Professional Documents
Culture Documents
Abstract—In this paper, an adaptive fuzzy fast terminal control (AF-FTSC) structure is suggested thus reducing the
synergetic voltage control scheme for DC/DC buck converter is reaching phase.
proposed based on recently developed synergetic theory and a The rest of the manuscript is structured as follows. In
terminal attractor method. The advantages of presented Section 2, the fundamental concepts related to synergetic
synergetic control include the characteristics of finite time control theory and fuzzy logic systems are reviewed briefly. In
convergence, insensitive to parameters variation and chattering Section 3, the basic model of the DC/DC buck converter in
free phenomena. Rendering the design more robust, fuzzy logic continuous conduction mode (CCM) is introduced. In Section
systems are used to approximate the unknown parameters in the 4, the adaptive fuzzy fast terminal synergetic control (AF-
proposed controller without calling upon usual model
FTSC) is developed for the DC/DC buck converter system. In
linearization and simplifications. Taking the DC/DC buck
converter in continuous conduction mode as an example, the
Section 5, simulation is performed on a DC/DC buck converter
algorithm of proposed synergetic control is analyzed in detail. All to verify the effectiveness and the applicability of the proposed
the simulation results demonstrate the effectiveness and the high AF-FTSC control technique over the FTSC strategy. In Section
dynamic capability of the proposed AF-FTSC control technique 6, the conclusions are given.
over the FTSC strategy.
II. PRELIMINARIES
Keywords-component; synergetic control, fuzzy logic system; A. The Basis of Synergetic Control(SC) strategy
terminal technique; finite time convergence; DC/DC buck
converter. In this section, we briefly introduce the fundamentals of
synergetic control synthesis for a nonlinear dynamic system
I. INTRODUCTION expressed as
Many nonlinear control approaches such as variable x = f (x, u, t) (1)
structure technique has attracted community attention in the
last decades. Due to its inherent robust features, this control Basically, the synergetic control design structure begins
technique has been extensively used in many nonlinear control with the definition of a macro-variable ߪ as a function of the
systems such as voltage regulation of DC/DC buck converter. state variables [3]
Problems of variable structure control are the chattering
phenomenon, steady state errors caused by high frequency σ = σ ( x, t ) (2)
switching actions, and the fact that it needs fairly high
bandwidths for the controller, which makes digital control The control purpose is to oblige the system to work on a
realization unfeasible [1]. On the others hand, this article pre-chosen manifold ߪ = 0.
introduces the fast terminal synergetic control (FTSC) method
which is utilized on the DC/DC buck converter voltage control Design a synergistic controller that would force the system
scheme. This approach has numerous advantages such as: it's states to exponentially approach the preferred manifold with an
excellent suitable to digital control, it functions at fixed evolution constraint which can be declared as
switching frequency which reduces the burden of filtering
design, and it shows low chattering as well as the inherent τ σ + σ = 0 ሺ͵ሻ
robustness features of sliding mode control [2].
However, the robustness of synergetic control structure is
not assured pending transient action when system states Where ߬ is a positive constant which imposes the designer
trajectories have not achieved the attractor. Therefore, a new select speed convergence to the specified attractor.
technique to avoid this difficulty intrinsic to fast terminal
synergetic control, an adaptive fuzzy fast terminal synergetic Taking into account the chain of differentiation which is
given by
Authorized licensed use limited to: Consortium - Algeria (CERIST). Downloaded on February 07,2024 at 20:18:39 UTC from IEEE Xplore. Restrictions apply.
dσ
T
ξ ( x) = ª¬ξ 1 ( x), ξ 2 ( x),..., ξ L ( x) º¼ ∈ R L ሺͳͲሻ
σ = ⋅ x ሺͶሻ
dx l
If y j are preferred as the free parameter, the above fuzzy
Substituting (1) and (3) into (4) gives (5) system (7) becomes an adaptive fuzzy system. Then, it can be
rewritten as
dσ
τ f ( x, u , t ) + σ = 0 ሺͷሻ y ( x) = θ T ξ ( x) (11)
dx
Where θ is ( L × m ) matrix, θ j and indicates the ( L × 1) j-th
Resolving (5) for u provides the synergetic control law as column of the matrix θ .
¦( ) °
L
∏ i =1 μ F ( xi )
n
l
® dVo ሺͳ͵ሻ
°̄ x2 = x1 = dt
i
l =1
And:
∏ μ (x )
n
Fil
ሺͻሻ
i =1 i
ξ ( x) = The state space average model of the DC/DC buck
¦ (∏ μ ( x ))
i L
n
i =1 Fil i converter can be rewritten as fellow
l =1
l
where y j is the point in V j at which μG ( y j ) achieves its x1 = x2
l
j °
maximum value (assuming μG ( y j ) = 1 ). ® x1 x2 Vin ሺͳͶሻ
°̄ x2 = − LC − RC + LC u
l
j
T
θ j = ª« y j , y j ,..., y j º» ∈ R L is called the adjustable parameter
1 2 L
Authorized licensed use limited to: Consortium - Algeria (CERIST). Downloaded on February 07,2024 at 20:18:39 UTC from IEEE Xplore. Restrictions apply.
e = x1 − Vref ሺͳͷሻ fˆ ( x θ f ) and gˆ( x θ g ) be the fuzzy approximates of the vector
functions ƒ(x) and g(x), respectively. Then, the controller (23)
is improved as follows
e = x2 − Vref ሺͳሻ
1 ª 1 p º
uFTSC = «Vref − fˆ ( x θ f ) − σ − α e − β e( p / q ) −1e» ሺʹͶሻ
To allow the application of AF-FTSC algorithm to a gˆ( x θ g ) ¬ τ q ¼
DC/DC buck converter, a special variable called the nonlinear
macro-variable function needs to be introduced as [7] Where
fˆ ( x θ f ) =θ Tf ξ f ( x) ሺʹͷሻ
σ = e + α e + β e p/q
ሺͳሻ
gˆ ( x θg ) =θ ξ ( x)
T
g g ሺʹሻ
Where and are the macro-variable constants, p and q are
positive odd numbers, that assure the following condition Where ξ f ( x) and ξ g ( x) are the fuzzy basis functions vector,
1<p/q<2. A constraint imposing desired dynamics to the
θ f and θ g are the parameters vector of fuzzy system.
macro-variable is selected as
τ σ + σ = 0 ሺͳͺሻ
Define the optimal estimation parameters of fuzzy system
as
Where߬ is a positive constant which imposes the designer
§ ·
θ *f = arg min ¨ sup fˆ ( x θ f ) − f ( x ) ¸
decide speed convergence to the preferred attractor. When the
macro-variable reaches ߪ = 0, the system dynamics (17) can be ሺʹሻ
θf ∈ Ωf © x∈ R n ¹
expressed as follows
§ ·
θ g* = arg min ¨ sup gˆ ( x θ g ) − g ( x ) ¸ ሺʹͺሻ
e + α e + β e p / q = 0 ሺͳͻሻ θg ∈ Ωg © x∈ R n ¹
On integrating (19), and through setting suitable parameters Where f and g represent constraint sets for f and g,
Ƚ, Ⱦ, p and q, the convergence time of the system from any respectively.
given initial condition e(0) Į 0 to the attractor ߪ = 0 is finite-
time tr ı 0 and is given by Defining the minimum estimation error as
1− ( p / q )
p § α e(0) +β ·
tr = ln ¨ ¸ ሺʹͲሻ
α ( p − q) ¨© β ¸ ε = ¬ª fˆ ( x θ *f ) − f ( x ) ¼º + ¬ª gˆ ( x θ g* ) − g ( x ) ¼º u ሺʹͻሻ
¹
The time derivative of ߪ can be rewritten as Substituting (29) into (18), the dynamic of macro-variable
is obtained as follows
d p/q
σ =
e + α e + β ሺʹͳሻ σ = ª¬ fˆ ( x θ *f ) − fˆ ( x θ f )º¼ + ª¬ gˆ ( x θg* ) − gˆ ( x θg )º¼ u − σ + ε ሺ͵Ͳሻ
e 1
dt τ
Combining (18) and (21), leads to (22) Leading after some simple manipulations to
p ( p / q) −1 1
e +α e + β
e e =− σ ሺʹʹሻ
σ = (θ *f T − θ Tf ) ξ ( x) + (θg*T − θgT ) ξ ( x)u − σ + ε ሺ͵ͳሻ
q τ 1
For the controller design we set the following simplification τ
Or in a more compact form
V x x
g ( x) = in , and f ( x ) = − 1 − 2
LC 1
LC RC
σ = θTf ξ ( x) + θgTξ ( x)u − σ + ε ሺ͵ʹሻ
τ
Solving for the terminal synergetic control algorithm uFTSC,
Where θf = θ f − θ f and θg = θ g − θ g
* *
guides to (23)
1 ª 1 p ( p / q ) −1 º
«Vref − f ( x) + σ − α e − β e e»
uFTSC = ሺʹ͵ሻ Define the Lyapunov function candidate V as
g ( x) ¬ τ q ¼
1§ 1 1 ·
Generally, the nonlinear system functions ƒ(x) and g(x) are V = ¨ σ 2 + θTf θf + θgTθg ¸ ሺ͵͵ሻ
difficult to determine precisely. So, a fuzzy logic system using 2© r1 r2 ¹
the universal approximation theorem is used to estimate the
nonlinear and unknown functions f(x) and g(x) in (23). Let Where r1 and r2 are positive constants.
Authorized licensed use limited to: Consortium - Algeria (CERIST). Downloaded on February 07,2024 at 20:18:39 UTC from IEEE Xplore. Restrictions apply.
The time derivative of V can be written as the proposed AF-FTSC algorithm has better performance in
terms of the settling time relatively to the FTSC controller.
1 1
V = σσ + θ Tf θ f + θgT θg ሺ͵Ͷሻ 40
r1 r2 Vref
Substituting (32) into (34), we can obtain (35) 30
0
1
τs
1
r1
(1
r2
)
V = − σ + σε + θTf r1σξ (x) + θf + θgT r2σξ (x)u + θg ሺ͵ሻ ( ) 0 0.01 0.02 0.03
Time (s)
0.04 0.05 0.06
(a)
To make V ≤ 0 , we choose the adaptive laws as
θf = −r1σξ ( x)
8
ሺ͵ሻ
1 1
V ≤ − σ + σε ≤ − σ ≤ 0 ሺ͵ͻሻ 0
τ τ 0 0.01 0.02 0.03 0.04 0.05 0.06
Time (s)
(b)
The block diagram of the designed AF-FTSC controller is Fig. 3. Simulation results of for a nominal start-up response.
presented in Fig. 2.
Vref
Fig. 4 shows the simulation responses of Vo when the load
resistance steps from nominal value of 20 to 5 and vice-
DC/DC Buck Converter System Vo versa. It is evident from Fig. 4 that under load resistance
x1 = x2 changes the AF-FTSC algorithm can respond faster than the
° _+
® x1 x2 Vin FTSC algorithm. Also, due to the independence of AF-FTSC
°̄ x2 = − LC − RC + LC u control law to load resistance R, the proposed controller shows
e a robust behavior against drastic changes of the load resistance.
FA-FTSC Controller
1 ª 1 p ( p / q ) −1 º
«Vref − f ( x θ f ) − σ − α e − β e
u FTSC = ˆ e » 40
gˆ ( x θ g ) ¬ τ q ¼
Decreasing of the load
30
Output voltage (V)
Adaptive Law
θf = −r1σξ ( x) Macro-Variable
σ = e + αe + βep/ q
θ = −r σξ ( x)u
20
f 2 Increasing of the load FTSC
AF-FTSC
10
Fig. 2. Block diagram for proposed AF-FTSC controller
V. SIMULATION RESULTS 0
0 0.01 0.02 0.03 0.04 0.05 0.06
In this section, some computer simulations are conducted Time (s)
using MATLAB/Simulink software to show the control Fig.4. Simulation results for a step-load response.
performance of the designed AF-FTSC control technique over
the FTSC strategy [7]. The DC-DC buck converter utilized in Figs 5(a)–5(c) depicts the output voltages tracking
simulation study has the following component values: L= 5 performance of the proposed AF-FTSC and FTSC controllers.
mH, C=1100 F, R = 20–5 , Vin= 50 V. The main parameters
used in designing controllers are: =100, Ⱦ =150, p/q =7/9, It is intuitive to see that under reference changes the
ɒ=0.005, r1=2.106 and r2=3.5.105. proposed AF-FTSC controller can respond faster than the
The start-up transient response of Vo and iL during the FTSC controller. Also, the above simulation comparisons
changes in the desired output voltage from 0 V to 30 V is effectively confirm the advantage of the proposed AF-FTSC
shown in Fig. 3.(a) and 3.(b), respectively. It is observed that algorithm over FTSC algorithm.
Authorized licensed use limited to: Consortium - Algeria (CERIST). Downloaded on February 07,2024 at 20:18:39 UTC from IEEE Xplore. Restrictions apply.
40 VI. CONCLUSIONS
FTSC
AF-FTSC
In this manuscript, an AF-FTSC control law has been
30 developed and implemented to deal with the voltage tracking
Output voltage (V)
20 REFERENCES
10 [1] Y.D. Son, T.W. Heo, E. Santi, A. Monti, “Synergetic control approach
for induction motor speed control,”, The 30th Annual Conference of the
IEEE Industrial Electronics Society, November 2-6, 2004, Busan, Korea.
0
0 0.01 0.02 0.03 0.04 0.05 0.06 [2] A. Kolesnikov, G.Veselov, A. Kolesnikov, A. Monti, F. Ponci, E. Santi
Time (s) and R. Dougal, “Synergetic synthesis of DC-DC boost converter
(b) controller: theory and experimental analysis,” IEEE APEC 2002,
pp.409–415, 2 DilZ.
[3] E. Santi, A. Monti, Li. Donghong, K. Proddutur, RA. Dougal,
40 “Synergetic control for DC-DC boost converter: implementation
options,” IEEE Trans Ind Appl, vol. 39, no. 6, pp. 1803–1813, 2003.
Output voltage (V)
Authorized licensed use limited to: Consortium - Algeria (CERIST). Downloaded on February 07,2024 at 20:18:39 UTC from IEEE Xplore. Restrictions apply.