1

DC-DC BOOST CONVERTER

CONTROL SYSTEM
Yerson David Rozo Giraldo
yerson.rozo@uao.edu.co
Engineering Faculty – Universidad Autónoma de Occidente

Abstract—This document aims to analyse, simulate, and design For left loop:
a control system for a DC-DC boost converter finding the
mathematical model, describing his behaviour, and obtaining a 𝑑𝑖
satisfactory control system for a non-lineal plant. 𝐸=𝐿 + 𝑖𝑅 (1)
𝑑𝑡

I. INTRODUCTION For the right loop:

S ome industrial applications require to obtain a high voltage

level from a constant DC power supply, in this case, we are
going to analyse a boost converter; in this type of converters, a
0=𝐶
𝑑𝑣 𝑣
+
𝑑𝑡 𝑅𝐿
(2)

higher voltage is obtained in the output. These kinds of circuits

are used to control electric motors, cranes, battery-powered
systems, and other power electronic devices.

II. MATHEMATICAL MODEL

A boost converter uses the store energy features in an inductor
and a capacitor to elevate the voltage in the output [1], the
inductor resists the abrupt changes in the current, therefore if
the current flowing in the inductor stops, the inductor reacts to Fig. 2. DC-DC boost converter MOSFT saturation
this abrupt change generating a voltage peak that will be store
in the capacitor. B. Cut-off
As can be seen, if the transistor is operating at the cut-off
The boost converter uses a MOSFET or IGBT transistor as a point, the transistor is open between collector and emitter also
switch to obtain these abrupt changes in the current and the inductor is subjected to an abrupt change in the current and
generate the voltages peaks, the basic structure of the circuit is the voltage peak is charging the capacitor and the load is
shown in Fig.1 diode protects the circuit for inverse current perceiving this voltage.
flow, in this case, it is an ideal diode.
For left loop:

𝑑𝑖
𝐸=𝐿 + 𝑖𝑅 + 𝑣 (3)
𝑑𝑡
For the node:

𝑑𝑣 𝑣
𝑖=𝐶 + (4)
𝑑𝑡 𝑅𝐿

Fig. 1. DC-DC boost converter

Notice that considering the commutation of the transistor

between cut off and saturation, there are two different cases to
analyse according to the transistor operating point.

A. Saturation
If the transistor is operating in saturation the diode is reverse Fig. 3. DC-DC boost converter MOSFET Cut off.
biased and the circuit can be rewritten as shown in Fig.2 at this
B. Mathematical model unification:
moment, the inductor in storing energy as a magnetic field and
Considering (4) and (2) it is easy to realize that the equations
the equations that describe the circuit are shown below.
can be unified in a single equation introducing a term 𝑢 that
represents the commutation effect in the transistor, this term
takes two values, 1 when the transistor is saturated or 0 when
is operating in the cut-off. This unification method can be
 2

applied to (1) and (3) in the same way, after the unification, the error redefines as the subtraction between the output and the set
system is described by (5) and (6). point.
• Relative degree
𝑑𝑣 𝑣
𝑖𝑢 = 𝐶 + (5) The relative degree (r) is the number of differentiations required
𝑑𝑡 𝑅𝐿 for the output (Y) until the input signal (U) appears, the output
𝑑𝑖 define by (7)
𝐸=𝐿 + 𝑖𝑅 + 𝑢𝑣 (6)
𝑑𝑡 𝑦 = 𝑥1
Solving these equations, it is easy to obtain the state variables 1 𝑥1
and the following block diagram Fig. 4. The system parameters 𝑦̇ = 𝑥1̇ = (𝑥 𝑢 − ) → 𝑟 = 1 (8)
𝐶 2 𝑅𝐿
are specified in Table 1.
• Sliding surface
The sliding surface is defined by the next summary:
𝑟−1
𝑑𝑣
𝑥1̇ = 𝑆(𝑒) = ∑ 𝛼(𝑖+1) 𝑒 𝑖
𝑑𝑡
𝑖=−1
𝑑𝑖
𝑥2̇ =
𝑑𝑡 𝑆(𝑒) = 𝛼0 𝑒 −1 + 𝛼1 𝑒 0 (9)
1 𝑥1 Besides, using the next definitions the sliding surface can be
𝑥1̇ = (𝑥 𝑢 − )
𝐶 2 𝑅𝐿 obtained:
1
𝑥2̇ = (𝐸 − 𝑥2 𝑅) 𝑒 −1 = ∫ 𝑒(𝑡) 𝑑𝑡
𝐿
𝑦 = 𝑥1 (7)
𝑒 0 = 𝑒(𝑡)
𝑑 𝑒(𝑡)
𝑒1 =
𝑑𝑥
𝑆̇ = −𝜂 𝑠𝑖𝑔𝑛(𝑠) (10)
Differentiating (9) and equalizing to (10) we obtain:

𝑆(𝑒) = 𝛼0 ∫ 𝑒(𝑡) 𝑑𝑡 + 𝑒(𝑡)

Fig. 4. DC-DC boost converter block diagram
𝑑 𝑒(𝑡)
TABLE I 𝑆̇ = 𝛼0 𝑒(𝑡) +
SYSTEM PARAMETERS 𝑑𝑥

Parameter Value Units 𝑆̇ = 𝛼0 𝑒(𝑡) + 𝑥1̇

𝐸 100 V 𝑥2 𝑢 𝑥1
𝑅𝐿 50 Ω 𝑆̇ = 𝛼0 𝑒(𝑡) + −
𝐶 𝐶𝑅𝐿
𝐿 400 𝜇𝐻
𝐶 25 𝜇𝐹 𝑥
(−𝜂 𝑠𝑖𝑔𝑛(𝑠) + 𝐶𝑅1 − 𝛼0 𝑒(𝑡) ) 𝐶
𝑅 0.1 Ω 𝐿
= 𝑢 (11)
𝑥2
At this moment, the non-linear system is correctly defined, and At this point, the sliding mode control can be applied to the
it is possible to start designing different kinds of controllers, plant Fig. 5, but it is mandatory to define 𝜂, and 𝛼0 they must
linearize the system or apply any other technique. be positive values, changing these values directly affect the
system transient response.
III. CONTROL SYSTEMS
𝜂 = 103 = 𝛼0
A. Sliding mode control (SMC)
This non-linear control method forces the system to “slide” until
the system's normal behaviour. Usually, the error defines as the
subtraction between the setpoint and the output, in SMC the
 3

Fig. 5. SMC

Fig. 8. Output

IV. CONCLUSIONS
Non-linear systems are especially difficult systems to control;
linearization is a helpful technique when the equilibrium point
is useful to the application, but many cases and problems
require more sophisticated techniques like sliding mode control
to be especially accurate and less restrictive than techniques that
use linearization.
Fig. 6. Voltage output
Linearization restricts the inputs to the equilibrium point, but
B. Linearization sliding mode control uses the error surface and allow the
designer to use diverse inputs, sliding mode control requires to
The next control methods require linearizing the system
adjust 𝜂, and 𝛼0 searching for the best performance.
around the equilibrium point and obtain the state space
representation, knowing the system parameters of the linearized
system define by (12) in this case, the plant has been simplified
into a SISO system.
−320𝑒3 𝑆 + 9.92𝑒9
𝐺(𝑆) = (12)
𝑆 2 + 1050 𝑆 + 2.52𝑒7
Using the transfer function, we start to design the controllers.
For the following methods, we are using MATLAB to obtain
the state space representation, integral gains, states observers
and feedback states gains.

Fig 7. State’s feedback, delay, and Smith predictor