CN108566086B - Two close cycles RBF neural sliding moding structure adaptive control system - Google Patents
Two close cycles RBF neural sliding moding structure adaptive control system Download PDFInfo
- Publication number
- CN108566086B CN108566086B CN201810329944.1A CN201810329944A CN108566086B CN 108566086 B CN108566086 B CN 108566086B CN 201810329944 A CN201810329944 A CN 201810329944A CN 108566086 B CN108566086 B CN 108566086B
- Authority
- CN
- China
- Prior art keywords
- controller
- control
- voltage
- real
- output
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02M—APPARATUS FOR CONVERSION BETWEEN AC AND AC, BETWEEN AC AND DC, OR BETWEEN DC AND DC, AND FOR USE WITH MAINS OR SIMILAR POWER SUPPLY SYSTEMS; CONVERSION OF DC OR AC INPUT POWER INTO SURGE OUTPUT POWER; CONTROL OR REGULATION THEREOF
- H02M3/00—Conversion of dc power input into dc power output
- H02M3/02—Conversion of dc power input into dc power output without intermediate conversion into ac
- H02M3/04—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters
- H02M3/10—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode
- H02M3/145—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode using devices of a triode or transistor type requiring continuous application of a control signal
- H02M3/155—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode using devices of a triode or transistor type requiring continuous application of a control signal using semiconductor devices only
- H02M3/156—Conversion of dc power input into dc power output without intermediate conversion into ac by static converters using discharge tubes with control electrode or semiconductor devices with control electrode using devices of a triode or transistor type requiring continuous application of a control signal using semiconductor devices only with automatic control of output voltage or current, e.g. switching regulators
-
- H—ELECTRICITY
- H02—GENERATION; CONVERSION OR DISTRIBUTION OF ELECTRIC POWER
- H02M—APPARATUS FOR CONVERSION BETWEEN AC AND AC, BETWEEN AC AND DC, OR BETWEEN DC AND DC, AND FOR USE WITH MAINS OR SIMILAR POWER SUPPLY SYSTEMS; CONVERSION OF DC OR AC INPUT POWER INTO SURGE OUTPUT POWER; CONTROL OR REGULATION THEREOF
- H02M1/00—Details of apparatus for conversion
- H02M1/32—Means for protecting converters other than automatic disconnection
Landscapes
- Engineering & Computer Science (AREA)
- Power Engineering (AREA)
- Dc-Dc Converters (AREA)
- Feedback Control In General (AREA)
Abstract
The present invention discloses two close cycles RBF neural sliding moding structure adaptive control system, and controller makes its output voltage stabilization to preset reference output voltage according to the desired output voltage of setting and acquired Real-time Feedback output voltage, Real-time Feedback output electric current output drive signal control Buck converter.Compared with prior art, reduce the buffeting generated during Sliding mode variable structure control using boundary layer sliding formwork control technology.Meanwhile more accurate description is carried out to the state equation of Buck converter during the work time due to the uncertain problem that system parameter variations and external unknown disturbances cause for Buck converter；And, for the uncertainty of system, the present invention increases self adaptive control on the basis of the Sliding Mode Controller of design, can carry out adaptive to external environment while can utmostly reduce the various influences interfered to Buck converter in external environment and not lose robustness.
Description
Technical field
The invention belongs to DC-DC converter automation fields more particularly to a kind of the double of Buck type DC-DC converter to close
Ring RBF neural sliding moding structure adaptive control system.
Background technique
With the development of human society, people are constantly increasing the demand of electric energy, while to the quality requirements of electric energy
Also it constantly improves.So the processing and conversion to electric energy have become indispensable one in the human lives of social development
Part.Electric energy Power Processing and transformation play increasingly important role to using electric energy aspect, thus its processing and transformation
Method has become the hot spot of area research.
Switch converters can be divided into following several citation forms: AC/DC according to power conversion type by electric energy processing unit
(rectifying conversion), AC/AC (AC-AC converter), DC/AC (inverse transformation), DC-DC (DC converting).Wherein DC-DC converter
Research belong to the scope of electronic power, theoretical method is related to the multiple subjects such as electron electric power, control theory and engineering.
Important branch of the DC-DC converter as electron electric power technology, in eighties of last century the seventies just in Europe, the U.S., day
The ground such as this started research upsurge, and be widely applied with computer, office automatic, data communication and industrial instrument and
The fields such as aerospace military.So far from last century the seventies, the research of theory analysis and control system has been achieved for big
Progress, DC-DC converter is just at an unprecedented rate towards the side such as efficient, high frequency, light-duty, green, integrated at present
To development.
DC-DC converter realizes output by changing the ratio of switching tube turn-on time as a kind of electric power converter
The adjusting of voltage, and its power bracket can from very low (baby battery) to very high (high voltage power transmission).DC-DC converter
Mainly there are pulse frequency modulated (PFM) and pulse width modulation (PWM) two ways, herein mainly with PWM type DC-DC transformation
Device is that object is studied.
Currently, DC-DC converter largely uses LINEARIZED CONTROL and Sliding mode variable structure control (SMC) technology.It is common
Proportional-integral-differential (PID) controller be that linearisation small-signal model based on controlled cell carries out performance design, no
It is suitble to that the nonlinear system of larger signal disturbance can be generated.Also, when system is there are when uncertain factor, to ensure system
There is good output performance, PID controller parameter needs passively to be adjusted repeatedly.In addition, when load wide variation, it is special
When not being band nonlinear load, switch converters have the shortcomings that dynamic responding speed is slow, output waveform has distortion.Sliding moding structure
Control is a kind of Non-Linear Control Theory, has natural applicability to the non-linear speciality of Power Electronic Switching Converters, uses
The converter of Sliding mode variable structure control have many advantages, such as stability range is wide, dynamic response is fast, strong robustness, control realize it is simple.
However, Sliding mode variable structure control process is similar to a high frequency, uncertain switch control signal, so passing through in control process
Chattering phenomenon can often occur near sliding-mode surface.So how effectively to reduce or eliminate buffeting is Sliding mode variable structure control process
In a problem frequently encountering.
Therefore in view of the drawbacks of the prior art, it is really necessary to propose a kind of technical solution to solve skill of the existing technology
Art problem.
Summary of the invention
In view of this, the present invention proposes two close cycles RBF neural sliding moding structure adaptive control system, to solve
The ineffective problem of existing Buck convertor controls realizes the good output performance of Buck converter.
In order to overcome technological deficiency of the existing technology, technical scheme is as follows:
A kind of two close cycles RBF neural sliding moding structure adaptive control system of Buck converter, includes at least
Buck converter, controller, power management module, drive module and AD sampling module, wherein the power management module is used for
Burning voltage is provided for the control system；
The drive module is used to carry out the output voltage control signal of controller driving enhancing to drive Buck to convert
Device；
The Buck converter is used for output voltage；
The AD sampling module is anti-in real time for being sampled and being will acquire to the output voltage and electric current of Buck converter
Feedback output voltage, Real-time Feedback output electric current are sent to the controller；
The controller is defeated according to the desired output voltage and acquired Real-time Feedback output voltage, Real-time Feedback of setting
Current output voltage controls signal out makes its output voltage stabilization to preset reference output voltage to control Buck converter；
The controller uses double-closed-loop control structure, including sliding moding structure adaptive controller and PID controller, institute
Stating sliding moding structure adaptive controller is outer ring voltage regulator, and the sliding moding structure approached using RBF neural is adaptive
Controller RBF-SMAC is answered, the reference inductive current as electric current loop, output equation are exported are as follows:
In formula, irFor with reference to inductive current, uiFor real-time input voltage, u be the sliding moding structure adaptive controller according to
Input parameter obtains the switching variable of sliding moding structure adaptive controller；L is inductor current value；
The PID controller is the current regulator of inner ring, PID controller PID control formula are as follows:
In formula, kp, ki, kdRespectively ratio, integral and derivative control coefficient；ei=ir-iL, iLFor Real-time Feedback inductance electricity
Stream；U is the output control amount of final system.
The sliding moding structure adaptive controller further comprises Sliding mode variable structure control as a preferred technical solution,
Device, controlled cell, Adaptable System and feedback loop；Wherein, feedback loop is missed according to reference value and output valve computing system
Difference, input of the obtained systematic error as Sliding Mode Controller；
The Adaptable System chooses the mode that RBF neural is approached, it passes through the reference value, output valve, cunning of system
Moding structure controller last time control output quantity and interference calculation go out Sliding Mode Controller it is current approach item
It is influenced for reducing error and interference to system bring；
The Sliding Mode Controller handles systematic error and adaptive fidelity term, and obtained result is as control
Amount processed is sent to controlled cell.
As a preferred technical solution, further include key module, the key module for preset desired output voltage with
And the control command for system inputs.
It as a preferred technical solution, further include display module, the display module is used to show the information of current system.
Following steps are executed in the controller as a preferred technical solution:
Step S1: the approximant sliding moding structure adaptive controller of design RBF neural realizes Voltage loop control, the cunning
Moding structure adaptive controller is according to reference output voltage urWith Real-time Feedback voltage uoObtain sliding moding structure self adaptive control
The switching variable u of device；Inductive current i is referred to as electric current loop using the output valve of Voltage looprOne of input parameter, ginseng
The calculation formula for examining inductive current is as follows:
Wherein, uiFor real-time input voltage, L is inductor current value；
Step S2: realize that current loop control, PID controller are obtained according to Voltage loop with reference to inductance electricity using PID controller
Flow irAnd the output control amount U of Real-time Feedback electric current output system that sampling obtains controls Buck converter, counts
It is as follows to calculate formula:
Wherein, ei=ir-iL, iLFor Real-time Feedback inductive current；kp, ki, kdRespectively ratio, integral and differential control system
Number.
Compared with prior art, the present invention has the following technical effect that
(1) present invention is generated in control process using boundary layer sliding formwork control technology to reduce Sliding mode variable structure control
It buffets.
(2) for Buck converter during the work time due to system parameter variations and external unknown disturbances cause not really
Qualitative question carries out more accurate description to the system mode of Buck converter, joined unknown bounded on the basis of original
Distracter, and adaptive controller is designed on this basis.
(3) it is directed to the uncertainty of system, the present invention increases RBF mind on the basis of the Sliding Mode Controller of design
It is controlled through network self-adapting, using systematic error and its derivative as the input of RBF network, the output of network is then approached as system
, and network weight is updated according to adaptive law, design a kind of novel sliding formwork change knot based on RBF neural
Structure adaptive control algorithm.Designed algorithm is that it is possible to carry out external environment adaptive same with the difference of pid algorithm maximum
When can utmostly reduce various influences of the interference to Buck converter in external environment and not lose robustness.
(4) of the invention to solve the problems such as single closed loop controlling structure stability is not strong, voltage responsive overshoot is also bigger
It uses using capacitor (output) voltage and inductive current as feedback quantity and forms respective closed loop configuration, to form double-closed-loop control
System.Wherein, outer ring is using the designed sliding moding structure adaptive controller approached based on RBF neural to output electricity
Pressure is adjusted, and inner ring is then adjusted inductive current using traditional PID controller.With the addition of current feedback amount,
It enables the system to carry out high-precision tracking, realizes the good dynamic and static characteristic of converter.
Detailed description of the invention
Fig. 1 is the whole functional block diagram of Buck control system of the present invention.
Fig. 2 is the control structure block diagram of controller in the present invention.
Fig. 3 is the functional block diagram of sliding moding structure adaptive controller of the present invention.
Fig. 4 is the program flow diagram of controller in the present invention.
Fig. 5 is Buck voltage booster circuit topological structure.
Equivalent circuit when Fig. 6 is switch conduction.
Equivalent circuit when Fig. 7 is switch OFF.
Fig. 8 is RBF neural network structure.
Fig. 9 (a)-(c) is respectively the voltage responsive of system under three kinds of control strategies, load disturbance and electric source disturbance emulation
Curve.
Figure 10 (a)-(c) is respectively that the voltage responsive of system under three kinds of control strategies, load disturbance and electric source disturbance are real
Survey curve.
Following specific embodiment will further illustrate the present invention in conjunction with above-mentioned attached drawing.
Specific embodiment
Technical solution provided by the invention is described further below with reference to attached drawing.
Nowadays, both direction is broadly divided into the research of DC-DC converter: first is that studying a kind of new converter topology knot
Structure improves energy conversion efficiency；Second is that it is good that a kind of new control performance is optimized or designed to the control algolithm first having
Good, strong robustness control strategy realizes the superior output performance of system and improves system effectiveness and stability.
Main research of the invention be the various conventional Control Methods of Buck type DC-DC converter are carried out analysis and
Comparison, finds out the control strategy of suitable Buck changer system, and combine with advanced control algolithm on this basis, proposes
A kind of superior performance, the control program of strong robustness, so as to improve the output performance of converter, to the non-linear control of supply convertor
The research of algorithm processed also has certain facilitation.
The present invention designs a kind of sliding moding structure adaptive controller of combination neural network, uses boundary layer sliding formwork control
Technology reduces shake, weakens the influence of systematic uncertainty using RBF neural Adaptable System, by the control method of design
It is applied in Buck converter, obtains good control effect.
The buffeting characteristic and converter that the present invention is directed to Sliding mode variable structure control are during the work time due to system parameter
The uncertain problem that variation and external unknown disturbances cause carries out more accurate description to state equation, on original basis
On joined the distracter of unknown bounded, and in such a way that Sliding mode variable structure control is combined with neural network algorithm to this not
Certainty item is adaptively approached.In Buck system model, using systematic error and its derivative as the input of RBF network,
The output of network is then used as system to approach item, and is updated according to adaptive law to network weight, devises a kind of based on RBF
The sliding moding structure adaptive controller (RBF-SMAC) that neural network is approached.Mind is added on the basis of Sliding mode variable structure control
It is controlled through network self-adapting, realizes faster response, effectively reduce the steady-state error of system, and being capable of self-adapting load
Variation and reduce system interference influence.
Referring to Fig. 1, show the whole functional block diagram of Buck control system of the present invention, the system include Buck converter,
Controller, power management module, drive module, AD sampling module, keyboard input module and display module, wherein power supply pipe
Reason module is used to provide burning voltage for controller system and Buck converter, and drive module is used for the output electricity of controller
Pressure carries out driving enhancing to drive Buck converter；AD sampling module is used to carry out the output voltage and electric current of Buck converter
Sample and will acquire Real-time Feedback output voltage, Real-time Feedback output electric current be sent to controller, to grasp converter in real time
Output state；This system uses double-closed-loop control structure, and controller handles the voltage and current sampled, obtains system
Voltage and current error respectively as system outer ring and inner ring feedback quantity.Key module is in addition to that can open system
The operations such as beginning, pause, reset, can be also used for the switching of control strategy.OLED display module is for showing the defeated of current system
Enter the information such as voltage, input voltage, output electric current, control algolithm type, duty ratio, facilitates observation and debugging.System protection mould
Block has the function of overcurrent protection, overheating protection, reverse connecting protection etc., implements safeguard measure immediately when system jam, prevents
System damage damage causes danger, the stability of maintenance system.
For DC-DC converter, a control structure that is reasonable, meeting the converter is selected to tend to improve converter
Stability, accuracy and conversion performance.The System control structures that prior art major part DC-DC converter uses are singly to close
Loop voltag control, this control structure design comparison is simply and readily realized, but the stability of system is not strong, voltage responsive overshoot
It measures also bigger.The control structure of controller in the present invention is shown referring to fig. 2 to solve the deficiency of single closed loop controlling structure
Block diagram, including sliding moding structure adaptive controller and PID controller, use using capacitor (output) voltage and inductive current as
Feedback quantity forms respective closed loop configuration to form double closed-loop control system.Wherein, outer ring is voltage regulator, and inner ring is electricity
Throttle regulator, both adjusters can select identical or different controller according to actual needs.With current feedback amount
It is added, enables the system to carry out high-precision tracking, realize the good dynamic and static characteristic of converter.
Wherein, the outer ring voltage regulator to play a leading role selects the sliding moding structure based on RBF neural adaptive
Controller RBF-SMAC, it can quickly track given reference voltage, have good control performance.The electric current tune of inner ring
Section device selects traditional PID controller, and the addition of electric current loop can be able to be not only that system realizes high-precision tracking, can be with
The maximum current and output power of limitation system, automatic protection converter and driving circuit guarantee the fortune of system safety and stability
Row.
Since this system uses double-closed-loop control structure, sliding moding structure adaptive controller (Voltage loop) output is as electricity
Flow the reference inductive current of ring, output equation are as follows:
In formula, irFor with reference to inductive current, i.e. the output control amount of sliding moding structure adaptive controller；uiIt is defeated in real time
Enter voltage, urFor reference output voltage, uoFor Real-time Feedback voltage, u is that the sliding moding structure adaptive controller is joined according to input
Number obtains the switching variable of sliding moding structure adaptive controller；L is inductor current value.
PID controller PID control formula are as follows:
In formula, kp, ki, kdRespectively ratio, integral and derivative control coefficient；ei=ir-iL, iLFor Real-time Feedback inductance electricity
Stream；U is the output control amount of final system.
Referring to Fig. 3, it show the functional block diagram of sliding moding structure adaptive controller of the present invention, including sliding moding structure control
Device, controlled cell, Adaptable System and feedback loop processed.Feedback loop according to reference value and output valve computing system error,
Input of the obtained systematic error as Sliding Mode Controller.Adaptable System chooses the mode that RBF neural is approached,
It controls output quantity and interference calculation by the reference value of system, output valve, controller last time and goes out that controller is current to force
Nearly itemIt is influenced for reducing error and interference to system bring.Sliding Mode Controller is to systematic error and adaptively
It approaches item to be handled, obtained result is sent to controlled cell as control amount.System, can also be by other than receiving control amount
To the influence of system interference, system interference mainly includes uncertainty and external environment as caused by Internal system parameters variation
Change caused unknown disturbances.A part of system interference is sent to Adaptable System, and Adaptable System can be effectively
Reduce influence of the interference to system, system is made to remain at optimal or suboptimum state at runtime.
In a preferred embodiment, controller uses single-chip microcontroller.
Referring to fig. 4, it show the program flow diagram of controller in the present invention, after system power supply, modules is carried out just
Beginningization is prepared for converter starting；When converter starts detection circuit whether overcurrent, overheat, then system if a failure occurs
It is out of service；After starting successfully, expectation output voltage is arranged by key module, then real-time output voltage is sampled, is obtained
Input quantity as RBF-SMAC sliding moding structure adaptive controller of result and setting value, adjust output voltage；Then again
Take PID control strategy that inductive current is adjusted.
Controller exports electricity according to the desired output voltage of setting and acquired Real-time Feedback output voltage, Real-time Feedback
Stream output drive signal control Buck converter makes its output voltage stabilization to preset reference output voltage, wherein controller
Middle execution following steps:
Step S1: the approximant sliding moding structure adaptive controller of design RBF neural realizes Voltage loop control, the cunning
Moding structure adaptive controller is according to reference output voltage urWith Real-time Feedback voltage uoObtain sliding moding structure self adaptive control
The switching variable u of device；And inductive current i is referred to as electric current loop using the output valve of Voltage looprOne of input parameter,
Calculation formula with reference to inductive current is as follows:
Wherein, uiFor real-time input voltage, L is inductor current value；
Step S2: realize that current loop control, PID controller are obtained according to Voltage loop with reference to inductance electricity using PID controller
Flow irAnd the Real-time Feedback electric current output drive signal U that sampling obtains controls Buck converter, the meter of driving signal
It is as follows to calculate formula:
Wherein, e=ir-iL；iLFor Real-time Feedback inductive current；kp, ki, kdRespectively ratio, integral and differential control system
Number.
Further, the design of controller further comprises following steps:
Step 1: establishing Buck changer system model；
It is Buck reduction voltage circuit topological structure and its isoboles, wherein u such as Fig. 5iFor input voltage；uoFor output (electricity
Hold) voltage；L is energy storage inductor；iLFor inductive current；VT is power switch tube；DT is freewheeling diode；C is filter capacitor；R
For load resistance.
Buck converter at electric current continuous operation mode (CCM) is studied, Buck circuit can be obtained according to analysis
State equation:
It is indicated with matrix equation are as follows:
In formula, u is switch function, is defined as follows:
In formula, T is switch periods, and D is duty ratio, and in ccm mode
Step 2: sliding variable design；
The approximant sliding moding structure adaptive controller of neural network is designed, first Sliding Mode Controller is carried out
Design.The error of system defines shown in an accepted way of doing sth (5):
E=ur-uo (5)
In formula: urRepresent desired output voltage, uoRepresent reality output (capacitor) voltage.Definition status variable:
x1=uo (6)
To x1State variable x is obtained into derivation2:
System state equation can be obtained according to formula (6) and formula (7) arrangement:
In formula, f (t) is system interference, | f (t) |≤F, F > 0.In order to simplify derivation process, enable Then formula (8) simplifies are as follows:
By sliding variable s is defined as:
In formula: λ > 0.After s derivation are as follows:
U can be designed to following expression by observation type (11):
In formula, η >=F,
Formula (12) is brought into formula (11) and is arranged:
Expression formula then can be obtained:
Formula (14) meets the stable condition of Lyapunov, shows the reasonability of u design.
Step 3: neural network approaches item design；
Fig. 3 is the functional block diagram of sliding moding structure adaptive control algorithm, it includes Sliding Mode Controller, controlled list
Member, Adaptable System and feedback loop.Feedback loop according to reference value and output valve computing system error, miss by obtained system
Input of the difference as Sliding Mode Controller.Adaptable System chooses the mode that RBF neural is approached, it passes through system
Reference value, output valve, controller last time control output quantity and interference calculation go out controller it is current approach itemFor
Reduce error and interference influences to system bring.Sliding Mode Controller to systematic error and adaptive fidelity term at
Reason, obtained result are sent to controlled system as control amount.System also suffers from system interference other than receiving control amount
It influences, system interference mainly includes that uncertainty and external environment as caused by Internal system parameters variation change and causes
Unknown disturbances.A part of system interference is sent to Adaptable System, and Adaptable System can effectively reduce interference to being
The influence of system makes system remain at optimal or suboptimum state at runtime.
In RBF network, input vector X=[x1,x2,...,xm]TIt indicates, and the output of network is the mark of input
Quasi-functionExpression formula is as follows:
In formula, n is the quantity of hidden layer neuron；Ci=[ci1,ci2,...,cim]TFor hidden neuron i center to
Amount；wiIt is the weight of the neuron i in linear convergent rate neuron.Radial basis function usually selects Gaussian function
hi(||X-Ci| |)=exp (- βi||X-Ci||2) i=1,2 ..., n (16)
In formula: βiFor the width of hidden layer neuron i, and βi> 0.Gaussian bases are local actions for center vector
, i.e.,
Influence of the change of one neuron parameter for the network layer input value far from the neuron is very little.It is given
The certain condition of activation primitive, RBF network can be used as RnThe general of compact subset approaches device.This means that one has enough
Any continuous function that the RBF network of the hidden neuron of quantity can be closed with the bounded closed set that arbitrary accuracy approaches.
After being completed to the Sliding Mode Controller design on basis, starts to carry out the design that neural network approaches item, use
In approaching due to system parameter (such as a1,a2,a3) the caused uncertain and system interference of variation.
If the input of network are as follows:
The output of network is arranged are as follows:
Nm(X)=WTH(X)+ε (19)
In formula, H (X) is the radial base vector of network hidden layer；W is the connection weight of network；ε approaches mistake for network
Difference, and | ε | < εN, εNFor the positive real number of very little.
H (X)=[h1 h2 h3 … hn]T (20)
Wherein, hi(i=1,2 ... expression n) is listed in formula (16).
W=[w1 w2 w3 … wn]T (21)
If network output approaches item in real time are as follows:
In formula,Term coefficient is approached for W:
Observation type (12) enables
Nm(X)=a1x2+a2x1 (24)
Then according to formula (12) and formula (24), u can be write as:
Formula (25) indicates that the Sliding mode variable structure control design that RBF neural approaches item is added.
Step 4: adaptive law design；
Formula (25) are substituted into formula (11) to obtain:
Definition
In formula,ThenIt may be expressed as:
The design acquires the expression formula of sliding moding structure adaptive controller using Lyapunov direct method, original
The adaptive item of neural network is added in liapunov function:
In formula: α1> 0, α2> 0.
Derivation is carried out to V to obtain:
According to formula (30), adaptive law is chosen are as follows:
Step 5: system stability analysis；
For access control device and the reasonability of adaptive design, formula (31) is brought into formula (30):
Due to α1> 0, so when choosing η >=εNWhen+F,Perseverance is set up, and meets the stable condition of Lyapunov at this time,
To enable systematic error to converge to zero.Theoretical foundation is provided for subsequent system emulation.η sign in formula (32)
(s) main function of item can be understood as reducing neural network approximate error and unknown disturbances to influence caused by system.
Step 6: voltage ring design；
For DC-DC converter, a control structure that is reasonable, meeting the converter is selected to tend to improve converter
Stability, accuracy and conversion performance.The System control structures that most of DC-DC converter uses are single closed loop voltage control
System, this control structure design comparison is simply and readily realized, but the stability of system is not strong, and voltage responsive overshoot also compares
Greatly.For the deficiency for solving single closed loop controlling structure, the present invention is used using capacitor (output) voltage and inductive current as feedback quantity
Respective closed loop configuration is formed to form double closed-loop control system.Wherein, outer ring is voltage regulator, and inner ring is current regulation
Device, both adjusters can select identical or different controller according to actual needs.With the addition of current feedback amount, make
The system of obtaining is able to carry out high-precision tracking, realizes the good dynamic and static characteristic of converter.
The control block diagram of system is as shown in Fig. 2, the outer ring voltage regulator to play a leading role is selected based on RBF neural
The sliding moding structure adaptive controller RBF-SMAC approached, it can quickly track given reference voltage, have good
Control performance.The current regulator of inner ring selects traditional PID controller, and the addition of electric current loop can be able to be not only system
It realizes high-precision tracking, the maximum current and output power of system, automatic protection converter and driving electricity can also be limited
Road guarantees that system is safely and steadily run.
Since this system uses double-closed-loop control structure, reference inductive current of the output of Voltage loop as electric current loop, root
It can be obtained according to formula (1):
In formula, irTo refer to inductive current (i.e. the output control amount of sliding moding structure adaptive controller RBF-SMAC), ui
For real-time input voltage.Formula (33) both sides are integrated and abbreviation can obtain the expression formula of sliding moding structure adaptive controller:
Sliding formwork control signal be it is discrete, in formula (25) contain sign function sign (s), when sliding variable reaches sliding formwork
Shake can be generated when plane, in order to reduce discrete shake, we use the symbol in saturation function sat (s) substituted (34)
Function sign (s):
In formula, the expression formula of saturation function sat (s) are as follows:
In formula, constant δ > 0, δ be diverter surface boundary layer the upper limit, using saturation function for boundary layer outside motor point
It is acted on to talk about switching, and the motor point of inside boundary is then linear change.By choosing suitable δ value, enable error
Zero is converged to, to reduce buffeting.Formula (35) is that the sliding moding structure adaptive controller of mode is approached using RBF neural
Final expression formula.
Step 7: design of current ring；
It is next then the design to electric current loop after the sliding moding structure adaptive controller design of Voltage loop is completed,
Current regulator selects PID controller.It is assumed that the error of electric current loop defines shown in an accepted way of doing sth (37):
ei=ir-iL (37)
In formula, iLFor actual inductive current.PID control formula are as follows:
In formula, kp, kp, kpRespectively ratio, integral and derivative control coefficient；U is the output control amount of final system.
Emulation experiment:
In order to verify design sliding moding structure adaptive controller effect, establish model in MATLAB and imitated
Very, certainly to the sliding moding structure of PID controller, traditional Sliding Mode Controller (CSMC) and RBF neural optimization
Adaptive controller (RBF-SMAC) compares and analyzes.
First to the present invention relates to some Buck transducer parameters be briefly described, shown in table 2 specific as follows.
2 Buck transducer parameters of table
Referring to Fig. 9, it show PID, CSMC, RBF-SMAC simulation curve, wherein (a), (b), (c) respectively correspond voltage
The case where response, load disturbance, electric source disturbance, specific test data is as shown in the following table 3.
The comparison of 3 PID, CSMC, RBF-SMAC simulation performance of table
Actual measurement experiment:
The program write is downloaded in system controller with emulator；By slide rheostat change its resistance value into
Row system load disturbance experiments；The data that system generates can be sent to host computer using serial ports convenient for observation in real time.By
Tracking curve of output under host computer is once only able to display a kind of control strategy, then by three kinds of control strategies in host computer
Experimental data under (PID, CSMC, RBF-SMAC) is acquired, and is handled using Matlab these three experimental datas, so
After be shown on same figure, make in this way measured result convenient for observation and compare.
The relevant parameter for adjusting PID, CSMC, RBF-SMAC, is optimal the control performance of three kinds of controllers, carries out real
Survey comparative experiments.Figure 10 (a)-(c) is respectively voltage responsive, load disturbance and the electric source disturbance of system under three kinds of control strategies
Measured curve.Specific measured data is as shown in table 4 below.
The comparison of 4 PID, CSMC, RBF-SMAC measured performance of table
Measured data can slightly increase than emulation data, but survey substantially uniform with the comparison result of emulation.Starting
Response phase, the response time ratio CSMC of RBF-SMAC lack about 3ms, about 24ms fewer than PID, and voltage overshoot only has 2.88%；
When the load disturbance stage, the regulating time ratio CSMC of RBF-SMAC has lacked about 7ms, fewer than PID 19ms, and RBF-SMAC is controlled
Voltage disturbance amount under system only has 1.05%, is better than CSMC and PID, these all embody FASMAC with very strong anti-interference energy
Power；In the electric source disturbance stage, voltage disturbance momentum caused by RBF-SMAC is minimum, and regulating time ratio CSMC lacks about 13ms, than
PID lacks about 22ms, and the strong robustness that RBF-SMAC is showed when further illustrating in the presence of interference is in CSMC and PID；When being
System reaches stable state, RBF-SMAC can adaptive approximation system indeterminate variation to reduce error.Pass through above actual measurement
Experimental verification RBF-SMAC control performance designed under identical conditions is better than PID and CSMC.
It is noted that the control performance in emulation is more preferable than the result in actual measurement.The reason is as follows that: 1) in emulation experiment, institute
Element be all it is ideal, such as inductance and capacitor, i.e., will not change when system is run；However, element in practice
True value with ideal value be it is devious, this is an important factor for influencing experimental precision；2) in the hardware design, reality output
Voltage obtains in such a way that electric resistance partial pressure is sampled by Chip Microcomputer A/D again, and there are errors in the process for this, this also affects control
Precision；3) it is worth noting that, the period of host computer acquisition experimental data is 5ms, this is that the control performance of system is caused to reduce
Other factors.Therefore, adjustment appropriate is carried out using the element of high-quality and to the Hardware Design, this is to further changing
Kind tracking performance plays very important effect.In addition, being needed in systems using the advanced microcontroller with high-speed sampling rate
Device and host computer, to obtain more accurately data.
The above description of the embodiment is only used to help understand the method for the present invention and its core ideas.It should be pointed out that pair
For those skilled in the art, without departing from the principle of the present invention, the present invention can also be carried out
Some improvements and modifications, these improvements and modifications also fall within the scope of protection of the claims of the present invention.
The foregoing description of the disclosed embodiments enables those skilled in the art to implement or use the present invention.
Various modifications to these embodiments will be readily apparent to those skilled in the art, as defined herein
General Principle can be realized in other embodiments without departing from the spirit or scope of the present invention.Therefore, of the invention
It is not intended to be limited to the embodiments shown herein, and is to fit to and the principles and novel features disclosed herein phase one
The widest scope of cause.
Claims (5)
1. two close cycles RBF neural sliding moding structure adaptive control system, which is characterized in that include at least Buck and convert
Device, controller, power management module, drive module and AD sampling module, wherein the power management module is used to be the control
System provides burning voltage；
The drive module is used to carry out the output voltage control signal of controller driving enhancing to drive Buck converter；
The Buck converter is used for output voltage；
The AD sampling module is defeated for being sampled to the output voltage and electric current of Buck converter and will acquire Real-time Feedback
Voltage, Real-time Feedback output electric current are sent to the controller out；
The controller exports electricity according to the desired output voltage of setting and acquired Real-time Feedback output voltage, Real-time Feedback
Flow output voltage control signal makes its output voltage stabilization to preset reference output voltage to control Buck converter；
The controller uses double-closed-loop control structure, including sliding moding structure adaptive controller and PID controller, the cunning
Moding structure adaptive controller is outer ring voltage regulator, and the sliding moding structure approached using RBF neural is self-adaptive controlled
Device RBF-SMAC processed exports the reference inductive current as electric current loop, output equation are as follows:
In formula, irFor with reference to inductive current, uiFor real-time input voltage, uoFor Real-time Feedback voltage, u be the sliding moding structure from
Adaptive controller obtains the switching variable of sliding moding structure adaptive controller according to input parameter；L is inductor current value；
The PID controller is the current regulator of inner ring, PID controller PID control formula are as follows:
In formula, kp, ki, kdRespectively ratio, integral and derivative control coefficient；ei=ir-iL, iLFor Real-time Feedback inductive current；U
For the output control amount of final system.
2. two close cycles RBF neural sliding moding structure adaptive control system according to claim 1, feature exist
In, the sliding moding structure adaptive controller further comprise Sliding Mode Controller, controlled cell, Adaptable System with
And feedback loop；Wherein, feedback loop is according to reference value and output valve computing system error, and obtained systematic error is as sliding formwork
The input of variable-structure controller；
The Adaptable System chooses the mode that RBF neural is approached, it is become by the reference value, output valve, sliding formwork of system
Structure controller last time control output quantity and interference calculation go out Sliding Mode Controller it is current approach itemFor
Reduce error and interference influences to system bring；
The Sliding Mode Controller handles systematic error and adaptive fidelity term, and obtained result is as control amount
It is sent to controlled cell.
3. two close cycles RBF neural sliding moding structure adaptive control system according to claim 1 or 2, feature
It is, further includes key module, the key module is used to preset desired output voltage and control command for system is defeated
Enter.
4. two close cycles RBF neural sliding moding structure adaptive control system according to claim 1 or claim 2, feature exist
In further including display module, the display module is used to show the information of current system.
5. two close cycles RBF neural sliding moding structure adaptive control system according to claim 1 or 2, feature
It is, executes following steps in the controller:
Step S1: the approximant sliding moding structure adaptive controller of design RBF neural realizes Voltage loop control, which becomes
Structure adaptive controller is according to reference output voltage urWith Real-time Feedback voltage uoObtain sliding moding structure adaptive controller
Switching variable u；Inductive current i is referred to as electric current loop using the output valve of Voltage looprOne of input parameter, with reference to electricity
The calculation formula of inducing current is as follows:
Wherein, uiFor real-time input voltage, L is inductor current value；
Step S2: realize that current loop control, PID controller are obtained according to Voltage loop with reference to inductive current i using PID controllerrWith
And the output control amount U of Real-time Feedback electric current output system that sampling obtains controls Buck converter, the calculation formula of U
It is as follows:
Wherein, ei=ir-iL, iLFor Real-time Feedback inductive current；kp, ki, kdRespectively ratio, integral and derivative control coefficient.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810329944.1A CN108566086B (en) | 2018-04-13 | 2018-04-13 | Two close cycles RBF neural sliding moding structure adaptive control system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810329944.1A CN108566086B (en) | 2018-04-13 | 2018-04-13 | Two close cycles RBF neural sliding moding structure adaptive control system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108566086A CN108566086A (en) | 2018-09-21 |
| CN108566086B true CN108566086B (en) | 2019-09-17 |
Family
ID=63534897
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810329944.1A Active CN108566086B (en) | 2018-04-13 | 2018-04-13 | Two close cycles RBF neural sliding moding structure adaptive control system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108566086B (en) |
Families Citing this family (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109149923B (en) * | 2018-10-08 | 2020-05-26 | 淮阴工学院 | APFC control system for estimating control rate on line |
| CN109884452B (en) * | 2019-03-25 | 2020-05-22 | 江南大学 | Fault diagnosis method of Buck converter based on high-order neural network |
| CN110048606B (en) * | 2019-05-23 | 2020-12-25 | 哈尔滨工业大学 | DC-DC boost converter dynamic sliding mode voltage control method based on interval two-type self-adaptive fuzzy neural network |
| CN113641096B (en) * | 2021-08-01 | 2024-02-02 | 西北工业大学 | Self-adaptive reconfigurable proportional-integral-derivative controller based on BP neural network |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102636995B (en) * | 2012-05-03 | 2014-07-30 | 河海大学常州校区 | Method for controlling micro gyro based on radial basis function (RBF) neural network sliding mode |
| CN102638186A (en) * | 2012-05-18 | 2012-08-15 | 上海三一精机有限公司 | Three-phase voltage type rectifier and control method thereof |
| CN104184149B (en) * | 2014-08-21 | 2017-05-24 | 上海电力学院 | Voltage fluctuation stabilizing method based on sliding mode control and super-capacitor |
| CN106950823A (en) * | 2016-01-07 | 2017-07-14 | 常州峰成科技有限公司 | A kind of Neural network PID System with Sliding Mode Controller for intelligent wheel chair |
| CN105576972B (en) * | 2016-01-26 | 2018-06-01 | 江苏大学 | A kind of buck converter without buffet sliding-mode control |
| CN106169754B (en) * | 2016-09-21 | 2018-10-26 | 河海大学常州校区 | Active Power Filter-APF neural network dynamic PID total-sliding-mode control methods |
| CN107017768B (en) * | 2017-05-21 | 2019-03-22 | 重庆大学 | Buck converter control system and method based on quasi-optimal sliding formwork control |
-
2018
- 2018-04-13 CN CN201810329944.1A patent/CN108566086B/en active Active
Also Published As
| Publication number | Publication date |
|---|---|
| CN108566086A (en) | 2018-09-21 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108566088B (en) | Two close cycles RBF neural sliding moding structure self-adaptation control method | |
| CN108566086B (en) | Two close cycles RBF neural sliding moding structure adaptive control system | |
| Dileep et al. | An improved particle swarm optimization based maximum power point tracking algorithm for PV system operating under partial shading conditions | |
| CN108539978B (en) | A kind of adaptive control system of Boost type DC-DC converter | |
| CN108566087B (en) | A kind of self-adaptation control method of Boost type DC-DC converter | |
| CN101604848A (en) | The modified fuzzy sliding mode controlling method of monopole three-phase photovoltaic grid-connected system | |
| CN107017768A (en) | Buck converter control systems and method based on quasi-optimal sliding formwork control | |
| Nikhar et al. | Review of various control techniques for DC-DC interleaved boost converters | |
| Hu et al. | Maximum power point tracking control of a high power dc-dc converter for PV integration in MVDC distribution grids | |
| Hariharan et al. | Near optimal controller tuning in a current-mode DPWM boost converter in CCM and application to a dimmable LED array driving | |
| Debnath et al. | Particle swarm optimization-based pid controller design for dc-dc buck converter | |
| Panda et al. | A comparative study of PI and fuzzy controllers for solar powered DC-DC boost converter | |
| Al-Qaisi et al. | High performance DC/DC buck converter using sliding mode controller | |
| CN109742941A (en) | DC-DC converter chaotic control method, system and medium based on supertwist control | |
| CN106681424B (en) | A kind of solar energy power generating MPPT control systems and control method | |
| Sarkar et al. | Comparison of different types of internal model controller architecture for a boost converter | |
| Jayawardana et al. | Boundary control with corrected second-order switching surface for buck converters connected to capacitive loads | |
| Tong et al. | Nonlinear characteristics of DAB converter and linearized control method | |
| Thirumeni et al. | Design and analysis of zeta converter for power factor correction using cascade PSO-GSA-tuned pi and reduced-order SMC | |
| Mohajery et al. | PSO-TVAC Based Optimized FOPID Controller for Cascaded DC-DC Boost Converter | |
| Podey et al. | Stability analysis of DC-DC boost converter using sliding mode controller | |
| Yun et al. | Double closed loop control of DC-DC boost converter | |
| Sarkar et al. | Comparison of advanced analog controllers for a DC-DC boost converter | |
| Ortatepe et al. | The performance analysis of AC-DC bridgeless converter using fuzzy self-tuning and comparing with PI control method | |
| Mushi et al. | Design for nonlinear current reference deadbeat control for boost converter |
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 | ||
| EE01 | Entry into force of recordation of patent licensing contract | ||
| EE01 | Entry into force of recordation of patent licensing contract |
Application publication date: 20180921 Assignee: HANGZHOU KONXIN SOC Co.,Ltd. Assignor: HANGZHOU DIANZI University Contract record no.: X2021330000825 Denomination of invention: Double closed loop RBF neural network sliding mode variable structure adaptive control system Granted publication date: 20190917 License type: Common License Record date: 20211220 |