CN113759700A - Fractional order PID self-adaptive adjustment method based on particle swarm and neural network - Google Patents

Fractional order PID self-adaptive adjustment method based on particle swarm and neural network Download PDF

Info

Publication number
CN113759700A
CN113759700A CN202111001128.6A CN202111001128A CN113759700A CN 113759700 A CN113759700 A CN 113759700A CN 202111001128 A CN202111001128 A CN 202111001128A CN 113759700 A CN113759700 A CN 113759700A
Authority
CN
China
Prior art keywords
fractional order
pid controller
controlled object
order pid
time
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
CN202111001128.6A
Other languages
Chinese (zh)
Inventor
黄壮壮
顾正晖
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.)
South China University of Technology SCUT
Original Assignee
South China University of Technology SCUT
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 South China University of Technology SCUT filed Critical South China University of Technology SCUT
Priority to CN202111001128.6A priority Critical patent/CN113759700A/en
Publication of CN113759700A publication Critical patent/CN113759700A/en
Pending legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G05—CONTROLLING; REGULATING
    • G05B—CONTROL OR REGULATING SYSTEMS IN GENERAL; FUNCTIONAL ELEMENTS OF SUCH SYSTEMS; MONITORING OR TESTING ARRANGEMENTS FOR SUCH SYSTEMS OR ELEMENTS
    • G05B11/00—Automatic controllers
    • G05B11/01—Automatic controllers electric
    • G05B11/36—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential
    • G05B11/42—Automatic controllers electric with provision for obtaining particular characteristics, e.g. proportional, integral, differential for obtaining a characteristic which is both proportional and time-dependent, e.g. P. I., P. I. D.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • Automation & Control Theory (AREA)
  • Feedback Control In General (AREA)

Abstract

The invention discloses a fractional order PID self-adaptive adjustment method based on particle swarm and neural network, which uses a fractal derivative defined by Hausdorff to carry out discretization processing on a fractional order PID controller. Then, a gradient descent method with momentum is used to derive the fraction order PID controller parameters and the order update formula. On the basis, in order to make the performance of the controller better, initial values need to be given to the parameters and orders of the fractional order PID controller before adaptive adjustment, so that unnecessary adjustment of the controlled object in the early stage of operation is avoided. The particle swarm optimization is used for carrying out parameter setting on the fractional order PID controller in advance to obtain an optimal value, and on the basis, the parameters of the fractional order PID controller are updated by using a self-adaptive adjustment strategy, so that a better control effect is achieved, the realization is easy, and the dynamic performance is good.

Description

Fractional order PID self-adaptive adjustment method based on particle swarm and neural network
Technical Field
The invention relates to the technical field of application of fractional order calculus in control engineering, in particular to a fractional order PID self-adaptive adjusting method based on particle swarm and neural network.
Background
In recent years, fractional calculus has been gaining increasing use in the scientific and engineering fields, which has led to interest in fractional calculus for researchers in the modeling and control fields. In the field of control theory and application, research and development on the fractional order controller are developed in order to obtain better control performance than the traditional integer order controller.
Fractional order PID controllers have the potential to provide better control performance than traditional PID controllers, but their controller parameter design becomes more complex due to the two degrees of freedom. In general, the design method of the fractional order PID controller can be mainly classified into analytic methods, such as a dominant pole method, a horizontal phase method at a cutoff frequency, and an optimization method. The method can obtain an effective parameter setting result to a certain extent, but the parameter setting method is an off-line calculation method and cannot perform on-line adjustment on the parameters of the fractional order PID controller.
Disclosure of Invention
The invention aims to overcome the difficulty in online parameter setting of the conventional fractional order PID controller, and provides a fractional order PID self-adaptive adjusting method based on a particle swarm and a neural network.
In order to achieve the purpose, the technical scheme provided by the invention is as follows: a fractional order PID self-adaptive adjusting method based on particle swarm and neural network comprises the following steps:
1) in the discrete calculation of the fractional order PID controller, a fractal derivative, namely a differential and integral discrete form of a Hausdorff derivative is used to replace the traditional fractional order calculus to simplify the complex calculation of the fractional order PID controller, so that the order update of the fractional order PID controller is simpler;
2) the method comprises the steps of modeling a controlled object by using a BP neural network and deducing a fractional order PID controller parameter and an update formula of an order by using a gradient descent method with momentum, so that the problem of self-adaptive update of the fractional order PID controller order is solved;
3) the parameters and the orders of the fractional order PID controller are endowed with initial values by using a particle swarm algorithm, unnecessary adjustment of a controlled object in the early stage of operation is avoided, and the parameters and the orders of the fractional order PID controller are adaptively updated by using a derived updating formula.
Further, in step 1), the Hausdorff derivative is used, which is defined as follows when the time argument t is greater than 0:
Figure BDA0003235671360000021
in the above formula, f (t)1) And f (t) are each variable t1Function value of and t, t1-αRepresents the variable t to the power of 1-alpha, tαAnd
Figure BDA0003235671360000022
respectively represent the variables t and t1To the power of alpha of (a),
Figure BDA0003235671360000023
denotes the alpha order differential, t, of Hausdorff1→ t denotes t1Approaching t;
the integral form of Hausdorff is defined as follows:
Figure BDA0003235671360000024
in the above formula, the first and second carbon atoms are,
Figure BDA0003235671360000025
representing the alpha-order integral of Hausdorff, f (τ) being the function value of the variable τ, τα-1Expressing the power alpha-1 of the variable tau, and carrying out discretization treatment, wherein the discretization transformation rule is as follows:
Figure BDA0003235671360000026
in the above equation, k is a sampling sequence, T is a sampling period, i is a variable, and i ═ 0,1,2, …, k, f (iT), f (kT), and f ((k-1) T) are function values at the time of iT, kT, and (k-1) T, respectively;
according to the discretized transformation rule above, for k > 0, the discrete form of the Hausdorff derivative is:
Figure BDA0003235671360000031
Figure BDA0003235671360000032
in the above formula, the first and second carbon atoms are,
Figure BDA0003235671360000033
and
Figure BDA0003235671360000034
respectively representing the alpha-order differential and integral, k, of discrete forms of the Hausdorff derivative1-αRepresents the sampling sequence k to the power of 1-alpha, and f (k) is a function value of k time;
and a fractional calculus controller corresponding to a time domain:
Figure BDA0003235671360000035
the fractional order PID controller converted into discrete time specifically comprises the following steps:
Figure BDA0003235671360000036
in the above formula, u (t) and e (t) are the control quantity and deviation of the controlled object at the time t,
Figure BDA0003235671360000037
and
Figure BDA0003235671360000038
lambda order integral and mu order differential respectively representing Hausdorff derivative, lambda and mu are integral and differential orders respectively, u (k) is controlled object control quantity, e (k) is controlled object deviation, k is controlled object deviationp、kiAnd kdRespectively, are parameters of the fractional order PID controller.
Further, in step 2), the controlled object can be expressed as:
y(k)＝g(y(k-1),y(k-2),…,y(k-ky),u(k),u(k-1),…,u(k-ku))
wherein the content of the first and second substances,kyand kuDetermined by the characteristics of a specific controlled object, g is a function for expressing the controlled object, y (k) and u (k) are respectively an output value and a control quantity of the controlled object at the moment k, and y (k-k)y) And u (k-k)u) Are respectively k-kyTime sum k-kuThe output value and the control quantity at the moment;
the BP neural network can be used for fitting, so that the purpose of modeling the controlled object is achieved;
for the whole control controlled object, the following loss function e (k) is defined:
E(k)＝0.5*e(k)＝0.5*(r(k)-y(k))
where k is a sampling sequence, k is 0,1,2,3, …, r (k) is a controlled object expected value, and e (k) is a controlled object deviation, the parameter regulation rule of the fractional order PID controller is as follows:
Figure BDA0003235671360000041
Figure BDA0003235671360000042
Figure BDA0003235671360000043
Figure BDA0003235671360000044
Figure BDA0003235671360000045
wherein eta isp、ηi、ηd、ηλAnd ημY (k) and u (k) represent the controlled object output and the controlled variable at the time k,
Figure BDA0003235671360000046
the value of (A) is not easy to be obtained, use
Figure BDA0003235671360000047
Instead of this, the user can,
Figure BDA0003235671360000048
for the BP neural network fitting output of the controlled object, lambda (k) and mu (k) are respectively integral and differential orders of k time, kp(k)、ki(k) And kd(k) Respectively, a parameter of a fractional order PID controller at time k, Δ kp(k+1)、Δki(k+1)、Δkd(k +1), Δ λ (k +1), and Δ μ (k +1) each represent a change amount at the next time;
defining the vector param (k) as [ kp(k),ki(k),kd(k),λ(k),μ(k)]At the value of time k, the update formula of the fractional order PID controller parameter and order is:
param(k+1)＝param(k)+Δparam(k+1)+βΔparam(k)
where β is the momentum coefficient, param (k +1) is the value at time k +1, and Δ param (k +1) and Δ param (k) are the amounts of change at times k +1 and k, respectively.
Further, in the step 3), a fitness function is designed to be in accordance with the control performance required by the controlled object, then, a particle swarm optimization is used, parameters and orders of the optimal fractional order PID controller are found through MATLAB simulation, and then, adaptive updating of the parameters and orders of the fractional order PID controller through an updating formula is achieved through writing a program.
Compared with the prior art, the invention has the following advantages and beneficial effects:
1. according to the invention, the Hausdorff derivative is used, so that the difficulty of fractional calculus in the fractional PID controller is reduced, and the calculation cost is saved.
2. The invention avoids the gamma function processing in the parameter updating process of the controller, so that the order updating of the fractional order PID controller is simpler
3. The invention can realize the online adjustment of the parameters and orders of the fractional order PID controller.
4. The invention introduces a particle swarm algorithm, and gives initial values to the parameters and the orders of the fractional order PID controller, thereby avoiding unnecessary adjustment of the controlled object in the early stage of operation.
5. The method provided by the invention can obtain better control effect and dynamic performance.
Drawings
FIG. 1 is a diagram of an embodiment of a vertical take-off and landing system.
Fig. 2 is a block diagram of fractional order PID adaptive adjustment in the embodiment.
FIG. 3 is a graph of the comparative follow-up response of the method and neural network adaptive PID controller of the present invention.
FIG. 4 is a graph of the comparative follow-up response error of the method and neural network adaptive PID controller of the present invention.
Detailed Description
The present invention will be described in further detail with reference to examples and drawings, but the present invention is not limited thereto.
The embodiment discloses a fractional order PID adaptive adjustment method based on a particle swarm and a neural network, which has the following specific conditions:
1) in the discrete calculation of the fractional order PID controller, a fractal derivative, namely a differential and integral discrete form of a Hausdorff derivative is used to replace the traditional fractional order calculus to simplify the complex calculation, so that the order update of the fractional order PID controller is more concise.
The derivative of Hausdorff used, when the time argument t is greater than 0, is defined as follows:
Figure BDA0003235671360000061
in the above formula, f (t)1) And f (t) are each variable t1Function value of and t, t1-αRepresents the variable t to the power of 1-alpha, tαAnd
Figure BDA0003235671360000062
respectively represent the variables t and t1To the power of alpha of (a),
Figure BDA0003235671360000063
denotes the alpha order differential, t, of Hausdorff1→ t denotes t1Approaching t;
the integral form of Hausdorff is defined as follows:
Figure BDA0003235671360000064
in the above formula, the first and second carbon atoms are,
Figure BDA0003235671360000065
representing the alpha-order integral of Hausdorff, f (τ) being the function value of the variable τ, τα-1Expressing the power alpha-1 of the variable tau, and carrying out discretization treatment, wherein the discretization transformation rule is as follows:
Figure BDA0003235671360000066
in the above equation, k is a sampling sequence, T is a sampling period, i is a variable, and i ═ 0,1,2, …, k, f (iT), f (kT), and f ((k-1) T) are function values at the time of iT, kT, and (k-1) T, respectively;
according to the discretized transformation rule above, for k > 0, the discrete form of the Hausdorff derivative is:
Figure BDA0003235671360000067
Figure BDA0003235671360000068
in the above formula, the first and second carbon atoms are,
Figure BDA0003235671360000069
and
Figure BDA00032356713600000610
each represents HDiscrete form alpha order differential and integral, k, of the ausdorff derivative1-αRepresents the sampling sequence k to the power of 1-alpha, and f (k) is a function value of k time;
and a fractional calculus controller corresponding to a time domain:
Figure BDA0003235671360000071
the fractional order PID controller converted into discrete time specifically comprises the following steps:
Figure BDA0003235671360000072
in the above formula, u (t) and e (t) are the control quantity and deviation of the controlled object at the time t,
Figure BDA0003235671360000073
and
Figure BDA0003235671360000074
lambda order integral and mu order differential respectively representing Hausdorff derivative, lambda and mu are integral and differential orders respectively, u (k) is controlled object control quantity, e (k) is controlled object deviation, k is controlled object deviationp、kiAnd kdRespectively, are parameters of the fractional order PID controller.
2) A controlled object is modeled by using a BP neural network and an updating formula of a fractional order PID controller parameter and an order is deduced by using a gradient descent method with momentum, so that the problem of self-adaptive updating of the fractional order PID controller order is solved.
In reality, the controlled object can be expressed as:
y(k)＝g(y(k-1),y(k-2),…,y(k-ky),u(k),u(k-1),…,u(k-ku))
wherein k isyAnd kuDetermined by the characteristics of a specific controlled object, g is a function for expressing the controlled object, y (k) and u (k) are respectively an output value and a control quantity of the controlled object at the moment k, and y (k-k)y) And u (k-k)u) Are respectively k-kyTime sum k-kuThe output value and the control quantity at the moment;
the BP neural network can be used for fitting, so that the purpose of modeling the controlled object is achieved;
for the whole control controlled object, the following loss function e (k) is defined:
E(k)＝0.5*e(k)＝0.5*(r(k)-y(k))
where k is a sampling sequence, k is 0,1,2,3, …, r (k) is a controlled object expected value, and e (k) is a controlled object deviation, the parameter regulation rule of the fractional order PID controller is as follows:
Figure BDA0003235671360000075
Figure BDA0003235671360000081
Figure BDA0003235671360000082
Figure BDA0003235671360000083
Figure BDA0003235671360000084
wherein eta isp、ηi、ηd、ηλAnd ημY (k) and u (k) represent the controlled object output and the controlled variable at the time k,
Figure BDA0003235671360000085
the value of (A) is not easy to be obtained, use
Figure BDA0003235671360000086
Instead of this, the user can,
Figure BDA0003235671360000087
being controlled objectsBP neural network fitting output, λ (k) and μ (k) are integral and differential orders at time k, respectivelyp(k)、ki(k) And kd(k) Respectively, a parameter of a fractional order PID controller at time k, Δ kp(k+1)、Δki(k+1)、Δkd(k +1), Δ λ (k +1), and Δ μ (k +1) each represent a change amount at the next time;
defining the vector param (k) as [ kp(k),ki(k),kd(k),λ(k),μ(k)]At the value of time k, the update formula of the fractional order PID controller parameter and order is:
param(k+1)＝param(k)+Δparam(k+1)+βΔparam(k)
where β is the momentum coefficient, param (k +1) is the value at time k +1, and Δ param (k +1) and Δ param (k) are the amounts of change at times k +1 and k, respectively.
3) The parameters and the orders of the fractional order PID controller are endowed with initial values by using a particle swarm algorithm, unnecessary adjustment of a controlled object in the early stage of operation is avoided, and the parameters and the orders of the fractional order PID controller are adaptively updated by using a derived updating formula, which specifically comprises the following steps:
the method comprises the steps of designing a fitness function to enable the fitness function to be in line with the control performance required by a controlled object, then using a particle swarm algorithm to find out the optimal parameters and orders of the fractional order PID controller through MATLAB simulation, and then writing a program to realize self-adaptive updating of the parameters and orders of the fractional order PID controller through an updating formula.
In the following, we select a vertical take-off and landing system as a control object, the vertical take-off and landing system is composed of a body rod, one side of the body rod is connected with a variable speed fan called a propeller actuator, and the other side of the body rod is connected with a counterweight, and a physical diagram of the vertical take-off and landing system is shown in fig. 1. Looking up documents and data, the transfer function of the vertical take-off and landing system can be solved as follows:
Figure BDA0003235671360000091
wherein s is a complex variable.
As shown in FIG. 2, BP neural net is usedAnd the network identifies the controlled object represented by the transfer function of the vertical take-off and landing system to obtain a neural network model. The neural network model selects a three-layer architecture, wherein an excitation function of a hidden layer is selected as a tanh function, and an input vector is selected as: x (k) ═ y (k-1), y (k-2), u (k-1), u (k-2)]TAnd y, (k) and u (k) are output and control quantity of the controlled object at the moment k, and the number of nodes of the hidden layer is selected to be 8.
For the fractional order PID controller parameters and the order update formula derived by using the gradient descent method with momentum, the learning rates of the parameters are respectively set as follows: etap＝0.2，ηi＝0.05，ηdThe learning rates of the order are set to 0.05: etaλ＝0.01，ημ＝0.01。
The parameters of the fractional order PID controller are adjusted through a particle swarm algorithm to obtain: k is a radical ofpIs 17.46, kiIs 22.48, kd24.89, λ 1.13, μ 0.87, i.e. the fractional order controller is:
Figure BDA0003235671360000092
and taking the value obtained by the setting as the initial value of the fractional order PID controller parameter and the order. And during MATLAB simulation operation, testing the vertical rising and falling controlled object model by using the set fractional order PID controller, and updating the parameters of the fractional order PID controller by using a designed gradient falling method with momentum.
To verify the effectiveness of the method of the present invention, a neural network adaptive PID controller (NNPID) was used in comparison with the method of the present invention (NNFOPID). In the vertical take-off and landing controlled object model, a sinusoidal signal of 1Hz is used as the desired pitch angle, and the resulting response is shown in fig. 3 and 4.
As can be seen from the output responses of the different controllers depicted in fig. 3, the method provided by the present invention can obtain a more accurate response and a better following effect than the output of the neural network adaptive PID controller. From fig. 4, it can be seen that the response error of the proposed method is within 0.05 and has a gradually decreasing trend, while the response error of the neural network adaptive PID controller is within 0.09 and remains in this range. The method provided by the invention can further improve the precision of the dynamic response of the controlled object and obtain better control performance.
The above embodiments are preferred embodiments of the present invention, but the present invention is not limited to the above embodiments, and any other changes, modifications, substitutions, combinations, and simplifications which do not depart from the spirit and principle of the present invention should be construed as equivalents thereof, and all such changes, modifications, substitutions, combinations, and simplifications are intended to be included in the scope of the present invention.

Claims (4)

1. The fractional order PID self-adaptive adjusting method based on the particle swarm and the neural network is characterized by comprising the following steps of:
1) in the discrete calculation of the fractional order PID controller, a fractal derivative, namely a differential and integral discrete form of a Hausdorff derivative is used to replace the traditional fractional order calculus to simplify the complex calculation of the fractional order PID controller, so that the order update of the fractional order PID controller is simpler;
2) the method comprises the steps of modeling a controlled object by using a BP neural network and deducing a fractional order PID controller parameter and an update formula of an order by using a gradient descent method with momentum, so that the problem of self-adaptive update of the fractional order PID controller order is solved;
3) the parameters and the orders of the fractional order PID controller are endowed with initial values by using a particle swarm algorithm, unnecessary adjustment of a controlled object in the early stage of operation is avoided, and the parameters and the orders of the fractional order PID controller are adaptively updated by using a derived updating formula.
2. The particle swarm and neural network based fractional order PID adaptive adjustment method of claim 1, characterized in that: in step 1), the Hausdorff derivative is used, which, when the time argument t is greater than 0, is defined as follows:
Figure FDA0003235671350000011
in the above formula, f (t)1) And f (t) are each variable t1Function value of and t, t1-αRepresents the variable t to the power of 1-alpha, tαAnd
Figure FDA0003235671350000012
respectively represent the variables t and t1To the power of alpha of (a),
Figure FDA0003235671350000013
denotes the alpha order differential, t, of Hausdorff1→ t denotes t1Approaching t;
the integral form of Hausdorff is defined as follows:
Figure FDA0003235671350000014
in the above formula, the first and second carbon atoms are,
Figure FDA0003235671350000015
representing the alpha-order integral of Hausdorff, f (τ) being the function value of the variable τ, τα-1Expressing the power alpha-1 of the variable tau, and carrying out discretization treatment, wherein the discretization transformation rule is as follows:
Figure FDA0003235671350000021
in the above equation, k is a sampling sequence, T is a sampling period, i is a variable, and i ═ 0,1,2, …, k, f (iT), f (kT), and f ((k-1) T) are function values at the time of iT, kT, and (k-1) T, respectively;
according to the discretized transformation rule above, for k > 0, the discrete form of the Hausdorff derivative is:
Figure FDA0003235671350000022
Figure FDA0003235671350000023
in the above formula, the first and second carbon atoms are,
Figure FDA0003235671350000024
and
Figure FDA0003235671350000025
respectively representing the alpha-order differential and integral, k, of discrete forms of the Hausdorff derivative1-αRepresents the sampling sequence k to the power of 1-alpha, and f (k) is a function value of k time;
and a fractional calculus controller corresponding to a time domain:
Figure FDA0003235671350000026
the fractional order PID controller converted into discrete time specifically comprises the following steps:
Figure FDA0003235671350000027
in the above formula, u (t) and e (t) are the control quantity and deviation of the controlled object at the time t,
Figure FDA0003235671350000028
and
Figure FDA0003235671350000029
lambda order integral and mu order differential respectively representing Hausdorff derivative, lambda and mu are integral and differential orders respectively, u (k) is controlled object control quantity, e (k) is controlled object deviation, k is controlled object deviationp、kiAnd kdRespectively, are parameters of the fractional order PID controller.
3. The particle swarm and neural network based fractional order PID adaptive adjustment method of claim 1, characterized in that: in step 2), the controlled object can be expressed as:
y(k)＝g(y(k-1),y(k-2),…,y(k-ky),u(k),u(k-1),…,u(k-ku))
wherein k isyAnd kuDetermined by the characteristics of a specific controlled object, g is a function for expressing the controlled object, y (k) and u (k) are respectively an output value and a control quantity of the controlled object at the moment k, and y (k-k)y) And u (k-k)u) Are respectively k-kyTime sum k-kuThe output value and the control quantity at the moment;
the BP neural network can be used for fitting, so that the purpose of modeling the controlled object is achieved;
for the whole control controlled object, the following loss function e (k) is defined:
E(k)＝0.5*e(k)＝0.5*(r(k)-y(k))
where k is a sampling sequence, k is 0,1,2,3, …, r (k) is a controlled object expected value, and e (k) is a controlled object deviation, the parameter regulation rule of the fractional order PID controller is as follows:
Figure FDA0003235671350000031
Figure FDA0003235671350000032
Figure FDA0003235671350000033
Figure FDA0003235671350000034
Figure FDA0003235671350000035
wherein eta isp、ηi、ηd、ηλAnd ημY (k) and u (k) represent the controlled object output and the controlled variable at the time k,
Figure FDA0003235671350000036
the value of (A) is not easy to be obtained, use
Figure FDA0003235671350000037
Instead of this, the user can,
Figure FDA0003235671350000038
for the BP neural network fitting output of the controlled object, lambda (k) and mu (k) are respectively integral and differential orders of k time, kp(k)、ki(k) And kd(k) Respectively, a parameter of a fractional order PID controller at time k, Δ kp(k+1)、Δki(k+1)、Δkd(k +1), Δ λ (k +1), and Δ μ (k +1) each represent a change amount at the next time;
defining the vector param (k) as [ kp(k),ki(k),kd(k),λ(k),μ(k)]At the value of time k, the update formula of the fractional order PID controller parameter and order is:
param(k+1)＝param(k)+Δparam(k+1)+βΔparam(k)
where β is the momentum coefficient, param (k +1) is the value at time k +1, and Δ param (k +1) and Δ param (k) are the amounts of change at times k +1 and k, respectively.
4. The particle swarm and neural network based fractional order PID adaptive adjustment method of claim 1, characterized in that: in the step 3), a fitness function is designed to be in accordance with the control performance required by the controlled object, then, a particle swarm optimization is used, the optimal parameters and orders of the fractional order PID controller are found through MATLAB simulation, and then, the parameters and orders of the fractional order PID controller are adaptively updated through an updating formula through a programming program.
CN202111001128.6A 2021-08-30 2021-08-30 Fractional order PID self-adaptive adjustment method based on particle swarm and neural network Pending CN113759700A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202111001128.6A CN113759700A (en) 2021-08-30 2021-08-30 Fractional order PID self-adaptive adjustment method based on particle swarm and neural network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202111001128.6A CN113759700A (en) 2021-08-30 2021-08-30 Fractional order PID self-adaptive adjustment method based on particle swarm and neural network

Publications (1)

Publication Number Publication Date
CN113759700A true CN113759700A (en) 2021-12-07

Family

ID=78791787

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202111001128.6A Pending CN113759700A (en) 2021-08-30 2021-08-30 Fractional order PID self-adaptive adjustment method based on particle swarm and neural network

Country Status (1)

Country Link
CN (1) CN113759700A (en)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114019985A (en) * 2021-12-20 2022-02-08 中国海洋大学 Unmanned ship rudder direction control design method based on fractional order PID and particle swarm algorithm
CN116184812A (en) * 2023-04-24 2023-05-30 荣耀终端有限公司 Signal compensation method, electronic equipment and medium

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103105774A (en) * 2013-01-30 2013-05-15 上海交通大学 Fractional order proportion integration differentiation (PID) controller setting method based on improved quantum evolutionary algorithm
CN106094910A (en) * 2016-08-22 2016-11-09 湖南科技大学 A kind of parameter tuning method separated based on particle cluster algorithm PID
CN106951617A (en) * 2017-03-10 2017-07-14 河海大学 A kind of point shape derivative analogue method of chlorion unusual dispersion ability data reconstruction in concrete
CN110531612A (en) * 2019-08-16 2019-12-03 佛山科学技术学院 A kind of parameter tuning method of Fractional Order PID Controller
GB202019112D0 (en) * 2020-10-26 2021-01-20 Univ Guizhou Fractional-order MEMS gyroscope acceleration adaptive backstepping control method without accurate reference trajectory

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103105774A (en) * 2013-01-30 2013-05-15 上海交通大学 Fractional order proportion integration differentiation (PID) controller setting method based on improved quantum evolutionary algorithm
CN106094910A (en) * 2016-08-22 2016-11-09 湖南科技大学 A kind of parameter tuning method separated based on particle cluster algorithm PID
CN106951617A (en) * 2017-03-10 2017-07-14 河海大学 A kind of point shape derivative analogue method of chlorion unusual dispersion ability data reconstruction in concrete
CN110531612A (en) * 2019-08-16 2019-12-03 佛山科学技术学院 A kind of parameter tuning method of Fractional Order PID Controller
GB202019112D0 (en) * 2020-10-26 2021-01-20 Univ Guizhou Fractional-order MEMS gyroscope acceleration adaptive backstepping control method without accurate reference trajectory

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
ZHE GAO: "A Tuning Method via Borges Derivative of a Neural Network-Based Discrete-Time Fractional-Order PID Controller with Hausdorff Difference and Hausdorff Sum", 《FRACTAL AND FRACTIONAL》 *
杜慧东: "基于改进自适应动态规划的六相永磁同步电机矢量控制***研究", 《中国优秀硕士学位论文全文数据库工程科技Ⅱ辑》 *
胡海波: "基于混合 PSO 神经网络的自整定分数阶 PID 控制器", 《微电子学与计算机》 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114019985A (en) * 2021-12-20 2022-02-08 中国海洋大学 Unmanned ship rudder direction control design method based on fractional order PID and particle swarm algorithm
CN114019985B (en) * 2021-12-20 2023-12-22 中国海洋大学 Unmanned rudder direction control design method based on fractional order PID and particle swarm algorithm
CN116184812A (en) * 2023-04-24 2023-05-30 荣耀终端有限公司 Signal compensation method, electronic equipment and medium

Similar Documents

Publication Publication Date Title
CN113759700A (en) Fractional order PID self-adaptive adjustment method based on particle swarm and neural network
CN109522602A (en) A kind of Modelica Model Parameter Optimization method based on agent model
Couceiro et al. Application of fractional algorithms in the control of a robotic bird
CN103472723A (en) Predictive control method and system based on multi-model generalized predictive controller
JP2007500379A (en) Software optimization device for intelligent control system
CN108008627A (en) A kind of reinforcement learning adaptive PID control method of parallel optimization
CN108828934A (en) A kind of fuzzy PID control method and device based on Model Distinguish
Li et al. Training a robust reinforcement learning controller for the uncertain system based on policy gradient method
CN109254530A (en) MFA control method based on grinding process basis circuit
CN204595644U (en) Based on the aluminum-bar heating furnace temperature of combustion automaton of neural network
CN114839880A (en) Self-adaptive control method based on flexible joint mechanical arm
CN109143853B (en) Self-adaptive control method for liquid level of fractionating tower in petroleum refining process
CN112696319A (en) Wind turbine model-based control and estimation with accurate online models
CN107180279B (en) QPSO-DMPC-based reaction regeneration system optimization control method
Kumar et al. Lyapunov stability-based control and identification of nonlinear dynamical systems using adaptive dynamic programming
Mfoumboulou Design of a model reference adaptive PID control algorithm for a tank system
Abdulameer et al. GUI based control system analysis using PID controller for education
Mu et al. Machine learning-based active flutter suppression for a flexible flying-wing aircraft
CN108351622A (en) The method for generating the fuzzy knowledge base for may be programmed fuzzy controller
Abdulla et al. Roll control system design using auto tuning LQR technique
CN114839874A (en) Parallel control method and system for system model partial unknown
Tajjudin et al. Model reference input for an optimal PID tuning using PSO
Wang et al. Event-based online learning control design with eligibility trace for discrete-time unknown nonlinear systems
Thampi et al. Multiple model based flight control design
Rayme et al. Control System based on Reinforcement Learning applied to a Klatt-Engell Reactor

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
WD01 Invention patent application deemed withdrawn after publication
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20211207