CN108983821A - A kind of PID automatic pilot parameter tuning method based on intelligent algorithm - Google Patents
A kind of PID automatic pilot parameter tuning method based on intelligent algorithm Download PDFInfo
- Publication number
- CN108983821A CN108983821A CN201810945271.2A CN201810945271A CN108983821A CN 108983821 A CN108983821 A CN 108983821A CN 201810945271 A CN201810945271 A CN 201810945271A CN 108983821 A CN108983821 A CN 108983821A
- Authority
- CN
- China
- Prior art keywords
- parameter
- automatic pilot
- aircraft
- pid
- model
- 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
Links
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
Landscapes
- Engineering & Computer Science (AREA)
- Aviation & Aerospace Engineering (AREA)
- Radar, Positioning & Navigation (AREA)
- Remote Sensing (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Feedback Control In General (AREA)
Abstract
The invention belongs to the parameter tuning methods of PID automatic pilot, and in particular to a kind of parameter tuning method based on intelligent algorithm.According to the Three Degree Of Freedom nonlinear model of aircraft, the Controlling model of aircraft is taken out；Build the actuator model and autopilot model of aircraft；The performance indicator of selected automatic pilot；Preset the pid parameter of automatic pilot；The parameter of setting genetic algorithm resolves the pid parameter of automatic pilot according to constraint condition and fitness function；The form of the penalty P (x) is as follows:R=(ct)α；Wherein, fjIt (x) is the penalty item for violating j-th of constraint, c, α, β is constant, and t is the evolutionary generation of optimization algorithm, and m is constraint sum；F (x)=f (x)+P (x), wherein f (x) is objective function, and p (x) is penalty, and F (x) is broad object function.A kind of pid parameter setting method based on genetic algorithm is proposed, the burden of staff is reduced, improves flight automatization level.
Description
Technical field
The invention belongs to the parameter tuning methods of PID automatic pilot, and in particular to a kind of parameter based on intelligent algorithm
Setting method.
Background technique
Autopilot Design method engineer application it is most be PID control method, i.e., traditional ratio, integral is micro-
The advantages that sub-control system, PID control have algorithm simple, high reliablity, do not depend on the model of control target, is used extensively
In production process, but in Practical Project, the ratio of PID controller, integral and differential adjustment parameter are often gathered using experiment additional examination
Method by manually adjusting, this not only needs a twist of the wrist, toward contact it is fairly time consuming.
Summary of the invention
Present invention solves the technical problem that: a kind of pid parameter setting method based on genetic algorithm is proposed, by parameter
Space encoding simultaneously uses random selection to carry out guiding search process as tool and develop towards more efficient direction, to improve pid parameter
The efficiency of adjusting reduces the burden of staff, improves flight automatization level.
Technical solution of the present invention: a kind of PID automatic pilot parameter tuning method based on intelligent algorithm, the side
Method includes the following steps:
According to the Three Degree Of Freedom nonlinear model of aircraft, the Controlling model of aircraft is taken out；
Construct the actuator model and autopilot model of aircraft；
The performance indicator of selected automatic pilot；
Preset the pid parameter of automatic pilot；
The parameter for setting genetic algorithm carries out the pid parameter of automatic pilot according to constraint condition and fitness function
It resolves；
The form of the penalty P (x) is as follows:
Wherein, fjIt (x) is the penalty item for violating j-th of constraint, c, α, β is constant, and t is the evolution generation of optimization algorithm
Number, m are constraint sum；
F (x)=f (x)+P (x)
Wherein, f (x) is objective function, and p (x) is penalty, and F (x) is broad object function.
Preferably, after the Controlling model for taking out aircraft, aircraft state parameter is calculated by quadravalence Long Gekutafa.It should
The calculation accuracy to aircraft state parameter can be improved in method, meets engineer application.
Preferably, the c, α, the value range of β are respectively as follows: β=2, c, a ∈ [0.5,1].It in the range can be in the hope of
Optimal or preferably result.
Beneficial effects of the present invention: the present invention is using PID automatic pilot as model, under traditional mode, by artificial
Experience carry out manual parameters adjusting it is time-consuming, it is laborious the disadvantages of；Using intelligent algorithm as optimization means, pass through reasonable model foundation
And the setting of respective algorithms parameter and fitness function, calculating adjusting quickly can be carried out to pid parameter, filter out and conform to
The pid parameter asked.This method can effectively mitigate the burden of staff, improve working efficiency.
Detailed description of the invention
Fig. 1 is the Diagram Model of steering engine；
Fig. 2 is automatic pilot block diagram；
Fig. 3 is the fundamental block diagram of PID controller.
Specific embodiment
Now in conjunction with attached drawing, embodiment, invention is further described in detail.
PID automatic pilot parameter tuning method based on genetic algorithm, specific steps are as follows:
Step 1: constructing the Three Degree Of Freedom nonlinear model of aircraft
The motion model of aircraft has the characteristics that the correlation of each motion state: attitude motion of the aircraft in space can
It is divided into three pitching, yaw, rolling channels, mutual coupling is occurred by inertia, damping, aerodynamic force or electrical link between them
It closes；The time variation of aerodynamic parameter: in flight course, quality, rotary inertia, centroid position of aircraft etc. and related with flight
The constantly variation at any time such as aerodynamic coefficient so that the equation of motion becomes the differential equation of one group of variable coefficient；Big compression ring
The complexity in border: aircraft interference suffered in flight course is mainly the uncertainty of flight environment of vehicle, such as wind, cloud cluster, mountain range
Landform etc..
For the above feature, model aircraft is divided into the kinetics equation of center of mass motion in the research to optimization method,
The kinetics equation of rotation around center of mass, the equation of motion of rotation around center of mass carry out modeling analysis respectively.
Step 2: taking out corresponding Controlling model
According to the airplane motion model that step 1 derives, with the angle of attack of aircraft, yaw angle, three angular speed of body axis
ωx,ωy,ωzAnd roll angle, yaw angle, pitch angle, as the state variable of pilot triple channel Controlling model, to model into
Row is abstract, can obtain Controlling model.Simultaneously for the ease of being designed and analyzing with mature linear control system design theory,
The nonlinear dynamical equation of aircraft is simplified to the Linear Time Invariant differential equation, has made following hypothesis:
Microvariations linearisation is assumed: i.e. the characteristic of the components such as hypothesis electronic circuit, rudder servo mechanism, instrument and movement side
Journey is all linear.Simultaneously assume missile aerodynamic force characteristic be it is linear, this can set up in the case where microvariations.
Only consider aircraft short-period motion: on the signature of flight path point studied, Aircraft Quality, rotary inertia, speed,
Atmospheric parameter etc. is considered as constant, that is, uses the principle of solidification.
In order to guarantee the solving precision of the differential equation, quadravalence Long Gekutafa can be used, formula is as follows:
Wherein, h is simulation step length.
Step 3: constructing the actuator model of aircraft
Model framework chart is as shown in Figure 1, the model is retouched using the two―step element of angle saturation limiting and angular speed saturation limiting
It states.Steering engine in model has the property that maximum angle of rudder reflection: 30 degree；Minimum angle of rudder reflection: -30 degree；Maximum angle of rudder reflection speed:
500 degrees seconds；Natural frequency: 150Hz；Bandwidth: 24 radian per seconds；Damped coefficient: 0.7；Transfer function model are as follows:
Step 4: building autopilot model
Automatic pilot is as shown in Fig. 2, in the automatic pilot: mainly by gain compensation, steering engine, and body, sensor,
Feedback factor and data processor composition；Steering engine controls the attitudes vibration of body, and sensor measurement goes out the occurrence of angle, carries out
Conversion feeds back to data processor, carries out error revision and issues new instruction.
The fundamental block diagram of PID controller, as shown in figure 3, mainly by ratio, integral, differential parameter；Integrator, differentiator
It is formed with summation module；Input instruction are as follows: αc=5；βc=0；γc=5.
Step 5: the performance indicator of selected automatic pilot
It must be selected corresponding according to the specific performance of aerial mission target and automatic pilot before optimizing
Performance indicator.The performance indicator of control system includes many aspects, such as: signal errors, energy consumption, the rise time, overshoot,
Steady state error etc..
Step 6:PID parameter is presetting
According to the characteristics of control system and the possible range of parameter to be optimized, by the value of parameter be limited to one it is reasonable
It in region, can not only guarantee algorithmic stability and convergence, improve optimization degree, operation time can also be reduced.
Step 7: setting the parameter of genetic algorithm
Parameter in genetic algorithm mainly has: determining the approximate range of each parameter and the number of code length, initial population
Individual each in population is decoded into corresponding parameter value, duplication, setting intersection and mutation probability, termination condition etc. by mesh.
Step 8: reasonably designing each constraint condition and fitness function
The quantity of state of known current aircraft, the position of target waypoint, flight plan and performance condition, it is reasonable selected winged
Machine satiable flying height in flight course, the flight time, the constraint conditions such as overload, constraint condition is because of selected model and flies
The difference of row task and slightly difference；When optimizing algorithm design, by the K of PIDp,Ki,KdThree parameter combinations are made together
For each of genetic algorithm group individual, the fitness value of each group of parameter is calculated according to fitness function；Then, to group
It is selected, intersects and mutation operation, continuous evolution obtain PID controller until finding optimal objective individual in group
Optimized parameter.
Objective function is exactly that each parameter in flight course, in aerial vehicle trajectory is meeting certain terminal and path about
Under conditions of beam, keep some performance indicator optimal by carrying out optimal selection to control variable, that is, the objective function chosen, mesh
Scalar functions can be the combination of a function single optimization or several functions.
MinJ=F (x)
Wherein, F (x) is broad object function.
Broad object functional form is as follows:
F (x)=f (x)+P (x)
Wherein, f (x) is objective function, and p (x) is penalty.
Since the penalty coefficient of normal static penalty is fixed and invariable, to different problems, need to choose difference
Numerical value repeatedly attempt, larger workload and when penalty coefficient is chosen not at that time is just unable to get satisfied result.A kind of improvement
Mode be to become penalty coefficient dynamically to allow to the variation with the number of iterations and change, adaptive basis
The information to develop in searching for automatically adjusts corresponding penalty coefficient.
The form of penalty P (x) is as follows:
fjIt (x) is the penalty item for violating j-th of constraint, c, α, β is constant, and t is the evolutionary generation of optimization algorithm.
This method advantage is to require parameter few, only 3.But the superiority and inferiority of optimizing effect, which compares, relies on this 3 parameters, is setting
Parameter is counted to need with caution.
Embodiment
Selection operator, crossover operator and mutation operator are first set in MATLAB programming.Selection operator when emulation are as follows: high
Position operator；Crossover operator are as follows: centre recombination two point intersects；Mutation operator are as follows: high position variation.Emulate sampling time for using for
2ms, input instruction are αc=5, βc=0, γc=5.Using binary coding mode, the binary coding for being 10 with length
String, respectively indicates 3 decision variable Kp,Ki,Kd；To obtain satisfied transient process dynamic characteristic, using the flat of Error Absolute Value
Minimum target function of the time integral performance indicator of side's sum as parameter selection.Performance indicator are as follows:
The range of 9 parameters of J (2) is respectively as follows:
[-6.5,0],[-12.5,0],[-0.2,0]；[-12.5,0],[-12.5.0],[-0.2,0]；[-8.5,0],[-
1.5,0],[-0.5,0]。
It evolves by 50 generations, the Optimal Parameters of acquisition are as follows:
Performance indicator, 9 pid parameters are as follows:
kp1=-6.4549, ki1=-12.9776, kd1=-0.0506；
kp2=-10.7751, ki2=0.6255, kd2=-0.0467；
kp3=-6.8280, ki3=0.1125, kd3=-0.1272
3 angles are difference: α=5.0002, β=0.0286, γ=4.9960.
Claims (3)
1. a kind of PID automatic pilot parameter tuning method based on intelligent algorithm, it is characterised in that the method includes such as
Lower step:
According to the Three Degree Of Freedom nonlinear model of aircraft, the Controlling model of aircraft is taken out；
Construct the actuator model and autopilot model of aircraft；
The performance indicator of selected automatic pilot；
Preset the pid parameter of automatic pilot；
The parameter of setting genetic algorithm solves the pid parameter of automatic pilot according to constraint condition and fitness function
It calculates；
The form of the penalty P (x) is as follows:
Wherein, fjIt (x) is the penalty item for violating j-th of constraint, c, α, β is constant, and t is the evolutionary generation of optimization algorithm, m
For constraint sum；
F (x)=f (x)+P (x)
Wherein, f (x) is objective function, and p (x) is penalty, and F (x) is broad object function.
2. a kind of PID automatic pilot parameter tuning method based on intelligent algorithm according to claim 1, feature
Are as follows: after taking out the Controlling model of aircraft, aircraft state parameter is calculated by quadravalence Long Gekutafa.
3. a kind of PID automatic pilot parameter tuning method based on intelligent algorithm according to claim 1, feature exist
In the c, the value range of α, β are respectively as follows: β=2, c, a ∈ [0.5,1].
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810945271.2A CN108983821A (en) | 2018-08-17 | 2018-08-17 | A kind of PID automatic pilot parameter tuning method based on intelligent algorithm |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810945271.2A CN108983821A (en) | 2018-08-17 | 2018-08-17 | A kind of PID automatic pilot parameter tuning method based on intelligent algorithm |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108983821A true CN108983821A (en) | 2018-12-11 |
Family
ID=64553727
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810945271.2A Pending CN108983821A (en) | 2018-08-17 | 2018-08-17 | A kind of PID automatic pilot parameter tuning method based on intelligent algorithm |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108983821A (en) |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20130126125A (en) * | 2012-05-10 | 2013-11-20 | 수원대학교산학협력단 | Method for controlling constant pressure using fuzzy pid controller |
| CN103645636A (en) * | 2013-11-25 | 2014-03-19 | 南京航空航天大学 | PID controller parameter optimization method |
| 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 |
| CN106054596A (en) * | 2016-06-23 | 2016-10-26 | 东南大学 | PID (Proportion Integration Differentiation) controller parameter optimization setting method based on improved performance indexes |
| CN107703749A (en) * | 2017-10-10 | 2018-02-16 | 常州大学 | A kind of pid parameter optimization method based on self-adapted genetic algorithm |
| CN108181802A (en) * | 2017-12-05 | 2018-06-19 | 东南大学 | A kind of controllable PID controller parameter optimization setting method of performance |
| CN108267958A (en) * | 2018-01-25 | 2018-07-10 | 北京控制工程研究所 | A kind of controller parameter optimization method, system and the medium of the H ∞ norm indexs decomposed based on modal structure |
-
2018
- 2018-08-17 CN CN201810945271.2A patent/CN108983821A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20130126125A (en) * | 2012-05-10 | 2013-11-20 | 수원대학교산학협력단 | Method for controlling constant pressure using fuzzy pid controller |
| CN103645636A (en) * | 2013-11-25 | 2014-03-19 | 南京航空航天大学 | PID controller parameter optimization method |
| 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 |
| CN106054596A (en) * | 2016-06-23 | 2016-10-26 | 东南大学 | PID (Proportion Integration Differentiation) controller parameter optimization setting method based on improved performance indexes |
| CN107703749A (en) * | 2017-10-10 | 2018-02-16 | 常州大学 | A kind of pid parameter optimization method based on self-adapted genetic algorithm |
| CN108181802A (en) * | 2017-12-05 | 2018-06-19 | 东南大学 | A kind of controllable PID controller parameter optimization setting method of performance |
| CN108267958A (en) * | 2018-01-25 | 2018-07-10 | 北京控制工程研究所 | A kind of controller parameter optimization method, system and the medium of the H ∞ norm indexs decomposed based on modal structure |
Non-Patent Citations (6)
| Title |
|---|
| 堵会晓，黄靓: "基于遗传算法的反导拦截弹助推段自动驾驶仪的设计", 《火力与指挥控制》 * |
| 朱苏朋: "《计算机控制及仿真》", 31 August 2009 * |
| 线博: "高超声速飞行器再入轨迹优化及仿真", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 胡占双: "无人机飞行姿态检测及控制研究", 《中国优秀硕士学位论文全文数据库 工程科技Ⅱ辑》 * |
| 蔡金狮: "《动力学系统辨识与建模》", 31 December 1991 * |
| 贝超，李芸，林维菘: "基于遗传算法的导弹稳定控制回路", 《系统工程与电子技术》 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Shim et al. | A comprehensive study of control design for an autonomous helicopter | |
| Kang et al. | Linear tracking for a fixed-wing UAV using nonlinear model predictive control | |
| CN109947124A (en) | Improve particle swarm algorithm Optimization of Fuzzy PID unmanned helicopter attitude control method | |
| CN109976156A (en) | Fixed-wing unmanned plane, which is dwelt, falls the modeling and forecast Control Algorithm of motor-driven track | |
| Sun et al. | UAV path planning based on improved rapidly-exploring random tree | |
| CN104597911A (en) | Adaptive optimal butt joint trajectory tracking flying control method for air refueling receiving machine | |
| CN110456821A (en) | Aerial vehicle trajectory method for optimally controlling and system based on Dynamic trigger mechanism | |
| CN109976378A (en) | Wind, which disturbs lower unmanned plane and dwells, falls motor-driven method for controlling trajectory | |
| Zhang et al. | Recurrent neural network-based model predictive control for multiple unmanned quadrotor formation flight | |
| CN110083168A (en) | Small-sized depopulated helicopter based on enhancing study determines high control method | |
| Gai et al. | Adaptive neural network dynamic inversion with prescribed performance for aircraft flight control | |
| Boughari et al. | New Methodology for Optimal Flight Control Using Differential Evolution Algorithms Applied on the Cessna Citation X Business Aircraft–Part 1. Design and Optimization | |
| Velagić et al. | Design of LQR controller for 3D trajectory tracking of octocopter unmanned aerial vehicle | |
| Lazim et al. | Intelligent observer-based feedback linearization for autonomous quadrotor control | |
| CN112231845B (en) | Stratospheric airship height control method and system | |
| CN114237268A (en) | Unmanned aerial vehicle strong robust attitude control method based on deep reinforcement learning | |
| CN108763779A (en) | A kind of method that the particle cluster algorithm of application enhancements controls quadrotor drone | |
| Abdulla et al. | Roll control system design using auto tuning LQR technique | |
| CN108983821A (en) | A kind of PID automatic pilot parameter tuning method based on intelligent algorithm | |
| CN117148851A (en) | Unmanned parafoil sectional type flight path planning method based on particle swarm optimization algorithm | |
| CN111580389A (en) | Three-degree-of-freedom helicopter explicit model prediction control method based on deep learning | |
| Lu et al. | Quadrotor inverted Pendulum control based on improved Particle Swarm Optimization | |
| CN115407661A (en) | Multi-unmanned aerial vehicle system nonlinear robust tracking control method based on azimuth measurement information | |
| Roy | Horizontal position control of a small scale helicopter using adaptive backstepping controller | |
| CN114047778A (en) | Small airplane short-distance automatic landing transverse and lateral control method |
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: 20181211 |