CN108536016A - A kind of network control method based on fuzzy inverse model - Google Patents
A kind of network control method based on fuzzy inverse model Download PDFInfo
- Publication number
- CN108536016A CN108536016A CN201810352904.9A CN201810352904A CN108536016A CN 108536016 A CN108536016 A CN 108536016A CN 201810352904 A CN201810352904 A CN 201810352904A CN 108536016 A CN108536016 A CN 108536016A
- Authority
- CN
- China
- Prior art keywords
- fuzzy
- belongs
- model
- output
- control
- 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.)
- Granted
Links
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
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/0265—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion
- G05B13/0275—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric the criterion being a learning criterion using fuzzy logic only
-
- 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
- G05B13/00—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion
- G05B13/02—Adaptive control systems, i.e. systems automatically adjusting themselves to have a performance which is optimum according to some preassigned criterion electric
- G05B13/04—Adaptive 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
- G05B13/042—Adaptive 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
Landscapes
- Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Physics & Mathematics (AREA)
- Medical Informatics (AREA)
- Health & Medical Sciences (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Computation (AREA)
- General Physics & Mathematics (AREA)
- Automation & Control Theory (AREA)
- Mathematical Physics (AREA)
- Fuzzy Systems (AREA)
- Feedback Control In General (AREA)
Abstract
The invention discloses a kind of network control methods based on fuzzy inverse model, and the inverse system of single-point model is obscured by solution, use the mode of iteration that influence of following control action with compensation network time delay to control system performance can be obtained.Linear system is both can be applied to, nonlinear system can also be applied to without understanding the mechanism of controlled process in depth using the present invention.
Description
Technical field
The present invention relates to a kind of network control methods based on fuzzy inverse model, and the inverse of single-point model is obscured by solution
System uses the mode of iteration that influence of following control action with compensation network time delay to control system performance can be obtained, belongs to
Technical field of automatic control.
Background technology
The progress of network technology has pushed the development of control theory.In a network environment, control structure is no longer traditional meaning
Point-to-point mode in justice, but a kind of distributed frame.This structure lower sensor, controller and actuator can pass through network
Medium forms closed loop, network consisting feedback control system.Have the characteristics that simple in structure, at low cost, easy to maintain.But due to
Inevitably there is phenomena such as time delay, data packetloss in network in the characteristic of network share medium, this is just traditional control
Theory proposes new challenge.
Meanwhile in practical application, the mechanism of some controlled processes is unclear, it is difficult to directly establish mechanism mathematical model, and show
Some network control methods are the method based on model mostly, and there is an urgent need to develop some control methods based on data.This
Fuzzy clustering modeling technique is applied in invention, can establish the fuzzy clustering of system directly according to the inputoutput data of controlled device
Model, then system is changed into fuzzy single-point model, a series of control action in futures is obtained iteration and by way of inverting, from
And can process end select appropriate control sequences by way of compensation network time delay and data packetloss etc. to control system
The influence of energy.Method by being then based on data, thus, have a wide range of applications space.
Invention content
(1) technical problems to be solved
In view of the above-mentioned problems of the prior art, the main purpose of the present invention is to provide one kind being based on fuzzy inverse model
Network control method, fully to improve the control performance of network control system.
(2) technical solution
In order to achieve the above objectives, the technical proposal of the invention is realized in this way:
Network control method based on fuzzy inverse model, this method include:
A, outputting and inputting according to controlled process obtains the dynamic mathematical modulo of system by the modeling method of fuzzy clustering
Type;
B, the mathematical model of fuzzy clustering is converted to fuzzy single-point model of equal value;
C, iteration is passed through based on fuzzy single-point model according to past the output of process and past control action
Mode can get a series of the output of process in futures;
D, judge that the inverse system for obscuring single-point model whether there is, continue to execute downwards if existing.It is such as not present, needs
Fuzzy single-point model is split, step B is returned to;
E, according to the output of process of given value and future, the inverse system of single-point model is obscured by solution, and a system can be obtained
Arrange following control action;
F, these control actions are packaged and process end is sent to by controller end by network, when process end passes through network
Prolonging compensator selects suitable control sequence to act on controlled process to compensate the time delay in feedforward network channel;
G, in next execution period, step E and F are repeated.
Preferably, the inputoutput data that the prediction model in the step A only need to fully be encouraged according to system
It obtains, wherein the degree of membership of former piece variable is obtained by G-K algorithms, and the parameter of consequent variable is obtained by least square method.
Preferably, the former piece variable of the fuzzy single-point model of the step B uses fragment triangle membership function, stroke
Fuzzy Cluster Model will be more than by dividing, and the single-point parameter of consequent be obtained by least square method, to make fuzzy single-point model and obscure
Clustering Model is of equal value.
Preferably, being easy to cause error accumulation after the step C successive ignitions, need to use internal model control in practical application
Structure increases feedback control.
Preferably, the time delay of the forward path is fixed delay either random delay.
Preferably, a maximum permissible delay is taken, if some control sequence has been more than the maximum permissible delay,
Then the control sequence under the maximum permissible delay can be used to be calculated.
(3) advantageous effect
It can be seen from the above technical proposal that the invention has the advantages that:
1, the unclear occasion of process mechanism can be adapted for using the Method Modeling of fuzzy clustering using the present invention, as long as
The process input and output data fully encouraged can be provided, so that it may obtain prediction model in a manner of by obscurity model building.
2, using the present invention, it is substantially a nonlinear model to obscure single-point model, thus the method for the present invention can
With in the control based on network applied to Invertible nonlinearity system.
3, using the present invention, fuzzy single-point model it is inverse in the absence of, its segmentation can be made singly by dividing domain again
The mode of tune obtains inversion model, thus has wider array of application prospect.
4, it using the present invention, selects suitable control sequence to act on by network delay compensator at process end and was controlled
Journey can compensate time delay existing for feedforward network channel.
5, using the present invention, in controller design, by directly seeking the inverse acquisition control action of fuzzy single-point model,
Simple with program, calculation amount is small, is convenient for commercial Application.
Description of the drawings
Fig. 1 is a kind of realization of the network control method overall technological scheme based on fuzzy inverse model provided by the invention
Flow chart;
Fig. 2 is the network control method control structure schematic diagram provided by the invention based on fuzzy inverse model;
Fig. 3 is the internal model control structure figure of the network control method based on fuzzy inverse model;
Fig. 4 is the membership function for the former piece variable that fuzzy clustering modeling obtains in specific embodiment
Fig. 5 is that the network control method based on fuzzy inverse model is deposited with regulatory PID control method in feedforward network channel
Control effect comparison diagram in the case of 10 step random delay;
Specific implementation mode
To make the objectives, technical solutions, and advantages of the present invention clearer, below in conjunction with specific embodiment, and reference
Attached drawing, the present invention is described in more detail.
As shown in Figure 1, 2, 3, the present invention provides a kind of network control method based on fuzzy inverse model, this method packet
It includes:
A, outputting and inputting according to controlled process obtains the dynamic mathematical modulo of system by the modeling method of fuzzy clustering
Type;
B, the mathematical model of fuzzy clustering is converted to fuzzy single-point model of equal value;
C, iteration is passed through based on fuzzy single-point model according to past the output of process and past control action
Mode can get a series of the output of process in futures;
D, judge that the inverse system for obscuring single-point model whether there is, continue to execute downwards if existing.It is such as not present, needs
Fuzzy single-point model is split, step B is returned to;
E, according to the output of process of given value and future, the inverse system of single-point model is obscured by solution, and a system can be obtained
Arrange following control action;
F, these control actions are packaged and process end is sent to by controller end by network, when process end passes through network
Prolonging compensator selects suitable control sequence to act on controlled process to compensate the time delay in feedforward network channel;
G, in next execution period, step E and F are repeated.
Preferably, the inputoutput data that the prediction model in the step A only need to fully be encouraged according to system
It obtains, wherein the degree of membership of former piece variable is obtained by G-K algorithms, and the parameter of consequent variable is obtained by least square method.
Preferably, the former piece variable of the fuzzy single-point model of the step B uses fragment triangle membership function, stroke
Fuzzy Cluster Model will be more than by dividing, and the single-point parameter of consequent be obtained by least square method, to make fuzzy single-point model and obscure
Clustering Model is of equal value.
Preferably, being easy to cause error accumulation after the step C successive ignitions, need to use internal model control in practical application
Structure increases feedback control.
Preferably, the time delay of the forward path is fixed delay either random delay.
Preferably, a maximum permissible delay is taken, if some control sequence has been more than the maximum permissible delay,
Then the control sequence under the maximum permissible delay can be used to be calculated.
The step A, it is assumed that data set Z includes the inputoutput data of controlled process, y (k), y (k-1) ..., y (k-ny
+ 1) it is the output of system, u (k), u (k-1) ..., u (k-nu+ 1) it is the input of system,For output
Fuzzy set,For the fuzzy set of input, then controlled process can be indicated with T-S fuzzy structure models:
Ri:If y (k) belongs to Ai1And y (k-1) belongs to Ai2And ... and
y(k-ny+ 1) belong toAnd u (k) belongs to Bi1And
U (k-1) belongs to Bi2And ... and u (k-nu+ 1) belong to
So
According to outputting and inputting for system, the mathematical model of controlled process is established using fuzzy clustering modeling technique, including:
C cluster of selection, 1 < c < N (N is integer), controlling fuzzy parameter m > 1, convergence criterion ε > 0, random initializtion point
Cutting torch battle array.
Enable l=1,2 ..., it calculates according to the following steps:
S1 fuzzy clustering center) is calculated
Wherein, v is fuzzy clustering center, and μ is membership function, and z is the observation for being blurred matrix, and k is integer, is indicated
K-th of fuzzy subset in i-th of fuzzy clustering, other symbols are as previously shown.
S2 fuzzy clustering variance matrix F) is calculated
Wherein, T represents the transposition of matrix or vector, other symbols are as previously shown.
S3 the norm of distance D) is calculated
Wherein, AiFor the norm matrix of ith cluster, ρiFor its binding occurrence, det indicates the value of solution matrix determinant, n
For integer, other symbols are as previously shown.
S4 subdivision matrix) is updated
If for 1≤i≤c, 1≤k≤N,Then
Other:
IfWithAnd
Steps be repeated alternatively until | U(l)-U(l-1)| | < ε, wherein U represent membership function matrix, and ε is convergence criterion,
His symbol is as previously shown.
S5) with least square method identification consequent variable aij,bij
S6 the output y (k+1) of system) is obtained
The step B includes:Fuzzy Cluster Model in step A is converted to the fuzzy single-point model of following form:
Ri:If y (k) belongs to A 'i1And y (k-1) belongs to A 'i2And ... and
Y (k-p+1) belongs to A 'ipAnd u (k) belongs to B 'i1And
U (k-1) belongs to B 'i2And ... and u (k-q+1) belongs to B 'iq
So y (k+1)=Ci (7)
State vector x (k) is introduced, is made it includes p-1 past outputs, q-1 past inputs and current output,
That is x (k)=[y (k) ..., y (k-p+1), u (k-1) ..., u (k-q+1)]T, then X=A1×…×Ap×B2×…×BqSimplify
It indicates, B is replaced with B1, then the rule obscured in single-point model is represented by following formula:
If x (k) belongs to X and u (k) belongs to B, y (k+1) belongs to C (8)
If M is fuzzy set XiNumber, N be fuzzy set BjNumber, then shared K=MN rules, as shown in table 1
Table 1:Fuzzy single-point model rule base
By using t- norm operators, the degree of membership of former piece rule is
The output y (k+1) of model can be by normalizing degree of membership βijAverage weighted consequent cijIt obtains
The step C includes:Based on the fuzzy single-point model in the step B, one is obtained by way of iteration
The serial following the output of process y (k+2) ..., y (k+P+1) and y (k).
The step D includes:Judge that the inverse system for obscuring single-point model whether there is according to following theorem, if existing
Continue to execute step E downwards.It is such as not present, needs to be split fuzzy single-point model, return to step B.
Theorem 1:If bj=core (Bj) it is BjCore, the fuzzy single-point model described in step B is and if only if following item
It is reversible when part meets.
1) each BjCore be all single-point, and | Bj|=1, j=1,2 ..., N, and
2)b1<b2<…<bN→ci1<ci2<…<ciN, or b1<b2<…<bN→ci1>ci2>…>ciN, i=1,2 ..., M
The step E includes:Rule-based model (8) is equivalent to following regression model
Y (k+1)=f (x (k), u (k)) (11)
The input of model be current state x (k)=[y (k) ..., y (k-p+1), u (k-1) ..., u (k-q+1)]TWith
Current input u (k), output are the one-step prediction value y (k+1) of process.The purpose of control algolithm is to calculate control input u (k),
So that the system output in next sampling instant is equal to desired output r (k+1).If it is known that current state x (k) and ginseng
Value r (k+1) is examined, then can complete the operation by Inverse Model, control input is given by:
U (k)=f-1(x(k),r(k+1)) (12)
In general, it is difficult to find inversion model f-1.However, using monomer model structure, multivariable can be mapped (11)
It is reduced to the single argument mapping of each particular state x (k),
Y (k+1)=fx(u(k)) (13)
Therefore, if model is reversible, inverse mapping can be easily foundThen may be used
A series of control action in futures is obtained according to following theorem and step.
Theorem 2:Assuming that controlled process can be indicated with fuzzy single-point model reversible in such as step B, appoint and take x, meetsAppoint and take u, meetsFor given state x (k), it is based on the single-point mould
The inverse control law of type is provided by following rule:
If r (k+1) belongs to Cj(k), then u (k) belongs to Bj, j=1,2 ..., N (14)
Wherein, Cj:Y → [0 1] are the fuzzy sets indicated by triangle membership function.
Wherein, core cjIt is obtained by formula (16)
And c1≤c2≤…≤cN, fuzzy set BjAlso it arranges in order.Then
Wherein, bjFor BjCore.Then it can follow these steps to obscuring single-point Inverse Model.
Step 1:Measure or estimate current system mode.For input/output model, it means that according to output y (k)
More new state x (k), referring to formula (8).
2nd step:Calculate core cj,
3rd step:Check invertibity.If detecting non-monotonic, it only includes dull rule that rule base (14), which is divided into,
Two or more rule bases.
4th step:Membership function is calculated using formula (15).If generating multiple rule bases by step 3, for each rule
Then library builds individual membership function.
Step 5:Control action u (k) is calculated using formula (17).If generating multiple rule bases by step 3, need for
Each rule base calculates control action, and is selected according to additional standard.By shown in formula (12)Then u (k+1), u (k+2) ..., u (k+P) can be obtained by way of iteration, i.e.,:
In fact, the dynamic change of practical control process cannot be fully described in single-point fuzzy model.Since process model loses
Match, interference or noise, the difference between model and process are inevitable, this will cause to generate stable state mistake in opened loop control
Difference.Fortunately, it can be compensated by internal model control strategy.
The step E includes:Assuming that the delay of forward channel is no more than control time domain.Since " the packet transmission " of internet is special
Property, a series of control action u (k) in futures, u (k+1) ..., u (k+P) are packaged with from controller end in sampling instant k
It is sent to process end.At process end, can be selected from newest control sequence suitable control action with compensation network when
Prolong.For example, the newest control sequence that process end obtains is
The control signal then selected will be
U (k)=u (k | k-t) (20)
Core of the invention thought is:1) fuzzy clustering modeling technique is used to establish the T-S fuzzy models of controlled device;2)
Fuzzy Cluster Model is converted to reversible fuzzy single-point model of equal value;3) pass through the side of iteration based on obscuring single-point model
Formula obtains a series of the output of process in futures;4) according to the output of process and following given value, one is obtained by way of inverting
The serial following control action, control action is packaged, process end is sent to by network by controller end;5) logical at process end
Cross the network delay for selecting suitable control sequence that can compensate forward path.Realize the control based on network to controlled process;6)
To reduce error in practical application, internal model control structure need to be used.
Specific embodiment is illustrated by taking servo control system as an example below.The system by direct current generator, turntable and
Angular transducer forms, and control targe is by motor driving disc to a given angle.It can be with by least square method
The discrete model between system control input (voltage) and angle position (degree) is obtained, the sampling time is 0.04 second.
Controlled process and controller are put in different physical locations, are referred to as process end and control terminal, and the two passes through net
Network forms closed loop configuration, as shown in Figure 2.There are 10 step random delay in forward path, backward channel is without time delay.
Selecting range is 0.4, and frequency is three sinusoidal superposed signals of the π radian per seconds of 0.06 π, 0.012 π, 0.03 as sharp
Source is encouraged to generate the data set modeled for fuzzy clustering.Data are split into two halves.Half is modeled for fuzzy clustering, the other half
For verifying.Choose the number c=3 of cluster, controlling fuzzy parameter m=2, end condition 0.01.It can be obtained by fuzzy clustering algorithm
The membership function for obtaining former piece variable y (k) and u (k) is as shown in Figure 4.Then, point of former piece variable in fuzzy single-point model can be designed
Section triangular membership functions:
Then, the consequent parameter for obscuring single-point model can be obtained with least square method, as shown in table 2.
Table 2:Single-point model rule base is obscured in embodiment
By reversibility condition it is found that the fuzzy single-point model established is reversible.Therefore, it can be obtained control according to abovementioned steps
It makes and uses.For access control effect, the method for the present invention is compared the (ginseng of PID controller with conventional PID control effect
Number is P=0.01, I=0.001, D=0.008), Fig. 5 is that the control effect of two kinds of control methods in embodiment compares, Cong Zhongke
To find out, control method of the invention has very good control effect.
Particular embodiments described above has carried out further in detail the purpose of the present invention, technical solution and advantageous effect
It describes in detail bright, it should be understood that the above is only a specific embodiment of the present invention, is not intended to restrict the invention, it is all
Within the spirit and principles in the present invention, any modification, equivalent substitution, improvement and etc. done should be included in the guarantor of the present invention
Within the scope of shield.
Claims (5)
1. a kind of network control method based on inversion model, which is characterized in that this method includes:
A, outputting and inputting according to controlled process obtains the dynamic mathematical models of system by the modeling method of fuzzy clustering;
B, the mathematical model of fuzzy clustering is converted to fuzzy single-point model of equal value;
C, according to past the output of process and past control action, based on fuzzy single-point model, by way of iteration
It can get a series of the output of process in futures;
D, judge that the inverse system for obscuring single-point model whether there is, continue to execute downwards if existing.It is such as not present, needs to mould
Paste single-point model is split, and returns to step B;
E, according to the output of process of given value and future, the inverse system of single-point model is obscured by solution, can be obtained it is a series of not
The control action come;
F, these control actions are packaged and process end is sent to by controller end by network, mended by network delay at process end
Repaying device selects suitable control sequence to act on controlled process to compensate the time delay in feedforward network channel;
G, in next execution period, step E and F are repeated.
2. according to the method described in claim 1, it is characterized in that, the step A, it is assumed that data set Z includes controlled process
Inputoutput data, y (k), y (k-1) ..., y (k-ny+ 1) it is the output of system, u (k), u (k-1) ..., u (k-nu+ 1) it is to be
The input of system,For the fuzzy set of output,For the fuzzy set of input, then it is controlled
Process can be indicated with T-S fuzzy structure models:
Ri:If y (k) belongs to Ai1And y (k-1) belongs to Ai2And ... and
y(k-ny+ 1) belong toAnd u (k) belongs to Bi1And
U (k-1) belongs to Bi2And ... and u (k-nu+ 1) belong to
So
According to outputting and inputting for system, the mathematical model of controlled process is established using fuzzy clustering modeling technique, including:
C cluster of selection, 1 < c < N (N is integer), controlling fuzzy parameter m > 1, convergence criterion ε>0, random initializtion divides square
Battle array,
Enable l=1,2 ..., it calculates according to the following steps:
S1 fuzzy clustering center) is calculated
Wherein, v is fuzzy clustering center, and μ is membership function, and z is the observation for being blurred matrix, and k is integer, indicates i-th
K-th of fuzzy subset in a fuzzy clustering,
S2 fuzzy clustering variance matrix F), is calculated
Wherein, T represents the transposition of matrix or vector, other symbols are as previously shown.
S3 the norm of distance D), is calculated
Wherein, AiFor the norm matrix of ith cluster, ρiFor its binding occurrence, det indicates that the value of solution matrix determinant, n are whole
Number,
S4 subdivision matrix), is updated
If for 1≤i≤c, 1≤k≤N,Then
Other:
IfWithAnd
Steps be repeated alternatively until | U(l)-U(l-1)||<ε, wherein U represent membership function matrix, and ε is convergence criterion.
S5), with least square method identification consequent variable aij,bij
S6 the output y (k+1) of system), is obtained
3. according to the method described in claim 1, it is characterized in that, the step B includes:By the fuzzy clustering mould in step A
Type is converted to the fuzzy single-point model of following form:
Ri:If y (k) belongs to A 'i1And y (k-1) belongs to A 'i2And ... and
Y (k-p+1) belongs to A 'ipAnd u (k) belongs to B 'i1And
U (k-1) belongs to B 'i2And ... and u (k-q+1) belongs to B 'iq
So y (k+1)=Ci (7)
State vector x (k) is introduced, makes that it includes p-1 past outputs, q-1 past inputs and current output, i.e. x
(k)=[y (k) ..., y (k-p+1), u (k-1) ..., u (k-q+1)] T, then X=A1×…×Ap×B2×…×BqSimplify table
Show, B1 is replaced with B, then the rule obscured in single-point model is represented by following formula:
If x (k) belongs to X and u (k) belongs to B, y (k+1) belongs to C (8)
If M is the number of fuzzy set Xi, N is fuzzy set BjNumber, then shared K=MN rules, by using t- norms
The degree of membership of operator, former piece rule is:
The output y (k+1) of model can be by normalizing degree of membership βijAverage weighted consequent cijIt obtains
4. according to the method described in claim 1, it is characterized in that, the step C includes:With the fuzzy list in the step B
Based on point model, a series of the output of process y (k+2) ... in futures, y (k+P+1) and y (k) are obtained by way of iteration.
5. according to the method described in claim 1, it is characterized in that, a maximum permissible delay is taken, if some control sequence
Row have been more than the maximum permissible delay, then can be used the control sequence under the maximum permissible delay to be calculated.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810352904.9A CN108536016B (en) | 2018-04-19 | 2018-04-19 | Networked control method based on fuzzy inverse model |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810352904.9A CN108536016B (en) | 2018-04-19 | 2018-04-19 | Networked control method based on fuzzy inverse model |
Publications (2)
Publication Number | Publication Date |
---|---|
CN108536016A true CN108536016A (en) | 2018-09-14 |
CN108536016B CN108536016B (en) | 2021-11-05 |
Family
ID=63477948
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810352904.9A Expired - Fee Related CN108536016B (en) | 2018-04-19 | 2018-04-19 | Networked control method based on fuzzy inverse model |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108536016B (en) |
Citations (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO1997042553A1 (en) * | 1996-05-06 | 1997-11-13 | Pavilion Technologies, Inc. | Method and apparatus for modeling dynamic and steady-state processes for prediction, control and optimization |
CN101082811A (en) * | 2006-06-02 | 2007-12-05 | 中国科学院自动化研究所 | Self-adapting state feedback forecasting control method based on noise computing |
US20090177291A1 (en) * | 2006-09-13 | 2009-07-09 | Rockwell Automation Technologies, Inc. | Dynamic controller utilizing a hybrid model |
CN101833284A (en) * | 2010-05-28 | 2010-09-15 | 上海交通大学 | Method for compensating system in real time by using hysteresis inversion model |
CN201611867U (en) * | 2010-02-05 | 2010-10-20 | 桂林电子科技大学 | Intelligent compensation control system of non-linear characteristic of a straight line voice coil motor |
US20130055021A1 (en) * | 1996-05-06 | 2013-02-28 | Rockwell Automation Technologies, Inc. | Method and apparatus for minimizing error in dynamic and steady-state processes for prediction, control, and optimization |
CN103345153A (en) * | 2013-06-16 | 2013-10-09 | 北京联合大学 | General networked prediction fuzzy control method |
CN103399487A (en) * | 2013-07-30 | 2013-11-20 | 东北石油大学 | Nonlinear MIMO (multiple input multiple output) system-based decoupling control method and device |
CN103473786A (en) * | 2013-10-13 | 2013-12-25 | 西安电子科技大学 | Gray level image segmentation method based on multi-objective fuzzy clustering |
CN103675010A (en) * | 2013-09-22 | 2014-03-26 | 浙江大学 | Supporting-vector-machine-based industrial melt index soft measuring meter and method |
CN105960026A (en) * | 2016-07-11 | 2016-09-21 | 辽宁大学 | Wireless sensor network MAC protocol based on fuzzy clustering in traffic monitoring |
CN106168762A (en) * | 2016-09-20 | 2016-11-30 | 北京联合大学 | A kind of residual oscillation suppressing method based on fuzzy clustering inversion model |
US20170019241A1 (en) * | 2015-07-15 | 2017-01-19 | Cisco Technology, Inc. | Full duplex network architecture in cable network environments |
CN106371321A (en) * | 2016-12-06 | 2017-02-01 | 杭州电子科技大学 | PID control method for fuzzy network optimization of coking-furnace hearth pressure system |
-
2018
- 2018-04-19 CN CN201810352904.9A patent/CN108536016B/en not_active Expired - Fee Related
Patent Citations (14)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
WO1997042553A1 (en) * | 1996-05-06 | 1997-11-13 | Pavilion Technologies, Inc. | Method and apparatus for modeling dynamic and steady-state processes for prediction, control and optimization |
US20130055021A1 (en) * | 1996-05-06 | 2013-02-28 | Rockwell Automation Technologies, Inc. | Method and apparatus for minimizing error in dynamic and steady-state processes for prediction, control, and optimization |
CN101082811A (en) * | 2006-06-02 | 2007-12-05 | 中国科学院自动化研究所 | Self-adapting state feedback forecasting control method based on noise computing |
US20090177291A1 (en) * | 2006-09-13 | 2009-07-09 | Rockwell Automation Technologies, Inc. | Dynamic controller utilizing a hybrid model |
CN201611867U (en) * | 2010-02-05 | 2010-10-20 | 桂林电子科技大学 | Intelligent compensation control system of non-linear characteristic of a straight line voice coil motor |
CN101833284A (en) * | 2010-05-28 | 2010-09-15 | 上海交通大学 | Method for compensating system in real time by using hysteresis inversion model |
CN103345153A (en) * | 2013-06-16 | 2013-10-09 | 北京联合大学 | General networked prediction fuzzy control method |
CN103399487A (en) * | 2013-07-30 | 2013-11-20 | 东北石油大学 | Nonlinear MIMO (multiple input multiple output) system-based decoupling control method and device |
CN103675010A (en) * | 2013-09-22 | 2014-03-26 | 浙江大学 | Supporting-vector-machine-based industrial melt index soft measuring meter and method |
CN103473786A (en) * | 2013-10-13 | 2013-12-25 | 西安电子科技大学 | Gray level image segmentation method based on multi-objective fuzzy clustering |
US20170019241A1 (en) * | 2015-07-15 | 2017-01-19 | Cisco Technology, Inc. | Full duplex network architecture in cable network environments |
CN105960026A (en) * | 2016-07-11 | 2016-09-21 | 辽宁大学 | Wireless sensor network MAC protocol based on fuzzy clustering in traffic monitoring |
CN106168762A (en) * | 2016-09-20 | 2016-11-30 | 北京联合大学 | A kind of residual oscillation suppressing method based on fuzzy clustering inversion model |
CN106371321A (en) * | 2016-12-06 | 2017-02-01 | 杭州电子科技大学 | PID control method for fuzzy network optimization of coking-furnace hearth pressure system |
Non-Patent Citations (2)
Title |
---|
YA-LING CHANG,等: "A TSK-Type Recurrent Fuzzy Neural Network Adaptive Inverse Modeling Control for a Class of Nonlinear Discrete-Time Time-Delay Systems", 《SICE ANNUAL CONFERENCE 2010》 * |
胡贵妹,等: "基于 RBF 网络逆辨识的 LF 炉电极系统控制策略", 《荆楚理工学院学报》 * |
Also Published As
Publication number | Publication date |
---|---|
CN108536016B (en) | 2021-11-05 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN103246702B (en) | A kind of complementing method of the industrial sequence data disappearance based on segmentation Shape Representation | |
CN102968055A (en) | Fuzzy PID (Proportion Integration Differentiation) controller based on genetic algorithm and control method thereof | |
Durvaux et al. | Efficient selection of time samples for higher-order DPA with projection pursuits | |
Transtrum et al. | Measurement-directed reduction of dynamic models in power systems | |
Jiang et al. | Sunspot Forecasting by Using Chaotic Time-series Analysis and NARX Network. | |
CN109709802A (en) | The control method of active electronic ladder circuit based on iterative learning control | |
Alzalg | Decomposition-based interior point methods for stochastic quadratic second-order cone programming | |
Linaro et al. | Continuous estimation of power system inertia using convolutional neural networks | |
Wang et al. | Short-term load forecasting of power system based on time convolutional network | |
Dahiya et al. | Event‐triggered based decentralised control for frequency regulation of power systems | |
Dong et al. | Data‐driven control based on simultaneous perturbation stochastic approximation with adaptive weighted gradient estimation | |
CN108536016A (en) | A kind of network control method based on fuzzy inverse model | |
CN106371321A (en) | PID control method for fuzzy network optimization of coking-furnace hearth pressure system | |
CN115728590A (en) | Power transmission line double-end fault distance measurement method and device using same | |
Wong et al. | Model Predictive Engine Air‐Ratio Control Using Online Sequential Relevance Vector Machine | |
CN109240085A (en) | Non-Gaussian filtering dynamic data rectification and system control performance optimization method | |
CN105527841B (en) | A kind of networking tracking and controlling method of time varying signal | |
Yu et al. | Analysis and Application of the Spatio-Temporal Feature in Wind Power Prediction. | |
JP2002091505A (en) | Model identifying device | |
Santos et al. | On the structure of the equilibrium price set of overlapping-generations economies | |
Liu et al. | State‐Feedback H∞ Control for LPV System Using T‐S Fuzzy Linearization Approach | |
Wang et al. | Data‐model driven rescheduling considering both rotor angle stability and transient voltage stability constraints | |
Tosatto et al. | A temporal-difference approach to policy gradient estimation | |
Guerra et al. | Value function computation in fuzzy models by differential evolution | |
Cao et al. | Optimal control with deep reinforcement learning for shunt compensations to enhance voltage stability |
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 | ||
CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20211105 |
|
CF01 | Termination of patent right due to non-payment of annual fee |