method based on feedforward and weight distribution
CN113885549B (en) Four-rotor gesture track control method based on dimension clipping PPO algorithm
CN113580127A (en) Multi-rotor aircraft humanoid double-mechanical-arm system and dynamic self-balancing control design method thereof

Legal Events

Date Code Title Description
PB01 Publication
PB01 Publication
SE01 Entry into force of request for substantive examination
SE01 Entry into force of request for substantive examination
RJ01 Rejection of invention patent application after publication
RJ01 Rejection of invention patent application after publication

Application publication date: 20200825