CN113836810B - Aircraft and attitude control model parameter identification method and device thereof and storage medium - Google Patents

Aircraft and attitude control model parameter identification method and device thereof and storage medium Download PDF

Info

Publication number
CN113836810B
CN113836810B CN202111138454.1A CN202111138454A CN113836810B CN 113836810 B CN113836810 B CN 113836810B CN 202111138454 A CN202111138454 A CN 202111138454A CN 113836810 B CN113836810 B CN 113836810B
Authority
CN
China
Prior art keywords
population
parameter
aircraft
identification
fitness
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.)
Active
Application number
CN202111138454.1A
Other languages
Chinese (zh)
Other versions
CN113836810A (en
Inventor
杜海铭
孙健
王钢
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.)
Chongqing Innovation Center of Beijing University of Technology
Original Assignee
Chongqing Innovation Center of Beijing University of Technology
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 Chongqing Innovation Center of Beijing University of Technology filed Critical Chongqing Innovation Center of Beijing University of Technology
Priority to CN202111138454.1A priority Critical patent/CN113836810B/en
Publication of CN113836810A publication Critical patent/CN113836810A/en
Application granted granted Critical
Publication of CN113836810B publication Critical patent/CN113836810B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/20—Design optimisation, verification or simulation
    • G06F30/27—Design optimisation, verification or simulation using machine learning, e.g. artificial intelligence, neural networks, support vector machines [SVM] or training a model
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F30/00—Computer-aided design [CAD]
    • G06F30/10—Geometric CAD
    • G06F30/15—Vehicle, aircraft or watercraft design
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F2119/00—Details relating to the type or aim of the analysis or the optimisation
    • G06F2119/02—Reliability analysis or reliability optimisation; Failure analysis, e.g. worst case scenario performance, failure mode and effects analysis [FMEA]
    • Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
    • Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
    • Y02T—CLIMATE CHANGE MITIGATION TECHNOLOGIES RELATED TO TRANSPORTATION
    • Y02T90/00—Enabling technologies or technologies with a potential or indirect contribution to GHG emissions mitigation

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Geometry (AREA)
  • Theoretical Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Evolutionary Computation (AREA)
  • General Engineering & Computer Science (AREA)
  • Computer Hardware Design (AREA)
  • Medical Informatics (AREA)
  • Software Systems (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Artificial Intelligence (AREA)
  • Automation & Control Theory (AREA)
  • Aviation & Aerospace Engineering (AREA)
  • Computational Mathematics (AREA)
  • Mathematical Analysis (AREA)
  • Mathematical Optimization (AREA)
  • Pure & Applied Mathematics (AREA)
  • Control Of Position, Course, Altitude, Or Attitude Of Moving Bodies (AREA)

Abstract

The invention provides an aircraft and a method, a device and a storage medium for identifying parameters of an attitude control model thereof, wherein a genetic algorithm is innovatively introduced to identify parameters of a characteristic model, the characteristic model is respectively established for three attitude angle channels of pitching, rolling and yawing of the aircraft, the characteristic models are respectively input into a genetic algorithm identification process, and a required identification parameter estimation result is output after reaching an accuracy requirement index and can be used for controller design; the on-line cyclic operation can be realized, but repeated matrix operation is not needed, so that the requirement on hardware computing force is reduced, and the parameter identification efficiency is improved; different from the traditional algorithm, the genetic algorithm considers a plurality of search results at the same time, takes an objective function as search information, is more effective, and can avoid a local optimal solution, so that the parameter identification precision is improved; furthermore, the genetic algorithm has minimal restriction on the objective function, and its parallel computing power increases its running speed.

Description

Aircraft and attitude control model parameter identification method and device thereof and storage medium
Technical Field
The invention relates to the technical field of flight attitude control, in particular to an aircraft, an attitude control model parameter identification method and device thereof, and a storage medium.
Background
The attitude control research of the aircraft is necessary, and the rapidness, the precision and the robustness are ensured by the attitude control. Most algorithms need to rely on accurate mathematical models of aircrafts, and selection of related parameters is not enough to follow mature rules, so that the change of the parameters is sensitive to influence of control effects, and limitations exist.
And establishing a model and designing a subsequent controller, wherein the characteristic parameters are required to be identified on line. The common identification method includes a least square method, a gradient projection method and the like, and the characteristic model parameters are obtained after the on-line repeated input of attitude state quantity circulation calculation and the stable fixation of the parameter requirement range are finally achieved.
The traditional parameter identification algorithm such as a least square method and a gradient algorithm is an acyclic single algorithm, data is subjected to offline identification processing, and the identification related to the change state quantity in the characteristic modeling process cannot acquire the change condition of parameters along with time, so that the method cannot be applied to the control modeling of the attitude parameters of the aircraft.
In order to be suitable for aircraft attitude parameter control modeling, the traditional parameter identification algorithm is improved at present, for example, the current common feature model parameter identification method comprises recursive least square method online identification, recursive least square method online identification containing forgetting factors and the like.
Recursive least squares identification: new parameter estimation value = old parameter estimation value + correction term. According to the time sequence, the state quantity changes with each time during the running of the aircraft are estimated, and the values are corrected with the advancement of time.
Recursive least squares identification with forgetting factors: in the identification process, old data are replaced by new data, so that forgetting factors are added on the basis of a recursive least square identification algorithm to adapt to the change of time-varying system parameters, the weight of the data at the earlier moment is reduced, the correlation degree of the new data is improved, and the reference effect of the old data is continuously reduced.
However, the improved feature model parameter identification algorithm still has the following problems:
because two factors need to be considered when selecting the parameter identification method: accuracy of identification and time required for parameter identification. The attitude control of the aircraft needs to be more real-time, and the identification of the characteristic parameters is an important loop in the control process, so that the identification speed is improved as much as possible. However, the conventional characteristic parameter identification method such as recursive least square requires a relatively long iteration number, and each round of calculation and estimation processes are relatively complex and the parameter convergence speed is relatively slow.
Secondly, the characteristic parameters of the identified characteristic model are not accurately reflected on the controlled object, or the response output error of the characteristic model and the object dynamic model which are possibly obtained is relatively larger.
Thirdly, matrix operation is involved in the least square algorithm, wherein the matrix P (k) has larger operation amount, and the calculation capability of a mainstream ARM architecture chip adopted by an actual unmanned aerial vehicle is a challenge. And the flight control system has more other parallel processes, and the real-time control effect can be influenced by occupying excessive calculation force.
Fourth, recursive least square with forgetting factors is difficult to choose the proper forgetting factor. When the forgetting factor lambda is larger, the identification result forgets slowly, and the algorithm loses the parameter estimation capability along with the increase of the identification data; when the forgetting factor lambda is smaller, the obtained identification result is forgotten fast, the identification data is increased, and each element in the identification result tends to be endless, so that the parameters to be identified can be fluctuated in a large range, and even the identification result is diverged.
Disclosure of Invention
The invention provides an aircraft and a method, a device and a storage medium for identifying parameters of an attitude control model thereof, which mainly solve the technical problems that: how to improve the accuracy and efficiency of the parameter identification of the attitude control model of the aircraft.
In order to solve the technical problems, the invention provides an aircraft attitude control model parameter identification method, which comprises the following steps:
s10, respectively establishing the following characteristic models for three attitude angle channels of pitching, rolling and yawing of the aircraft:
X(k+1)＝α 1 (k)X(k)+α 2 (k)X(k-1)+β 0 (k)U(k)； (1)
wherein X (k+1) is the estimated attitude angle at the next moment, X (k) is the actual attitude angle at the current moment, X (k-1) is the actual attitude angle at the previous moment, U (k) is the control moment, and alpha 1 (k)、α 2 (k) And beta 0 (k) Is a parameter to be identified;
s11, defining an objective function according to the characteristic model error;
s12, constructing an adaptability function based on the reciprocal of the objective function;
s20, acquiring identification parameter estimation output at the previous moment, performing binary coding on the identification parameter estimation, and inputting a genetic algorithm model for genetic operation to generate a next generation population;
s21, calculating the fitness of the population based on the fitness function;
s22, judging whether the fitness of the population meets the precision requirement or not; if yes, go to step S23; if not, go to step S24;
s23, decoding the population to obtain parameter estimation of the parameters to be identified;
s24, updating the evolution algebra;
s25, judging whether the maximum evolution generation is reached; if yes, go to step S23; if not, go to step S26;
s26, judging whether the current evolution algebra is more than or equal to 2; if yes, go to step S27; if not, go to step S28;
s27, judging whether the fitness of the current population is better than that of the previous generation population; if yes, go to step S291; if not, go to step S292;
s28, taking the first generation population as a parent population, and continuing to perform gene operation;
s291, continuing to perform gene operation by taking the current population as a parent population;
s292, taking the previous generation population as a parent population, and continuing gene operation;
s30, saving parameter estimation of the parameter to be identified, and being used for estimating corresponding attitude parameters of the channel at the next moment.
Optionally, the step S11 of defining an objective function according to the feature model error includes:
the characteristic model error is as follows:
e＝X M -X τ ； (2)
wherein X is M For the actual attitude angle output, X τ For characteristic model attitude angle estimation, the sampling frequency is N, and an objective function is defined as follows:
optionally, the fitness function is:
wherein the population isM is population size.
Alternatively, the genetic manipulation includes a replication manipulation, a crossover manipulation, and a mutation manipulation.
Optionally, the copying operation selects a wheel selection algorithm:
wherein Pc is the crossover probability, pm is the mutation probability, f max 、f avg F 'is the maximum fitness and average fitness in the population, f' is the greater fitness in the two intersecting individuals;
the crossing operation is uniform crossing;
individual decision method for mutation operation selection based on mutation probability P m Randomly selecting a corresponding number of mutant individuals in the population to carry out mutation operation.
Optionally, the method further comprises: when the identification parameter estimation output at the last moment does not exist, randomly generating an initial population in a set parameter range for binary coding, and inputting a genetic algorithm model for genetic operation to generate a next generation population.
The invention also provides an aircraft attitude control model parameter identification device, which comprises:
the characteristic model creation module is used for respectively creating the following characteristic models for three attitude angle channels of pitching, rolling and yawing of the aircraft:
X(k+1)＝α 1 (k)X(k)+α 2 (k)X(k-1)+β 0 (k)U(k)； (7)
wherein X (k+1) is the estimated attitude angle at the next moment, X (k) is the actual attitude angle at the current moment, X (k-1) is the actual attitude angle at the previous moment, U (k) is the control moment, and alpha 1 (k)、α 2 (k) And beta 0 (k) Is a parameter to be identified;
defining an objective function according to the characteristic model error; constructing an fitness function based on the inverse of the objective function;
the parameter identification module is used for acquiring the identification parameter estimation output at the previous moment from the storage module, performing binary coding on the identification parameter estimation, and inputting a genetic algorithm model for genetic operation so as to generate a next generation population; calculating the fitness of the population based on the fitness function; judging whether the fitness of the population meets the precision requirement or not; if so, decoding the population to obtain parameter estimation of the parameters to be identified, and sending the parameter estimation to the storage module; if not, updating the evolution algebra; judging whether the maximum evolution generation is reached; if yes, decoding the population; if not, judging whether the current evolution algebra is more than or equal to 2; if not, taking the first generation population as a parent population, and continuing gene operation to generate a next generation population; if the current evolution algebra is more than or equal to 2, judging whether the adaptability of the current population is better than that of the previous generation population; if so, continuing gene operation by taking the current population as a parent population to generate a next generation population; if not, taking the previous generation population as a parent population, and continuing gene operation to generate the next generation population;
the storage module is used for storing parameter estimation of the parameter to be identified so as to be used for estimating corresponding attitude parameters of the channel at the next moment.
The invention also provides an aircraft, which comprises the aircraft attitude control model parameter identification device.
The present invention also provides a storage medium storing one or more programs executable by one or more processors to implement the steps of the aircraft attitude control model parameter identification method as described above.
The beneficial effects of the invention are as follows:
according to the aircraft and the attitude control model parameter identification method, device and storage medium thereof provided by the invention, a genetic algorithm is innovatively introduced to identify parameters of a characteristic model, the characteristic model is respectively established for three attitude angle channels of pitching, rolling and yawing of the aircraft, the characteristic model is respectively input into a genetic algorithm identification process, and a required identification parameter estimation result is output after reaching an accuracy requirement index, so that the aircraft can be used for controller design; the on-line cyclic operation can be realized, but repeated matrix operation is not needed, so that the requirement on hardware computing force is reduced, and the parameter identification efficiency is improved; different from the traditional algorithm, the genetic algorithm (Genetic Algorithm, GA) considers a plurality of search results at the same time, takes an objective function as search information, is more effective, and can avoid a local optimal solution, so that the parameter identification precision is improved; furthermore, the GA has minimal restriction on the objective function, and its parallel computing power increases its running speed.
Drawings
FIG. 1 is a schematic flow chart of a method for identifying parameters of an aircraft attitude control model according to a first embodiment of the present invention;
fig. 2 is a schematic diagram of an aircraft attitude control principle according to a first embodiment of the present invention;
FIG. 3 is a diagram illustrating a recognition result according to a first embodiment of the present invention;
FIG. 4 is a diagram showing the identification result of a comparison method according to the first embodiment of the present invention;
FIG. 5 is a schematic structural diagram of an aircraft attitude control model parameter identification device according to a second embodiment of the present invention;
fig. 6 is a schematic view of an aircraft structure according to a third embodiment of the invention.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention will be further described in detail by the following detailed description with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are for purposes of illustration only and are not intended to limit the scope of the invention.
Embodiment one:
in order to improve the accuracy and efficiency of identifying parameters of an aircraft attitude control model, the embodiment provides a method for identifying parameters of an aircraft attitude control model, please refer to fig. 1-2, which mainly includes the following steps:
s10, respectively establishing the following characteristic models for three attitude angle channels of pitching, rolling and yawing of the aircraft:
X(k+1)＝α 1 (k)X(k)+α 2 (k)X(k-1)+β 0 (k)U(k)； (8)
wherein X (k+1) is the estimated attitude angle at the next moment, X (k) is the actual attitude angle at the current moment, X (k-1) is the actual attitude angle at the previous moment, U (k) is the control moment, and alpha 1 (k)、α 2 (k) And beta 0 (k) Is the parameter to be identified. The control moment refers to a direct control quantity input to a controlled object attitude ring, and for three attitude angle direction calculation methods, taking a four-rotor aircraft as an example, the control moment can be determined by calculating the rotation speeds of four motors, and the specific calculation process is not repeated here.
S11, defining an objective function according to the characteristic model error.
Definition of decision variablesConstraints and objective functions (recognition accuracy). Wherein the error in defining the pose is e=x M -X τ Wherein X is M For outputting the attitude angle of the actual object, X τ For feature model attitude angle estimation, define the objective function as +.>Wherein, N is the sampling times, namely the number of data sampled in one cycle when calculating the fitness. The identification process needs to solve the problem of minima of the objective function J.
S12, constructing an adaptability function based on the inverse of the objective function.
Designing a fitness function based on the inverse of the objective function:
setting genetic parameters including M: population size, G: maximum evolution generation, P c : crossover probability, P m : mutation probability.
In the embodiment, the population size is set at 50-150; maximum genetic passage is 200-500; crossover probability 1 > Pc2 > 0, mutation probability Pm1:0.1 to 0.5, pm2:0.001 to 0.01.
S20, acquiring identification parameter estimation output at the previous moment, performing binary coding on the identification parameter estimation, and inputting a genetic algorithm model for genetic operation to generate a next generation population.
When the attitude parameters of the aircraft characteristic model are identified for the first time, i.e. the identification parameter estimation output at the last moment does not exist, the initial population is randomly generated in the set parameter rangeWherein M is the population scale, and the space dimension of the solution is the number of vectors to be identified. For binary encoding thereof, for genetic manipulation with input genetic algorithm models to generate next generation populations.
In this embodiment, the set parameter range can be set to f according to the definition of feature modeling 1 (k)∈(1，2]、f 2 (k)∈[-1，0)，g 0 (k) 1; when the aircraft with different objects is applied, the specific setting is performed according to the characteristics of the aircraft, and the detailed description is omitted.
Genetic manipulation includes replication manipulation, crossover manipulation and mutation manipulation.
The copying operation adopts a wheel disc selection algorithm, so that the diversity of population individuals is enriched, and the occurrence of local optimization is reduced:
wherein f max 、f avg F 'is the maximum fitness and average fitness in the population, f' is the greater fitness in the crossing individuals;
the crossing operation is uniform crossing;
individual decision method for mutation operation selection based on mutation probability P m Randomly selecting a corresponding number of mutant individuals in the population to carry out mutation operation.
S21, calculating the fitness of the population based on the fitness function.
S22, judging whether the fitness of the population meets the precision requirement or not; if yes, go to step S23; if not, go to step S24.
The adaptability accuracy requirement can be flexibly set based on the characteristics of the identified system and the environment, which is not limited in the embodiment.
S23, decoding the population to obtain parameter estimation of the parameters to be identified
S24, updating the evolution algebra.
S25, judging whether the maximum evolution generation is reached; if yes, go to step S23; if not, go to step S26.
S26, judging whether the current evolution algebra is more than or equal to 2; if yes, go to step S27; if not, go to step S28.
S27, judging whether the fitness of the current population is better than that of the previous generation population; if yes, go to step S291; if not, go to step S292.
S28, taking the first generation population as a parent population, and continuing to perform gene operation.
S291, continuing to perform gene operation by taking the current population as a parent population.
S292, taking the previous generation population as a parent population, and continuing the gene operation.
S30, saving parameter estimation of the parameter to be identified, and being used for estimating corresponding attitude parameters of the channel at the next moment.
Based on the characteristic parameter identification process, parameter estimation of a certain attitude angle channel of the aircraft can be realized, and other attitude angles can be processed in parallel to realize parameter estimation of a corresponding angle channel. In order to embody the advantages of the scheme in terms of parameter identification precision and efficiency, the embodiment establishes a characteristic model for a four-rotor aircraft attitude control system, and identifies the characteristic model for the pitching attitude angle theta in the attitude ring rolling, pitching and yawing, and the verification method is structured as shown in figure 2. The same signals (selected unit step signals) are input to the characteristic model established by identification and the actual pitch angle attitude model at the same time, and 3 characteristic parameters (alpha) of the channel are drawn 1 (k)、α 2 (k)、β 0 (k) A) identifying change rule, and solving the maximum error by taking the difference of the responses output by the two models; as a control group, a recursive least square method with forgetting factors is selected to perform the same operation on the same system, and the result is compared with the result of the method of the invention to obtain the result as shown in fig. 3-4:
the maximum error between the identification of the genetic algorithm and the output of the original system model is 0.014, and the average error is 3 multiplied by 10 -6 The method comprises the steps of carrying out a first treatment on the surface of the The maximum output error of the comparison recursive least square method is 0.572, and the average error is 5 multiplied by 10 -6 。
The two sets of parameter identification results are shown in fig. 3 and 4 respectively, wherein the final parameter results are basically consistent, but compared with the comparison method, the identification speed of the method is faster, and the process of reaching the final value is more direct.
The result analysis can be used for considering that the feature parameter identification based on the genetic algorithm improves the real-time performance and accuracy of the feature parameter identification, and compared with the traditional online identification method, the feature parameter identification is quicker, and the model is closer to a real object; meanwhile, the identification process is simplified through an improved algorithm, and the challenges of complex parameter selection and calculation capability in actual engineering of the existing method are avoided.
Embodiment two:
the present embodiment provides an aircraft attitude control model parameter identification apparatus based on the first embodiment, for implementing the steps of the aircraft attitude control model parameter identification method in the first embodiment, please refer to fig. 5, the apparatus mainly includes a feature model creation module 51, a parameter identification module 52, and a storage module 53, wherein:
the characteristic model creation module is used for respectively creating the following characteristic models for three attitude angle channels of pitching, rolling and yawing of the aircraft:
X(k+1)＝α 1 (k)X(k)+α 2 (k)X(k-1)+β 0 (k)U(k)； (12)
wherein X (k+1) is the estimated attitude angle at the next moment, X (k) is the actual attitude angle at the current moment, X (k-1) is the actual attitude angle at the previous moment, U (k) is the control moment, and alpha 1 (k)、α 2 (k) And beta 0 (k) Is a parameter to be identified;
defining an objective function according to the characteristic model error; constructing an fitness function based on the inverse of the objective function;
the parameter identification module is used for acquiring the identification parameter estimation output at the previous moment from the storage module, performing binary coding on the identification parameter estimation, and inputting a genetic algorithm model for genetic operation so as to generate a next generation population; calculating the fitness of the population based on the fitness function; judging whether the fitness of the population meets the precision requirement or not; if so, decoding the population to obtain parameter estimation of the parameters to be identified, and sending the parameter estimation to the storage module; if not, updating the evolution algebra; judging whether the maximum evolution generation is reached; if yes, decoding the population; if not, judging whether the current evolution algebra is more than or equal to 2; if not, taking the first generation population as a parent population, and continuing gene operation to generate a next generation population; if the current evolution algebra is more than or equal to 2, judging whether the adaptability of the current population is better than that of the previous generation population; if so, continuing gene operation by taking the current population as a parent population to generate a next generation population; if not, taking the previous generation population as a parent population, and continuing gene operation to generate the next generation population;
the storage module is used for storing parameter estimation of the parameter to be identified so as to be used for estimating corresponding attitude parameters of the channel at the next moment.
Please refer to the description of the first embodiment, and the description is omitted here.
Embodiment III:
the present embodiment provides an aircraft, please refer to fig. 6, which includes the aircraft attitude control model parameter identification apparatus 61 according to the second embodiment. In this embodiment, the aircraft includes, but is not limited to, a quad-rotor drone. Please refer to the description of the second embodiment, and the description is omitted herein.
Embodiment four:
the present embodiment provides a storage medium storing one or more programs executable by one or more processors to implement the steps of the aircraft attitude control model parameter identification method as described in embodiment one. Please refer to the description of the first embodiment, and the description is omitted here.
It will be appreciated by those skilled in the art that the modules or steps of the invention described above may be implemented in a general purpose computing device, they may be centralized on a single computing device, or distributed across a network of computing devices, or they may alternatively be implemented in program code executable by computing devices, such that they may be stored on a computer storage medium (ROM/RAM, magnetic or optical disk) for execution by computing devices, and in some cases, the steps shown or described may be performed in a different order than what is shown or described herein, or they may be individually manufactured as individual integrated circuit modules, or a plurality of modules or steps in them may be manufactured as a single integrated circuit module. Therefore, the present invention is not limited to any specific combination of hardware and software.
The foregoing is a further detailed description of the invention in connection with specific embodiments, and it is not intended that the invention be limited to such description. It will be apparent to those skilled in the art that several simple deductions or substitutions may be made without departing from the spirit of the invention, and these should be considered to be within the scope of the invention.

Claims (7)

1. An aircraft attitude control model parameter identification method is characterized by comprising the following steps:
s10, respectively establishing the following characteristic models for three attitude angle channels of pitching, rolling and yawing of the aircraft:
X(k+1)＝α 1 (k)X(k)+α 2 (k)X(k-1)+β 0 (k)U(k)；
wherein X (k+1) is the estimated attitude angle at the next moment, X (k) is the actual attitude angle at the current moment, X (k-1) is the actual attitude angle at the previous moment, U (k) is the control moment, and alpha 1 (k)、α 2 (k) And beta 0 (k) Is a parameter to be identified;
s11, defining an objective function according to the characteristic model error;
s12, constructing an adaptability function based on the reciprocal of the objective function;
s20, acquiring identification parameter estimation output at the previous moment, performing binary coding on the identification parameter estimation, and inputting a genetic algorithm model for genetic operation to generate a next generation population;
s21, calculating the fitness of the population based on the fitness function;
s22, judging whether the fitness of the population meets the precision requirement or not; if yes, go to step S23; if not, go to step S24;
s23, decoding the population to obtain parameter estimation of the parameters to be identified;
s24, updating the evolution algebra;
s25, judging whether the maximum evolution generation is reached; if yes, go to step S23; if not, go to step S26;
s26, judging whether the current evolution algebra is more than or equal to 2; if yes, go to step S27; if not, go to step S28;
s27, judging whether the fitness of the current population is better than that of the previous generation population; if yes, go to step S291; if not, go to step S292;
s28, taking the first generation population as a parent population, and continuing to perform gene operation;
s291, continuing to perform gene operation by taking the current population as a parent population;
s292, taking the previous generation population as a parent population, and continuing gene operation;
s30, saving parameter estimation of the parameter to be identified, so as to be used for estimating corresponding attitude parameters of the channel at the next moment;
wherein, the step S11 of defining the objective function according to the characteristic model error includes:
the characteristic model error is as follows:
e＝X M -X τ ；
wherein X is M For the actual attitude angle output, X τ For characteristic model attitude angle estimation, the sampling frequency is N, and an objective function is defined as follows:
the fitness function is as follows:
wherein the population isM is population size.
2. The aircraft attitude control model parameter identification method according to claim 1, wherein the genetic manipulation includes a copy manipulation, a crossover manipulation and a mutation manipulation.
3. The method for identifying parameters of an aircraft attitude control model according to claim 2, wherein the copying operation is performed by using a wheel selection algorithm:
wherein Pc is the crossover probability, pm is the mutation probability, f max 、f avg F 'is the maximum fitness and average fitness in the population, f' is the greater fitness in the two intersecting individuals;
the crossing operation is uniform crossing;
individual decision method for mutation operation selection based on mutation probability P m Randomly selecting a corresponding number of mutant individuals in the population to carry out mutation operation.
4. The aircraft attitude control model parameter identification method according to claim 1, further comprising: when the identification parameter estimation output at the last moment does not exist, randomly generating an initial population in a set parameter range for binary coding, and inputting a genetic algorithm model for genetic operation to generate a next generation population.
5. An aircraft attitude control model parameter identification apparatus, comprising:
the characteristic model creation module is used for respectively creating the following characteristic models for three attitude angle channels of pitching, rolling and yawing of the aircraft:
X(k+1)＝α 1 (k)X(k)+α 2 (k)X(k-1)+β 0 (k)U(k)；
wherein X (k+1) is the estimated attitude angle at the next moment, X (k) is the actual attitude angle at the current moment, X (k-1) is the actual attitude angle at the previous moment, U (k) is the control moment, and alpha 1 (k)、α 2 (k) And beta 0 (k) Is a parameter to be identified;
defining an objective function according to the characteristic model error; constructing an fitness function based on the inverse of the objective function;
the parameter identification module is used for acquiring the identification parameter estimation output at the previous moment from the storage module, performing binary coding on the identification parameter estimation, and inputting a genetic algorithm model for genetic operation so as to generate a next generation population; calculating the fitness of the population based on the fitness function; judging whether the fitness of the population meets the precision requirement or not; if so, decoding the population to obtain parameter estimation of the parameters to be identified, and sending the parameter estimation to the storage module; if not, updating the evolution algebra; judging whether the maximum evolution generation is reached; if yes, decoding the population; if not, judging whether the current evolution algebra is more than or equal to 2; if not, taking the first generation population as a parent population, and continuing gene operation to generate a next generation population; if the current evolution algebra is more than or equal to 2, judging whether the adaptability of the current population is better than that of the previous generation population; if so, continuing gene operation by taking the current population as a parent population to generate a next generation population; if not, taking the previous generation population as a parent population, and continuing gene operation to generate the next generation population;
the storage module is used for storing parameter estimation of the parameter to be identified so as to be used for estimating corresponding attitude parameters of the channel at the next moment;
wherein, the characteristic model error is:
e＝X M -X τ ；
wherein X is M For the actual attitude angle output, X τ For characteristic model attitude angle estimation, the sampling frequency is N, and an objective function is defined as follows:
the fitness function is as follows:
wherein the population isM is population size.
6. An aircraft comprising the aircraft attitude control model parameter identification apparatus of claim 5.
7. A storage medium storing one or more programs executable by one or more processors to implement the steps of the aircraft attitude control model parameter identification method according to any one of claims 1 to 4.
CN202111138454.1A 2021-09-27 2021-09-27 Aircraft and attitude control model parameter identification method and device thereof and storage medium Active CN113836810B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111138454.1A CN113836810B (en) 2021-09-27 2021-09-27 Aircraft and attitude control model parameter identification method and device thereof and storage medium

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111138454.1A CN113836810B (en) 2021-09-27 2021-09-27 Aircraft and attitude control model parameter identification method and device thereof and storage medium

Publications (2)

Publication Number Publication Date
CN113836810A CN113836810A (en) 2021-12-24
CN113836810B true CN113836810B (en) 2023-12-19

Family

ID=78970706

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111138454.1A Active CN113836810B (en) 2021-09-27 2021-09-27 Aircraft and attitude control model parameter identification method and device thereof and storage medium

Country Status (1)

Country Link
CN (1) CN113836810B (en)

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108490965A (en) * 2018-04-19 2018-09-04 东华大学 Rotor craft attitude control method based on Genetic Algorithm Optimized Neural Network

Family Cites Families (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2005059656A (en) * 2003-08-08 2005-03-10 Fuji Heavy Ind Ltd Landing control device and method for flying object
US8977484B1 (en) * 2013-08-22 2015-03-10 The Boeing Company Using aircraft trajectory data to infer aircraft intent

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108490965A (en) * 2018-04-19 2018-09-04 东华大学 Rotor craft attitude control method based on Genetic Algorithm Optimized Neural Network

Non-Patent Citations (2)

* Cited by examiner, † Cited by third party
Title
基于自适应遗传算法的小型无人旋翼机系统辨识方法;雷旭升;白浪;洪晔;杜玉虎;;机器人(05);全文 *
姜霞，曾宪琳，孙健，陈杰.多飞行器的分布式油画研究现状与展望.2020,第42卷(第04期),全文. *

Also Published As

Publication number Publication date
CN113836810A (en) 2021-12-24

Similar Documents

Publication Publication Date Title
Al Osman et al. Machine learning in measurement part 2: Uncertainty quantification
CN113705793B (en) Decision variable determination method and device, electronic equipment and medium
CN113302605A (en) Robust and data efficient black box optimization
US20240185025A1 (en) Flexible Parameter Sharing for Multi-Task Learning
Guo et al. A fully-pipelined expectation-maximization engine for Gaussian mixture models
CN112258557B (en) Visual tracking method based on space attention feature aggregation
CN111178486A (en) Hyper-parameter asynchronous parallel search method based on population evolution
CN114580747A (en) Abnormal data prediction method and system based on data correlation and fuzzy system
CN113743474A (en) Digital picture classification method and system based on cooperative semi-supervised convolutional neural network
CN116415177A (en) Classifier parameter identification method based on extreme learning machine
CN116776148A (en) QUIC network abnormal behavior detection method, system and equipment
CN111461353A (en) Model training method and system
CN113836810B (en) Aircraft and attitude control model parameter identification method and device thereof and storage medium
Zhang et al. Noise-aware local model training mechanism for federated learning
Liu et al. Towards understanding acceleration tradeoff between momentum and asynchrony in nonconvex stochastic optimization
Deepika et al. Machine learning algorithms: A background artifact
Gu et al. Parallelizing machine learning optimization algorithms on distributed data-parallel platforms with parameter server
CN115794405A (en) Dynamic resource allocation method of big data processing framework based on SSA-XGboost algorithm
CN115972197A (en) Robot action demonstration learning method based on joint angle trajectory coding
US11726817B2 (en) Scheduling multiple processes with varying delay sensitivity
CN114818281A (en) Section inversion method and system based on convolutional neural network
Niu et al. A Novel Distributed Duration-Aware LSTM for Large Scale Sequential Data Analysis
CN111428876B (en) Image classification method of mixed cavity convolution neural network based on self-walking learning
WO2024066791A1 (en) Data processing method, apparatus and system, and medium and program product
CN109784477B (en) Sampling method and system for contrasting neural network training

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
GR01 Patent grant
GR01 Patent grant