WO2022041317A1 - 无模型自适应控制的改进方法 - Google Patents

无模型自适应控制的改进方法 Download PDF

Info

Publication number
WO2022041317A1
WO2022041317A1 PCT/CN2020/114129 CN2020114129W WO2022041317A1 WO 2022041317 A1 WO2022041317 A1 WO 2022041317A1 CN 2020114129 W CN2020114129 W CN 2020114129W WO 2022041317 A1 WO2022041317 A1 WO 2022041317A1
Authority
WO
WIPO (PCT)
Prior art keywords
control
formula
algorithm
following
mfac
Prior art date
Application number
PCT/CN2020/114129
Other languages
English (en)
French (fr)
Inventor
孙希明
温思歆
杜宪
马艳华
刘小雨
郝育闻
Original Assignee
大连理工大学
大连理工大学人工智能大连研究院
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 大连理工大学, 大连理工大学人工智能大连研究院 filed Critical 大连理工大学
Priority to US17/440,052 priority Critical patent/US20220276620A1/en
Publication of WO2022041317A1 publication Critical patent/WO2022041317A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
    • G05B13/04Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
    • G05B13/04Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators
    • G05B13/042Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric involving the use of models or simulators in which a parameter or coefficient is automatically adjusted to optimise the performance
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B11/00Automatic controllers
    • G05B11/01Automatic controllers electric
    • G05B11/36Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential
    • G05B11/42Automatic 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.
    • GPHYSICS
    • G05CONTROLLING; REGULATING
    • G05BCONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B13/00Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
    • G05B13/02Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
    • G05B13/0205Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric not using a model or a simulator of the controlled system
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10Complex mathematical operations
    • G06F17/15Correlation function computation including computation of convolution operations
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F17/00Digital computing or data processing equipment or methods, specially adapted for specific functions
    • G06F17/10Complex mathematical operations
    • G06F17/16Matrix or vector computation, e.g. matrix-matrix or matrix-vector multiplication, matrix factorization
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06NCOMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N20/00Machine learning

Definitions

  • the invention discloses an improved method for compact dynamic linearization model-free adaptive control based on a multi-input multi-output system, and belongs to the field of control algorithm design.
  • MFAC is a data-driven control method, and its parameter design does not depend on the structure of the control object, that is, no modeling or parameter identification of the control object is required, and the control parameter design is only carried out through the input and output data of the control system.
  • This method was first proposed by Hou Zhongsheng, including a new dynamic linearization method and the concept of pseudo Jacobian matrix (PJM).
  • JM pseudo Jacobian matrix
  • the pseudo Jacobian matrix can be directly estimated from the input and output data.
  • this method has achieved important research results in both theory and application. Previous studies have shown that the MFAC method is easier to use and has better control effects than other DDC methods, such as IFT and VRFT methods.
  • MFAC algorithm such as adaptive iterative learning control, adaptive online learning control and model-free adaptive decoupling control, etc.
  • the adaptive iterative learning control method is used to solve the macroscopic highway traffic density control problem, random packet loss problem, and train station parking control and train trajectory tracking control problems for ramp control.
  • the application of neural network combined with MFAC has also been applied.
  • most previous studies on aero-engine control have focused on model-based control. Therefore, it is of great practical significance to extend MFAC to the field of aero-engine control.
  • the MFAC control strategy uses the input and output data of the system to update the PJM of the system in real time through the parameter estimation algorithm, so that the controller parameters are updated in real time, so that when the flight environment changes seriously, the controller can also perform timely and stable adjustments to the aircraft. control, so as to ensure the safe flight of the aircraft at different flight altitudes and atmospheric conditions.
  • the present invention proposes a compact dynamic linearization model-free adaptive control based on multiple-input multiple-output system.
  • the improved method is suitable for the field of control system design and application, and can be used to improve the performance of the control system.
  • Step A Analyzing the existing model-free adaptive control method based on tight-format dynamic linearization, through the experimental results, it is concluded that there are deficiencies in response time and stability in the application process;
  • the multiple-input multiple-output discrete-time nonlinear system is represented as follows:
  • u(k) and y(k) are the system input and system output at time k, respectively;
  • n y and n u are two unknown integers;
  • formula (1) is expressed in the form of the following CFDL data model:
  • ⁇ c (k) as the pseudo Jacobian of the system should be a diagonally dominant matrix, satisfying the following conditions:
  • the constants of , i and j are the row and column indices of the matrix, and the signs of all elements in ⁇ c (k) remain unchanged for any time k;
  • control input criterion function is shown in formula (3):
  • ⁇ >0 represents the weight factor, which is used to punish the change of the excessive control input
  • y * (k+1) is the desired output signal
  • is the weight factor, which is used to punish the excessive change of the PJM estimated value, is the estimated value of ⁇ c (k);
  • control parameter estimation algorithm described above needs to perform parameter estimation at each k, and then the control input at that moment can be given; however, the calculation of the parameter estimation algorithm takes a certain amount of time, resulting in a slow system response, resulting in the requirement for the control period
  • the control algorithm is limited in use, and from the experimental results, the system oscillates greatly under non-ideal conditions;
  • Step B Consider the following improvement plan based on the above problems of slow response and oscillation
  • um (k)′ is the output of the MFAC controller
  • ⁇ up (k) is the output of the proportional controller
  • u_max and u_min are the upper and lower limits of the actuator.
  • Step C for the improved control algorithm described above, through theoretical derivation, analyze the convergence of the tracking error and the stability of the bounded input and bounded output;
  • Equation (17) is equivalent to Equation (18);
  • Equation (30) is satisfied, under this condition, e(k) shows a downward trend.
  • Step D Apply the above control algorithm to the control of the aero-engine model, and select three different situations to compare the results to verify the effectiveness and superiority of the control algorithm; first, compare the MFAC+Kp, CFDL- The control effects of MFAC and PID illustrate the effectiveness of the improved controller; then, different heights and different delays are selected to compare the control effects to illustrate the superiority of the controller.
  • control effects of different algorithms are compared under standard conditions.
  • the second case is used to illustrate that the controller can adaptively control the wide flight envelope of the aircraft, and select different flight heights to analyze the control effect.
  • the results are shown in Figure 3. From the simulation results, the MFAC+Kp algorithm can achieve stable control at different flight heights. The higher the flight altitude, the greater the overshoot, but the algorithm can still quickly stabilize the system output and has strong adaptive ability. In addition, compared with the MFAC control effect under the same conditions, it shows that the control algorithm has stronger stability.
  • FIG. 1 is a configuration diagram of a controller.
  • Figure 2 is a comparison chart of the effects of three control algorithms of MFAC+Kp, MFAC and PID.
  • Figure 3 is a comparison of control effects at different flight heights.
  • Figure 4 is a comparison of control effects under different delays.
  • the structural block diagram of the improved control algorithm of the present invention is shown in Figure 1, wherein the controller mainly includes three parts: MFAC, proportional control and anti-saturation control.
  • the control algorithm combines the advantages of the three algorithms, and can achieve stable and fast control even for very complex nonlinear models, and has good robustness.
  • (1) MFAC algorithm the parameters of the control algorithm are updated by the estimation algorithm at each sampling time point, so that the control algorithm can be adaptively changed to achieve a good control effect on the control object, and has a certain robustness.
  • the response time of the controller is slow and susceptible to disturbances.
  • it is considered to add a proportional control link on the basis of the control algorithm.
  • Proportional control algorithm simple operation, short time-consuming, can reduce steady-state error, speed up control response, make up for the deficiency of MFAC algorithm, and improve control performance.
  • Anti-saturation algorithm due to the upper and lower limits of the actuator in the control system, the output of the control algorithm may exceed the execution capability of the actuator, causing the actuator to fall into saturation, which will affect the response speed and control of the controller. precision.
  • the anti-saturation algorithm can stop the operation when the actuator reaches saturation, so that when the control algorithm gives a normal command, the actuator can respond as quickly as when it is not saturated.
  • the basic criteria for measuring the control algorithm are the accuracy, stability and rapidity of its control.
  • the present invention also has anti-saturation properties while satisfying the above criteria.
  • the improved method for model-free adaptive control of the present invention mainly has the following advantages:
  • Step A Analyzing the existing model-free adaptive control method based on tight-format dynamic linearization, through the experimental results, it is concluded that there are deficiencies in response time and stability in the application process;
  • the multiple-input multiple-output discrete-time nonlinear system is represented as follows:
  • u(k) and y(k) are the system input and system output at time k, respectively;
  • n y and n u are two unknown integers;
  • formula (1) is expressed in the form of the following CFDL data model:
  • ⁇ c (k) as the pseudo Jacobian of the system should be a diagonally dominant matrix, satisfying the following conditions:
  • the constants of , i and j are the row and column indices of the matrix, and the signs of all elements in ⁇ c (k) remain unchanged for any time k;
  • control input criterion function is shown in formula (3):
  • ⁇ >0 represents the weight factor, which is used to punish the change of the excessive control input
  • y * (k+1) is the desired output signal
  • is the weight factor, which is used to punish the excessive change of the PJM estimated value, is the estimated value of ⁇ c (k);
  • control parameter estimation algorithm described above needs to perform parameter estimation at each k, and then the control input at that moment can be given; however, the calculation of the parameter estimation algorithm takes a certain amount of time, resulting in a slow system response, resulting in the requirement for the control cycle
  • the control algorithm is limited in use, and from the experimental results, the system oscillates greatly under non-ideal conditions;
  • Step B Consider the following improvement plan based on the above problems of slow response and oscillation
  • um (k)′ is the output of the MFAC controller
  • ⁇ up (k) is the output of the proportional controller
  • the following anti-saturation algorithm is proposed as part of the proposed control algorithm: when the actuator is at the upper saturation limit and there is still a growing trend, or when the actuator is at the lower saturation limit and is still falling, the integrator will stop updating; otherwise, The integrator is functioning properly; that is, in the case of saturation, only the integration operations that help reduce the saturation are performed; it is represented by the following formula:
  • u_max and u_min are the upper and lower limits of the actuator
  • Step C for the improved control algorithm described above, through theoretical derivation, analyze the convergence of the tracking error and the stability of the bounded input and bounded output;
  • Equation (17) is equivalent to Equation (18);
  • Equation (30) is satisfied, under this condition, e(k) shows a downward trend.
  • Step D Apply the above control algorithm to the control of the aero-engine model, and select three different situations to compare the results to verify the effectiveness and superiority of the control algorithm; first, compare the MFAC+Kp, CFDL- The control effects of MFAC and PID illustrate the effectiveness of the improved controller; then, different heights and different delays are selected to compare the control effects to illustrate the superiority of the controller.
  • control effects of different algorithms are compared under standard conditions.
  • the second case is used to illustrate that the controller can adaptively control the wide flight envelope of the aircraft, and select different flight heights to analyze the control effect.
  • the results are shown in Figure 3. From the simulation results, the MFAC+Kp algorithm can achieve stable control at different flight heights. The higher the flight altitude, the greater the overshoot, but the algorithm can still quickly stabilize the system output and has strong adaptive ability. In addition, compared with the MFAC control effect under the same conditions, it shows that the control algorithm has stronger stability.
  • the improved method for model-free adaptive control of the present invention proposes a new model-free adaptive control method, which improves MFAC overshoot and oscillation by adding proportional control. At the same time, it also integrates the idea of integral anti-windup to improve the control performance.
  • the improved control algorithm has tracking error convergence and BIBO stability under the condition of satisfying the assumptions.
  • the improved MFAC is applied to the control of the aero-engine model, and three experiments are carried out from different angles to verify the anti-saturation, rapidity and stability of the control algorithm of the present invention under different flight altitudes and different delays. The results are better than the MFAC algorithm and the PID algorithm. The results show that the control algorithm proposed in this paper has a stable and fast control effect on the aero-engine control system, which verifies the effectiveness of the algorithm.

Landscapes

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Pure & Applied Mathematics (AREA)
  • Data Mining & Analysis (AREA)
  • Theoretical Computer Science (AREA)
  • Mathematical Optimization (AREA)
  • Mathematical Analysis (AREA)
  • Computational Mathematics (AREA)
  • Software Systems (AREA)
  • Computing Systems (AREA)
  • Algebra (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Automation & Control Theory (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Evolutionary Computation (AREA)
  • Artificial Intelligence (AREA)
  • Health & Medical Sciences (AREA)
  • Medical Informatics (AREA)
  • Feedback Control In General (AREA)

Abstract

本发明公开一种无模型自适应控制的改进方法,基于多输入多输出系统的紧格式动态线性化无模型自适应控制的改进方法,属于控制算法设计领域。首先,在CFDL-MFAC中加入比例控制,用来改善原控制系统的响应速度慢、超调大的问题;其次,在以上控制结构中加入执行机构抗饱和控制算法,使得执行机构在达到上限或下限饱和时不再进行超限运算,当控制指令再次进入非饱和区时,执行机构能够快速做出控制响应,提高系统的控制精度;接着,通过严格的分析证明了改进的控制算法可以保证一定条件下跟踪误差和BIBO稳定性;最后,将上述控制算法应用于航空发动机控制系统,通过数值实验可以得出上述控制算法的有效性和优越性。

Description

无模型自适应控制的改进方法 技术领域
本发明公开一种基于多输入多输出系统的紧格式动态线性化无模型自适应控制的改进方法,属于控制算法设计领域。
背景技术
随着技术创新和产业进步,人们对飞机的安全,稳定,高效控制的要求越来越高,基于模型的控制理念受到建模精度的极大影响。此外,由于未建模的动力学问题和发动机机械磨损,通过数学建模获得的模型将逐渐偏离实际的控制对象。这些控制盲点导致控制效果随着使用时间的延长而越来越差。因此,无模型自适应控制(MFAC)算法应运而生。
MFAC是一种数据驱动的控制方法,其参数设计不依赖于控制对象结构,即无需对控制对象进行建模或参数辨识,而只通过控制系统的输入输出数据进行控制参数设计。该方法最早由侯忠生提出,包括新的动态线性化方法和伪Jacobian矩阵(PJM)的概念。其中伪Jacobian矩阵可由输入输出数据直接估计得出。在过去的二十年中,该方法在理论和应用方面均已经取得了重要的研究成果。已有研究表明,MFAC方法比其它DDC方法,如IFT和VRFT方法,更易于使用且具有更好的控制效果。
最近,许多文献都提到了MFAC算法的扩展研究和应用,例如自适应迭代学习控制、自适应在线学习控制和无模型自适应解耦控制等等。另外,近五年,MFAC算法与迭代学习相结合的研究热潮逐渐增加。例如,采用自适应迭代学习控制方法来解决坡道控制的宏观公路交通密度控制问题、随机丢包问题和火车站停车控制和火车轨迹跟踪控制问题。此外,神经网络与MFAC相结合的应用也已得到应用。然而,以往大多数航空发动机控制研究都集中在基于模型的控制上。因此,将MFAC扩展到航空发动机控制领域具有重要的现实意义。MFAC控制策略通过参数估计算法,运用系统的输入和输出数据实时更新系统的PJM,从而使得控制器参数得到实时更新,以便在飞行环境发生严重变化时,控制器也可以对飞机进行及时、稳定的控制,从而确保飞机在不同的飞行高度和大气环境下安全飞行。
发明内容
针对现有基于紧格式动态线性化无模型自适应控制方法在复杂模型上的离散仿真应用存在的不足,本发明提出一种基于多输入多输出系统的紧格式动态线性化无模型自适应控制的改进方法,适用于控制系统设计与应用领域,可用于提高控制系统性能,主要解决无模型自适应控制方法中存在的响应速度慢、超调大和执行机构饱和问题。
本发明的技术方案:
一种基于多输入多输出系统的紧格式动态线性化无模型自适应控制(CFDL-MFAC)的 改进方法,步骤如下:
步骤A、分析已有的基于紧格式动态线性化无模型自适应控制方法,通过实验结果得出,其应用过程中在响应时间和稳定性方面存在不足;
多输入多输出离散时间非线性系统表示如下形式:
y(k+1)=f(y(k),…,y(k-n y),u(k),…,u(k-n u))       (1)
其中,u(k)和y(k)分别是k时刻的系统输入和系统输出;n y和n u是两个未知整数;f(…)=(f 1(…),…,f m(…))是一个未知的非线性函数;
当f存在连续偏导数的条件且公式(1)满足广义Lipschiz条件时,公式(1)表示为如下的CFDL数据模型的形式:
Δy(k+1)=Φ c(k)Δu(k)(2)
其中,
Figure PCTCN2020114129-appb-000001
首先提出如下假设:
假设1:Φ c(k)作为系统的伪Jacobian应为对角占优矩阵,满足如下条件:|φ ij|≤b 1,b 2≤|φ ii(k)|≤αb 2,α≥1,b 2>b 1(2α+1)(m-1),i=1,...,m,j=1,...,m,i≠j,设b 1、b 2为有界的常数,i和j为矩阵的行、列索引,且Φ c(k)中所有元素的符号对任意时刻k保持不变;
控制输入准则函数如式(3)所示:
J(u(k)=||y *(k+1)-y(k+1)|| 2+λ||u(k)-u(k-1)|| 2       (3)
其中,λ>0代表权重因子,用于惩罚过度控制输入量的变化;y *(k+1)是期望的输出信号;
将式(2)带入式(3),对u(k)求导,并令其等于0,得控制输入算法如下:
Figure PCTCN2020114129-appb-000002
考虑如下参数估计算法准则函数:
Figure PCTCN2020114129-appb-000003
其中,μ为权重因子,用来惩罚PJM估计值的过大变化,
Figure PCTCN2020114129-appb-000004
为Φ c(k)的估计值;
将式(5)对Φ c(k)求导,并另其等于0,得参数估计算法如下:
Figure PCTCN2020114129-appb-000005
以上所述控制参数估计算法在每个k都要进行参数估计,而后才能给出该时刻的控制输入;然而参数估计算法的计算需要占用一定的时间,导致系统响应变慢,导致对于控制周期要求较小的系统,该控制算法使用受限,且从实验结果来看系统在非理想条件下震荡较大;
步骤B、基于以上存在响应慢和存在震荡的问题考虑如下改进方案;
Δu(k)=Δu m(k)′+Δu p(k)        (7)
其中,u m(k)′是MFAC控制器输出,Δu p(k)是比例控制器输出,用以下公式表示:
Figure PCTCN2020114129-appb-000006
Δu p(k)=βK(y *(k+1)-y(k))-βK(y *(k)-y(k-1))   (9)
并提出以下抗饱和算法作为所提出的控制算法的一部分:当致动器处于饱和上限且仍有增长趋势时,或者当致动器处于饱和下限且仍在下降时,将停止更新积分器;否则,积分器正常工作;也就是说,在饱和的情况下,只执行有助于降低饱和程度的积分运算;它由以下公式表示:
Figure PCTCN2020114129-appb-000007
其中,u_max和u_min是致动器的上限和下限。
基于公式(6)、(7)、(8)、(9),提出以下控制方案:
Figure PCTCN2020114129-appb-000008
如果
Figure PCTCN2020114129-appb-000009
Figure PCTCN2020114129-appb-000010
Figure PCTCN2020114129-appb-000011
Figure PCTCN2020114129-appb-000012
如果
Figure PCTCN2020114129-appb-000013
Figure PCTCN2020114129-appb-000014
Figure PCTCN2020114129-appb-000015
其中,
Figure PCTCN2020114129-appb-000016
Figure PCTCN2020114129-appb-000017
的初值;
步骤C、对于以上所述的改进控制算法,通过理论推导,分析跟踪误差的收敛性和有界输入有界输出的稳定性;
首先,定义以下系统输出误差:
e(k)=y *(k)-y(k)(15)
将公式(2)和公式(14)替换为公式(15),当f(k)=1时,得:
Figure PCTCN2020114129-appb-000018
Figure PCTCN2020114129-appb-000019
其中z是矩阵
Figure PCTCN2020114129-appb-000020
的特征值,D j,j=1,2,...,m是Gerschgorin圆盘;
公式(17)等价于公式(18);
Figure PCTCN2020114129-appb-000021
由重置算法(12)和(13)得,
Figure PCTCN2020114129-appb-000022
Figure PCTCN2020114129-appb-000023
由假设1得,|φ ij|≤b 1,b 2≤|φ ii(k)|≤ab 2,i=1,…,m;j=1,...,m;i≠j;
由以上条件得下式
Figure PCTCN2020114129-appb-000024
Figure PCTCN2020114129-appb-000025
Figure PCTCN2020114129-appb-000026
由重置算法公式(11)和假设1得,
Figure PCTCN2020114129-appb-000027
因此存在λ min,使得当λ>λ min时,下式成立。
Figure PCTCN2020114129-appb-000028
因此选择0<ρ≤1和λ>λ min,使得
Figure PCTCN2020114129-appb-000029
对于任意的λ>λ min,显然下式成立
Figure PCTCN2020114129-appb-000030
由式(21)、(23)、(24),知
Figure PCTCN2020114129-appb-000031
由式(18)和式(24)得,
Figure PCTCN2020114129-appb-000032
其中s(M)是M的谱半径;
Figure PCTCN2020114129-appb-000033
B=||βΦ c(k)K)|| v,由矩阵谱半径的结论得,存在一个 任意小的正数ε 1,使得
Figure PCTCN2020114129-appb-000034
其中,||M|| v是矩阵M的相容范数;
存在β使得B满足如下不等式:
1>1-A≥M 11>B>0       (28)
由式(16)和式(28)得:
||e(k+1)|| υ≤A||e(k)|| υ+B||e(k-1)|| υ<(1-B)||e(k)|| υ+B||e(k-1)|| υ(29)
移项后得:
||e(k+1)|| υ-||e(k)|| υ<-B(||e(k)|| υ-||e(k-1)|| υ)(30)
基于式(30),从以下四个方面讨论e(k)的形式:
1.当||e(k+1)|| v>||e(k)|| v且||e(k)|| v>||e(k-1)|| v时,可得
||e(k+1)|| v-||e(k)|| v>-B(||e(k)|| v-||e(k-1)|| v)           (31)不等式结果与式(30)相反,因此,此假设不存在。
2.当||e(k+1)|| v>||e(k)|| v且||e(k)|| v<||e(k-1)|| v时,由式(30)得:
Figure PCTCN2020114129-appb-000035
即在三个相邻采样时刻,e(k)的减小量大于增加量,所以总体趋势呈下降趋势。
3.当||e(k+1)|| v<||e(k)|| v且||e(k)|| v<||e(k-1)|| v时,可得
Figure PCTCN2020114129-appb-000036
满足式(30),此条件下,e(k)呈下降趋势。
4.当||e(k+1)|| v<||e(k)|| v且||e(k)|| v>||e(k-1)|| v时,由式(30)可得,该情况有可能存在。在k+2时刻存在两种可能性:如果存在||e(k+2)|| v>||e(k+1)|| v,我们可以得到与第二种情况相同的结论;如果||e(k+2)|| v<||e(k+1)|| v,我们可以得到与第三种情况相同的结论。简而言之,在这种情况下,e(k)仍然呈下降趋势。
当f(k)=0时,上述证明方法依然适用。综上所述,误差e(k)总体呈下降趋势。因此,误差的收敛性得证。
步骤D、将以上控制算法应用于航空发动机模型的控制,选取三种不同的情况进行结果对比,用以验证控制算法的有效性和优越性;首先,比较标准条件下的MFAC+Kp、CFDL-MFAC和PID的控制效果,说明了所述改进控制器的有效性;然后,选择不同的高度和不同的延迟来比较控制效果,以说明控制器的优越性。
第一种情况,在标准条件下进行不同算法的控制效果对比。在飞行高度H=0、Ma=0、无噪声和无延迟的标称条件下三种算法的控制效果如图2所示。可以看出,MFAC+Kp算法的上升时间介于MFAC算法和PID算法之间,但其优于MFAC算法的优点是超调量小,满足控制算法对性能的严格稳定性要求。
第二种情况,用于说明控制器可以自适应地控制飞机的宽飞行包线,选取不同的飞行高度进行控制效果分析,结果如图3所示。从仿真结果来看,MFAC+Kp算法可以实现不同飞行高度的稳定控制。飞行高度越高,超调越大,但该算法仍能快速稳定系统输出,具有较强的自适应能力。此外,与相同条件下的MFAC控制效果相比,说明该控制算法具有更强的稳定性。
[根据细则91更正 10.11.2020] 
第三种情况,验证了控制算法在存在延时的情况下对模型的稳定控制。在H=10和Ma=1的飞行条件下,选择四种不同大小的延时进行模拟。结果表明,MFAC+Kp算法能够对不同程度的延时进行稳定快速的控制。从图4中可看出,当执行结构饱和时,所提出的抗饱和算法可以使模型快速脱离饱和区,但在相同条件下,MFAC算法由于在饱和后仍然继续运算,故需要很长时间才能脱离饱和区域。
本发明的有益效果:
(1)CFDL-MFAC+Kp控制算法,提高了原MFAC的响应速度和鲁棒性。并在现有MFAC稳定性证明的基础上进行了理论分析,证明改进算法的稳定性。
(2)在上述控制算法结构下,融入执行机构抗饱和算法被同时考虑在内,通过实验分析验证了上述控制算法的抗饱和效果。
附图、表说明
图1是控制器结构图。
图2是MFAC+Kp、MFAC、PID三种控制算法效果对比图。
图3是不同飞行高度下的控制效果对比。
图4是不同延时下的控制效果对比。
[根据细则91更正 10.11.2020] 
具体实施方式
为使本发明提出的技术方案、解决的技术问题更加清晰,以下结合附图对本发明的技术方案进行具体阐述。
本发明所述改进控制算法结构框图如图1所示,其中控制器主要包括三部分:MFAC、比例控制和抗饱和控制。该控制算法融合了三种算法的优点,即使对十分复杂的非线性模型也能实现稳定、快速的控制,且具有较好的鲁棒性。
该控制算法各部分的具体组成如下:
(1)MFAC算法,在每一个采样时间点都会由估计算法更新控制算法的参数,使得控制算法能够自适应地改变以达到对控制对象的良好控制效果,具有一定的鲁棒性。但由于其估计算法的加入,控制器的响应时间变慢且易受干扰影响。为了满足对控制器快速性及鲁棒性的需求,考虑在该控制算法的基础上加入比例控制环节。
(2)比例控制算法,运算简单,耗时短,且能够减小稳态误差,加快控制响应,弥补MFAC算法的不足,改善控制性能。
(3)抗饱和算法,由于控制系统中,执行机构的上、下限限制,因此控制算法的输出有可能超出执行机构的执行能力,使得执行机构陷入饱和,这将影响控制器的响应速度和控制精度。抗饱和算法可在执行机构达到饱和时,停止运算,使得控制算法在给出正常指令时,执行机构能够像未陷入饱和时一样快速做出响应。
衡量控制算法的基本标准是其控制的准确性、稳定性和快速性,本发明在满足以上标准的同时还兼具抗饱和性,本发明无模型自适应控制的改进方法主要具有以下优点:
(1)准确性好。由图3和图4可知,本发明所述控制算法能够在不同的高度和不同的延时条件下达到良好的控制效果,说明该算法具备良好的准确性。
(2)稳定性优。由图2、图3和图4可知,通过相同情况下与MFAC和PID算法的对比,本发明所述控制算法具有优异的稳定性,在所述的不同飞行高度和不同延时情况下都能实现稳定的控制,且稳定性明显优于原MFAC算法。
(3)快速性强。图3和图4可知,通过相同情况下与MFAC和PID算法的对比,本发明所述控制算法具有优异的快速性,在所述的不同飞行高度和不同延时情况下都能实现稳定的控制,且稳定性明显优于原MFAC算法。
[根据细则91更正 10.11.2020] 
(4)抗饱和性好。由图4可知,本发明所述抗饱和算法在执行机构达到饱和后会停止继续累加,防止进一步陷入饱和,并在控制器输出正常值后,执行机构能够快速响应,与原MFAC算法相比响应速度更快。
以下即为本发明中提出的无模型自适应控制的改进方法,具体步骤如下:
步骤A、分析已有的基于紧格式动态线性化无模型自适应控制方法,通过实验结果得出,其应用过程中在响应时间和稳定性方面存在不足;
多输入多输出离散时间非线性系统表示如下形式:
y(k+1)=f(y(k),…,y(k-n y),u(k),…,u(k-n u))     (1)
其中,u(k)和y(k)分别是k时刻的系统输入和系统输出;n y和n u是两个未知整数;f(…)=(f 1(…),…,f m(…))是一个未知的非线性函数;
当f存在连续偏导数的条件且公式(1)满足广义Lipschiz条件时,公式(1)表示为如下的CFDL数据模型的形式:
Δy(k+1)=Φ c(k)Δu(k)(2)
其中,
Figure PCTCN2020114129-appb-000037
首先提出如下假设:
假设1:Φ c(k)作为系统的伪Jacobian应为对角占优矩阵,满足如下条件:|φ ij|≤b 1,b 2≤|φ ii(k)|≤αb 2,α≥1,b 2>b 1(2α+1)(m-1),i=1,...,m,j=1,...,m,i≠j,设b 1、b 2为有界的常数,i和j为矩阵的行、列索引,且Φ c(k)中所有元素的符号对任意时刻k保持不变;
控制输入准则函数如式(3)所示:
J(u(k))=||y *(k+1)-y(k+1)|| 2+λ||u(k)-u(k-1)|| 2           (3)
其中,λ>0代表权重因子,用于惩罚过度控制输入量的变化;y *(k+1)是期望的输出信号;
将式(2)带入式(3),对u(k)求导,并令其等于0,得控制输入算法如下:
Figure PCTCN2020114129-appb-000038
考虑如下参数估计算法准则函数:
Figure PCTCN2020114129-appb-000039
其中,μ为权重因子,用来惩罚PJM估计值的过大变化,
Figure PCTCN2020114129-appb-000040
为Φ c(k)的估计值;
将式(5)对Φ c(k)求导,并另其等于0,得参数估计算法如下:
Figure PCTCN2020114129-appb-000041
以上所述控制参数估计算法在每个k都要进行参数估计,而后才能给出该时刻的控制输入;然而参数估计算法的计算需要占用一定的时间,导致系统响应变慢,导致对于控制周期要求较小的系统,该控制算法使用受限,且从实验结果来看系统在非理想条件下震荡较大;
步骤B、基于以上存在响应慢和存在震荡的问题考虑如下改进方案;
Δu(k)=Δu m(k′)+Δu p(k)        (7)
其中,u m(k)′是MFAC控制器输出,Δu p(k)是比例控制器输出,用以下公式表示:
Figure PCTCN2020114129-appb-000042
Δu p(k)=βK(y *(k+1)-y(k))-βK(y *(k)-y(k-1))    (9)
基于公式(6)和公式(7),提出以下控制方案:
Figure PCTCN2020114129-appb-000043
如果
Figure PCTCN2020114129-appb-000044
Figure PCTCN2020114129-appb-000045
Figure PCTCN2020114129-appb-000046
Figure PCTCN2020114129-appb-000047
如果
Figure PCTCN2020114129-appb-000048
Figure PCTCN2020114129-appb-000049
Figure PCTCN2020114129-appb-000050
其中,
Figure PCTCN2020114129-appb-000051
Figure PCTCN2020114129-appb-000052
的初值。
提出以下抗饱和算法作为所提出的控制算法的一部分:当致动器处于饱和上限且仍有增长趋势时,或者当致动器处于饱和下限且仍在下降时,将停止更新积分器;否则,积分器正常工作;也就是说,在饱和的情况下,只执行有助于降低饱和程度的积分运算;它由以下公式表示:
Δu m(k)′=Δu m(k)f(k)        (13)
Figure PCTCN2020114129-appb-000053
其中,u_max和u_min是致动器的上限和下限;
步骤C、对于以上所述的改进控制算法,通过理论推导,分析跟踪误差的收敛性和有界输入有界输出的稳定性;
首先,定义以下系统输出误差:
e(k)=y *(k)-y(k)(15)
将公式(2)和公式(12)替换为公式(15),当f(k)=1时,得:
Figure PCTCN2020114129-appb-000054
Figure PCTCN2020114129-appb-000055
其中z是矩阵
Figure PCTCN2020114129-appb-000056
的特征值,D j,j=1,2,...,m是Gerschgorin圆盘;
公式(17)等价于公式(18);
Figure PCTCN2020114129-appb-000057
由重置算法(10)和(11)得,
Figure PCTCN2020114129-appb-000058
Figure PCTCN2020114129-appb-000059
由假设1得,|φ ij|≤b 1,b 2≤|φ ii(k)|≤ab 2,i=1,…,m;j=1,...,m;i≠j;
由以上条件得下式
Figure PCTCN2020114129-appb-000060
Figure PCTCN2020114129-appb-000061
Figure PCTCN2020114129-appb-000062
由重置算法公式(11)和假设1得,
Figure PCTCN2020114129-appb-000063
因此存在λ min,使得当λ>λ min时,下式成立。
Figure PCTCN2020114129-appb-000064
因此选择0<ρ≤1和λ>λ min,使得
Figure PCTCN2020114129-appb-000065
对于任意的λ>λ min,显然下式成立
Figure PCTCN2020114129-appb-000066
由式(21)、(23)、(24),知
Figure PCTCN2020114129-appb-000067
由式(18)和式(24)得,
Figure PCTCN2020114129-appb-000068
其中s(M)是M的谱半径;
Figure PCTCN2020114129-appb-000069
B=||βΦ c(k)K)|| v,由矩阵谱半径的结论得,存在一个 任意小的正数ε 1,使得
Figure PCTCN2020114129-appb-000070
其中,||M|| v是矩阵M的相容范数;
存在β使得B满足如下不等式:
1>1-A≥M 11>B>0                (28)
由式(16)和式(28)得:
||e(k+1)|| υ≤A||e(k)|| υ+B||e(k-1)|| υ<(1-B)||e(k)|| υ+B||e(k-1)|| υ(29)
移项后得:
||e(k+1)|| υ-||e(k)|| υ<-B(||e(k)|| υ-||e(k-1)|| υ)(30)
基于式(30),从以下四个方面讨论e(k)的形式:
1.当||e(k+1)|| v>||e(k)|| v且||e(k)|| v>||e(k-1)|| v时,可得
||e(k+1)|| v-||e(k)|| v>-B(||e(k)|| v-||e(k-1)|| v)     (31)
不等式结果与式(30)相反,因此,此假设不存在。
2.当||e(k+1)|| v>||e(k)|| v且||e(k)|| v<||e(k-1)|| v时,由式(30)得:
Figure PCTCN2020114129-appb-000071
即在三个相邻采样时刻,e(k)的减小量大于增加量,所以总体趋势呈下降趋势。
3.当||e(k+1)|| v<||e(k)|| v且||e(k)|| v<||e(k-1)|| v时,可得
Figure PCTCN2020114129-appb-000072
满足式(30),此条件下,e(k)呈下降趋势。
4.当||e(k+1)|| v<||e(k)|| v且||e(k)|| v>||e(k-1)|| v时,由式(30)可得,该情况有可能存在。在k+2时刻存在两种可能性:如果存在||e(k+2)|| v>||e(k+1)|| v,我们可以得到与第二种情况相同的结论;如果||e(k+2)|| v<||e(k+1)|| v,我们可以得到与第三种情况相同的结论。简而言之,在这种情况下,e(k)仍然呈下降趋势。
当f(k)=0时,上述证明方法依然适用。综上所述,误差e(k)总体呈下降趋势。因此,误差的收敛性得证。
步骤D、将以上控制算法应用于航空发动机模型的控制,选取三种不同的情况进行结果对比,用以验证控制算法的有效性和优越性;首先,比较标准条件下的MFAC+Kp、CFDL-MFAC和PID的控制效果,说明了所述改进控制器的有效性;然后,选择不同的高度和不同的延迟来比较控制效果,以说明控制器的优越性。
第一种情况,在标准条件下进行不同算法的控制效果对比。在飞行高度H=0、Ma=0、无噪声和无延迟的标称条件下三种算法的控制效果如图2所示。可以看出,MFAC+Kp算法的上升时间介于MFAC算法和PID算法之间,但其优于MFAC算法的优点是超调量小,满足控制算法对性能的严格稳定性要求。
第二种情况,用于说明控制器可以自适应地控制飞机的宽飞行包线,选取不同的飞行高度进行控制效果分析,结果如图3所示。从仿真结果来看,MFAC+Kp算法可以实现不同飞行高度的稳定控制。飞行高度越高,超调越大,但该算法仍能快速稳定系统输出,具有较强的自适应能力。此外,与相同条件下的MFAC控制效果相比,说明该控制算法具有更强的稳定性。
[根据细则91更正 10.11.2020] 
第三种情况,验证了控制算法在存在延时的情况下对模型的稳定控制。在H=10和Ma=1饿飞行条件下,选择四种不同大小的延时进行模拟。结果表明,MFAC+Kp算法能够对不同程度的延时进行稳定快速的控制。从图4中可以看出,当执行结构饱和时,所提出的抗饱和算法可以使模型快速脱离饱和区,但在相同条件下,MFAC算法由于在饱和后仍然继续运算,故需要很长时间才能脱离饱和区域。
综上,本发明无模型自适应控制的改进方法提出了一种新的无模型自适应控制方法,通过增加比例控制来改善MFAC超调和振荡。同时,它还集成了积分抗饱和的思想来提高了控制性能。通过严格分析,证明了改进后的控制算法在满足假设的条件下具有跟踪误差收敛性和BIBO稳定性。最后,将改进的MFAC应用于航空发动机模型的控制,从不同角度进行了三个实验,验证了本发明所述控制算法在不同飞行高度和不同延时情况下的抗饱和性、快速性和稳定性,结果优于MFAC算法和PID算法。结果表明,本文提出的控制算法对航空发动机控制系统具有稳定快速的控制效果,验证了算法的有效性。
需要说明的是,本领域的技术人员应当理解:以上各实施例仅用以说明本发明的技术方案,而非对其限制,在不同实施例中出现的不同技术特征可以进行组合,以取得有益效果。本领域的研究人员在说明书、权利要求书的基础上结合附图应能理解并实现所揭示的实施例的其他变化的实施例。需要指出的是,以上已参照前述各实施例对本发明进行了详细的说明, 其对前述各实施例提到的技术方案进行修改,或者对其中部分或者全部技术特征进行等同替换,并不使相应技术方案的本质脱离本发明各实施例技术方案的范围。

Claims (1)

  1. 一种无模型自适应控制的改进方法,其特征在于,步骤如下:
    步骤A、分析已有的基于紧格式动态线性化无模型自适应控制方法,通过实验结果得出,其应用过程中在响应时间和稳定性方面存在不足;
    多输入多输出离散时间非线性系统表示如下形式:
    y(k+1)=f(y(k),…,y(k-n y),u(k),…,u(k-n u))    (I)
    其中,u(k)和y(k)分别是k时刻的系统输入和系统输出;n y和n u是两个未知整数;f(…)=(f 1(…),…,f m(…))是一个未知的非线性函数;
    当f存在连续偏导数的条件且公式(1)满足广义Lipschiz条件时,公式(1)表示为如下的CFDL数据模型的形式:
    Δy(k+1)=Φ c(k)Δu(k)    (2)
    其中,
    Figure PCTCN2020114129-appb-100001
    首先提出如下假设:
    假设1:Φ c(k)作为系统的伪Jacobian应为对角占优矩阵,满足如下条件:|φ ij|≤b 1,b 2≤|φ ii(k)|≤αb 2,α≥1,b 2>b 1(2α+1)(m-1),i=1,...,m,j=1,...,m,i≠j,设b 1、b 2为有界的常数,i和j为矩阵的行、列索引,且Φ c(k)中所有元素的符号对任意时刻k保持不变;
    控制输入准则函数如式(3)所示:
    Figure PCTCN2020114129-appb-100002
    其中,λ>0代表权重因子,用于惩罚过度控制输入量的变化;y *(k+1)是期望的输出信号;
    将式(2)带入式(3),对u(k)求导,并令其等于0,得控制输入算法如下:
    Figure PCTCN2020114129-appb-100003
    考虑如下参数估计算法准则函数:
    Figure PCTCN2020114129-appb-100004
    其中,μ为权重因子,用来惩罚PJM估计值的过大变化,
    Figure PCTCN2020114129-appb-100005
    为Φ c(k)的估计值;
    将式(5)对Φ c(k)求导,并另其等于0,得参数估计算法如下:
    Figure PCTCN2020114129-appb-100006
    以上所述控制参数估计算法在每个k都要进行参数估计,而后才能给出该时刻的控制输入;然而参数估计算法的计算需要占用一定的时间,导致系统响应变慢,导致对于控制周期要求较小的系统,该控制算法使用受限,且从实验结果来看系统在非理想条件下震荡较大;
    步骤B、基于以上存在响应慢和存在震荡的问题考虑如下改进方案;
    Δu(k)=Δu m(k)′+Δu p(k)    (7)
    其中,Δu m(k)′是MFAC控制器输出,Δu p(k)是比例控制器输出,用以下公式表示:
    Figure PCTCN2020114129-appb-100007
    Δu p(k)=βK(y *(k+1)-y(k))-βK(y *(k)-y(k-1))    (9)
    并提出以下抗饱和算法作为所提出的控制算法的一部分:当致动器处于饱和上限且仍有增长趋势时,或者当致动器处于饱和下限且仍在下降时,将停止更新积分器;否则,积分器正常工作;也就是说,在饱和的情况下,只执行有助于降低饱和程度的积分运算;它由以下公式表示:
    Δu m(k)′=Δu m(k)f(k)    (10)
    Figure PCTCN2020114129-appb-100008
    其中,u_max和u_min是致动器的上限和下限;
    基于公式(6)、(7)、(8)、(9),提出以下控制方案:
    Figure PCTCN2020114129-appb-100009
    如果
    Figure PCTCN2020114129-appb-100010
    Figure PCTCN2020114129-appb-100011
    Figure PCTCN2020114129-appb-100012
    Figure PCTCN2020114129-appb-100013
    如果
    Figure PCTCN2020114129-appb-100014
    Figure PCTCN2020114129-appb-100015
    Figure PCTCN2020114129-appb-100016
    其中,
    Figure PCTCN2020114129-appb-100017
    Figure PCTCN2020114129-appb-100018
    的初值;
    步骤C、对于以上所述的改进控制算法,通过理论推导,分析跟踪误差的收敛性和有界输入有界输出的稳定性;
    首先,定义以下系统输出误差:
    e(k)=y *(k)-y(k)    (15)
    将公式(2)和公式(14)替换为公式(15),当f(k)=1时,得:
    Figure PCTCN2020114129-appb-100019
    Figure PCTCN2020114129-appb-100020
    其中z是矩阵
    Figure PCTCN2020114129-appb-100021
    的特征值,D j,j=1,2,...,m是Gerschgorin圆盘;
    公式(17)等价于公式(18);
    Figure PCTCN2020114129-appb-100022
    由重置算法(12)和(13)得,
    Figure PCTCN2020114129-appb-100023
    Figure PCTCN2020114129-appb-100024
    由假设1得,|φ ij|≤b 1,b 2≤|φ ii|(k)|≤αb 2,i=1,…,m;j=1,...,m;i≠j;
    由以上条件得下式
    Figure PCTCN2020114129-appb-100025
    Figure PCTCN2020114129-appb-100026
    Figure PCTCN2020114129-appb-100027
    由重置算法公式(11)和假设1得,
    Figure PCTCN2020114129-appb-100028
    因此存在λ min,使得当λ>λ min时,下式成立;
    Figure PCTCN2020114129-appb-100029
    因此选择0<ρ≤1和λ>λ min,使得
    Figure PCTCN2020114129-appb-100030
    对于任意的λ>λ min,显然下式成立
    Figure PCTCN2020114129-appb-100031
    由式(21)、(23)、(24),知
    Figure PCTCN2020114129-appb-100032
    由式(18)和式(24)得,
    Figure PCTCN2020114129-appb-100033
    其中s(M)是M的谱半径;
    Figure PCTCN2020114129-appb-100034
    B=||βΦ c(k)K)|| v,由矩阵谱半径的结论得,存在一个任意小的正数ε 1,使得
    Figure PCTCN2020114129-appb-100035
    其中,||M|| v是矩阵M的相容范数;
    存在β使得B满足如下不等式:
    1>1-A≥M 11>B>0    (28)
    由式(16)和式(28)得:
    ||e(k+1)|| v≤A||e(k)|| v+B||e(k-1)|| v<(1-B)||e(k)|| v+B||e(k-1)|| v    (29)移项后得:
    ||e(k+1)|| v-||e(k)|| v<-B(||e(k)|| v-||e(k-1)|| v)    (30)
    基于式(30),从以下四个方面讨论e(k)的形式:
    第一种情况,当||e(k+1)|| v>||e(k)|| v且||e(k)|| v>||e(k-1)|| v时,得
    ||e(k+1)|| v-||e(k)|| v>-B(||e(k)|| v-||e(k-1)|| v)    (31)
    不等式结果与式(30)相反,因此,此假设不存在;
    第二种情况,当||e(k+1)|| v>||e(k)|| v且||e(k)|| v<||e(k-1)|| v时,由式(30)得
    Figure PCTCN2020114129-appb-100036
    即在三个相邻采样时刻,e(k)的减小量大于增加量,所以总体趋势呈下降趋势;
    第三种情况,当||e(k+1)|| v<||e(k)|| v且||e(k)|| v<||e(k-1)|| v时,得
    Figure PCTCN2020114129-appb-100037
    满足式(30),此条件下,e(k)呈下降趋势;
    第四种情况,当||e(k+1)|| v<||e(k)|| v且||e(k)|| v>||e(k-1)|| v时,由式(30)得,该情况有可能存在;在k+2时刻存在两种可能性:如果存在||e(k+2)|| v>||e(k+1)|| v,得到与第二种情况相同的结论;如果||e(k+2)|| v<||e(k+1)|| v,得到与第三种情况相同的结论;简而言之,在这种情况下,e(k)仍然呈下降趋势;
    当f(k)=0时,上述证明方法依然适用;综上所述,误差e(k)总体呈下降趋势;因此,误差的收敛性得证;
    步骤D、将以上控制算法应用于航空发动机模型的控制,选取三种不同的情况进行结果对比,用以验证控制算法的有效性和优越性;首先,比较标准条件下的MFAC+Kp、CFDL-MFAC和PID的控制效果,说明了所述改进控制器的有效性;然后,选择不同的高度和不同的延迟来比较控制效果,以说明控制器的优越性。
PCT/CN2020/114129 2020-08-25 2020-09-09 无模型自适应控制的改进方法 WO2022041317A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US17/440,052 US20220276620A1 (en) 2020-08-25 2020-09-09 Improved model-free adaptive control method

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010863628.XA CN112034707B (zh) 2020-08-25 2020-08-25 无模型自适应控制的改进方法
CN202010863628.X 2020-08-25

Publications (1)

Publication Number Publication Date
WO2022041317A1 true WO2022041317A1 (zh) 2022-03-03

Family

ID=73581261

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/114129 WO2022041317A1 (zh) 2020-08-25 2020-09-09 无模型自适应控制的改进方法

Country Status (3)

Country Link
US (1) US20220276620A1 (zh)
CN (1) CN112034707B (zh)
WO (1) WO2022041317A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118092293A (zh) * 2024-04-28 2024-05-28 西北工业大学 基于激励响应映射的月面轮式机器人数据驱动控制方法

Families Citing this family (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113110519B (zh) * 2021-05-26 2021-11-30 哈尔滨工程大学 一种舰船用非增量型无模型自适应艏向控制方法
CN115185178B (zh) * 2022-06-23 2023-06-30 江南大学 一种基于虚拟传感器的离散非线性系统的mfac容错控制方法
CN114815588B (zh) * 2022-06-28 2022-09-27 浙江大学 Pics补偿器与pid控制器设计方法及装置
CN116243610B (zh) * 2023-05-12 2023-08-01 青岛大学 一种数据驱动车辆队列容错跟踪控制追踪方法及系统

Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070181701A1 (en) * 2005-10-31 2007-08-09 Cheng George S Arrangement and method for automatically determined time constant for a control device
CN105487385A (zh) * 2016-02-01 2016-04-13 金陵科技学院 基于无模型自适应内模控制方法
CN109144066A (zh) * 2018-09-05 2019-01-04 哈尔滨工程大学 一种舰船用积分分离式pi型紧格式无模型自适应航向控制算法

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8594813B2 (en) * 2009-08-14 2013-11-26 General Cybernation Group, Inc. Dream controller
US20140309793A1 (en) * 2013-04-15 2014-10-16 General Cybernation Group, Inc. Method and apparatus of self-organizing actuation and control
CN106054594B (zh) * 2016-06-12 2018-12-11 金陵科技学院 基于控制输入饱和的无模型自适应控制方法
CN109507885B (zh) * 2018-12-20 2020-09-25 中国海洋大学 基于自抗扰的无模型自适应auv控制方法
CN110737197B (zh) * 2019-09-25 2022-09-06 江苏航运职业技术学院 一种基于无模型的柴油机转速自适应自抗扰控制方法

Patent Citations (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20070181701A1 (en) * 2005-10-31 2007-08-09 Cheng George S Arrangement and method for automatically determined time constant for a control device
CN105487385A (zh) * 2016-02-01 2016-04-13 金陵科技学院 基于无模型自适应内模控制方法
CN109144066A (zh) * 2018-09-05 2019-01-04 哈尔滨工程大学 一种舰船用积分分离式pi型紧格式无模型自适应航向控制算法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
LIU XIAO-YU; SUN XI-MING; ZHANG YONG-LIANG; WANG XUE-FANG; WEN SI-XIN: "An Improved Model-Free Adaptive Control Algorithm and the Application in Aero-engines", 2020 39TH CHINESE CONTROL CONFERENCE (CCC), TECHNICAL COMMITTEE ON CONTROL THEORY, CHINESE ASSOCIATION OF AUTOMATION, 27 July 2020 (2020-07-27), pages 1940 - 1945, XP033821223, DOI: 10.23919/CCC50068.2020.9189377 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118092293A (zh) * 2024-04-28 2024-05-28 西北工业大学 基于激励响应映射的月面轮式机器人数据驱动控制方法

Also Published As

Publication number Publication date
CN112034707B (zh) 2021-08-20
US20220276620A1 (en) 2022-09-01
CN112034707A (zh) 2020-12-04

Similar Documents

Publication Publication Date Title
WO2022041317A1 (zh) 无模型自适应控制的改进方法
Wang et al. Fixed-time disturbance observer based fixed-time back-stepping control for an air-breathing hypersonic vehicle
Sun et al. Finite-time tracking control of hypersonic vehicle with input saturation
Ohri GA tuned LQR and PID controller for aircraft pitch control
CN102621889A (zh) 一种压电陶瓷定位的复合控制方法
CN110161841A (zh) 一种适用于暂冲式跨声速风洞的前馈-模糊pid控制方法
Singh et al. State feedback control of an aeroelastic system with structural nonlinearity
CN110376902B (zh) 一种欠驱动机械系统伺服约束跟踪控制器的设计方法
Banerjee et al. Stabilization of double link inverted pendulum using LQR
Zhao et al. A novel tensor product model transformation-based adaptive variable universe of discourse controller
CN111007724A (zh) 一种基于区间ii型模糊神经网络的高超音速飞行器指定性能量化跟踪控制方法
CN102410138B (zh) 一种获取风力发电机组最优控制输入的方法
Sun et al. Event-triggered intelligent critic control with input constraints applied to a nonlinear aeroelastic system
Suzuki et al. Simultaneous structure/control design optimization of a wing structure with a gust load alleviation system
An et al. Event-triggered adaptive control of hypersonic vehicles subject to actuator nonlinearities
Zheng et al. Adaptive fractional order non-singular terminal sliding mode anti-disturbance control for advanced layout carrier-based UAV
CN104155876B (zh) 一种pid控制器的分离实现方法
Dillsaver et al. Constrained control of very flexible aircraft using reference and extended command governors
Wahid et al. Comparative assesment using LQR and fuzzy logic controller for a pitch control system
Singh et al. Design and modeling of controllers for aircraft pitch contol movement
Simões et al. Nonsmooth multi-objective synthesis with applications
CN109244510A (zh) 一种基于未建模动态补偿的固体氧化物燃料电池控制方法
Yao et al. Revisiting the separation principle for improved transition control
CN114839882A (zh) 一种输入约束下的非线性系统复合自适应控制方法
Midhat Optimal LQR controller design for wing rock motion control in delta wing aircraft

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20950986

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20950986

Country of ref document: EP

Kind code of ref document: A1