Power Electronics

Complex Engineering Problem

Report

Design and Control of a Buck Converter

Submitted to:
Engr. Dr. Arsalan Ansari

Submitted By:
D-19F-ES-22 --- Abu Bakar
D-19F-ES-08 --- Muhammad Affan
D-19F-ES-06 --- Ayesha Jawaid
D-19F-ES-14 --- Javeria Shariq
 Introduction
A direct current to direct current (DC-DC) converter is an electrical system (device) that
transforms direct current (DC) sources from one voltage level to the other. In other words, a
DC-DC converter takes a DC input voltage and produces a distinct DC voltage. The output DC
voltage might be more or less than the input DC voltage. A DC-DC converter, as the name
indicates, only works with direct current (DC) sources and not with alternating current (AC).
A DC-DC converter may also be referred to as a DC-DC power converter, voltage regulator, or
choppers. We will look at the Step-Down Chopper or Buck converter in this report, which
decreases the input DC voltage to a particular DC output voltage.
The polynomial controller is chosen as the control approach to maintaining the buck converter's
output voltage. The polynomial controller compares the actual output voltage to the reference
value. The difference in voltages prompts the control element to adjust the output voltage to
the fixed reference voltage level. This is referred to as voltage regulation.
The main goal of this report is to construct a buck converter controller based on discrete
polynomial controller theory. The converter control signal is implemented as a PWM signal,
converting the system to a switched system. It is thus relevant to use simulations to determine
the nonlinear system parameters. The control design should be assessed based on the accuracy
of the output voltage level in the presence of various types of disturbances. The disturbances
are fluctuations in load current, measurement noise, and parameter variations.
A Buck converter is made consisting of a transistor and a diode that apply the supply voltage to
an inductor-capacitor, or LC, circuit. The voltage across the capacitor is the output voltage.

The LC circuit's input voltage is regulated using pulse width modulation, or PWM. In other
words, during a portion of a cycle time, the voltage applied to the LC circuit is Vg, and for the
remainder of the cycle time, the input voltage is zero. The duty cycle range, d=[0, 1], is the
portion of the cycle time when the supply voltage is connected to the circuit, i.e. the input is Vg.
The control signal to the Buck converter is the duty cycle ‘d’.
 Literature Review
Scenario-Based Approach for Control Design for DC-DC Buck Converter:
This approach entails the study of a wide variety of control system analysis and design issues for
uncertain systems that are capable of solving numerical problems efficiently provided the
requirements for robustness are enforced in a probabilistic sense inside the concept. Changes in
load current can affect the converter's output voltage. In the dynamic model of the converter, this
fluctuation in load appears as a linear unknown parameter with a known range. The challenge of
keeping the output voltage constant in the presence of load changes is expressed as an LPV H-
infinity optimization problem. However, rather than covering the entire uncertainty set with the
conventional polytopic approach, we use the scenario-based approach to extract the scenarios
explicitly and provide a bound on the number of scenarios, resulting in a design to ensure
particular probabilistic robustness in the design of a control problem for a DC-DC Buck
Converter
The main idea behind the development of a control scheme is to guarantee the working of a
system in a real-time environment. H-infinity technique has the advantage over classical control
methods as we can deal with convex optimization and in that it is easily applicable to the
problems involving multivariable systems (MIMO).
The control technique based on the scenario approach by solving linear matrix inequalities
(LMIs) has been applied. LMIs may be used to assess and solve control constraints such as
disturbance rejection of external signals, resilience in uncertain systems, and limited control
signals. For example, design performance, stability, and control restrictions can all be evaluated
concurrently, phrased in the same way, and solved in a single optimization problem.

[1]
CONTROL OF BUCK CONVERTER BY POLYNOMIAL, PID, AND PD CONTROLLERS:
The control methods are used to keep the buck converter's output voltage steady. Controlling a
buck converter may be achieved utilizing a variety of controllers. In our thesis, we employ the
Polynomial Controller and the Proportional Integral Derivative [PID] Controller. The actual
output voltage is compared to the reference voltage using these controlling approaches. The
difference between these values causes the control element to dampen changes in the output
voltage to the set reference voltage, a process called voltage regulation.
The control technique of the converter is based on polynomial correctors for energy management
between the HEVs (electric load) and the hybrid sources (battery and supercapacitors). The
suggested control approach is divided into two parts. The first stage is concerned with current
control for the SC, while the second is concerned with dc-bus voltage management using battery
current control as an inner loop. To manage the SC current, the bidirectional converter control
laws obtained from the buck-boost converters modeling are used.

For buck converter;

To develop controller structure from the pole placement controller, which gives the closed
system of exact poles defined by own dynamics characteristics.
This does not include any direct comparison between the set-point and actual value according
to rules of error. But it exposes the set-point value and actual value for the more individual
treatment doing so, is and polynomials with the delay operator z−1. As our buck converter plant
is of third order we have considered five parameters, from the design and output of the second-
order polynomial pole placement controller i.e., and to achieve necessary design freedom.
 [2][3]
Polynomial Controller Design and Its Application: Experimental Validation on a Laboratory
Setup of Nonideal DC-DC Buck Converter:
One of the most significant challenges in the design of a control system is transient response
control. The link between characteristic ratios and time-domain criteria was used to define the
transient response. The characteristic ratios (CR) approach's design parameters are the
characteristic ratios. This paper introduces a new polynomial controller for all forms of linear
second-order systems. The suggested controller is built with CRA. This controller's key strengths
are its straightforward design processes, set-point tracking, and disturbance rejection capabilities.
To show the performance and durability of the proposed controller, it is initially constructed for
an unstable non-minimum-phase second-order system. The provided controller is next examined
and validated on a nonideal dc-dc buck converter.
Characteristic ratio assignment (CRA) is a controller design technique in which the design
parameters are characteristic ratios and the generalized time constant.  The CRA is a polynomial
control design method. Using the CRA technique, one may get decreased settling time and non-
over-shoot reaction, both of which are desirable criteria in any design. Furthermore, the CRA is
resistant to parametric uncertainty in system dynamics. Furthermore, the CRA-based controller
can reduce the influence of unknown disturbances and noise on the system. Finally, in the
control system, CRA is a strong control approach for set-point tracking and disturbance
rejection.

[4]
An Intelligent Adaptive Control of DC-DC Buck Converters:
In many applications, a buck DC-DC converter is utilized to deliver a set quantity of DC voltage.
They are extremely sensitive to rapidly changing loading conditions. Such a condition
necessitates a powerful control system capable of ensuring the buck converter's good functioning
in the face of a rapidly changing load. This is made feasible by constructing an adaptive control
method capable of estimating the correct values of the unknown load parameters in the shortest
amount of time. This paper proposes a backstepping control technique based on an adaptive
Chebyshev neural network (CNN) for the output voltage regulation of a DC-DC buck converter.
A controller is created by integrating a backstepping control mechanism with an adaptive
Chebyshev neural network to track the output voltage of a buck DC-DC converter (CNN).
Because the loading circumstances are significantly variable, typical backstepping control action
is unreliable. As a result, an adaptable neural network is combined with Chebyshev's algorithm.

[5]
 Design Methodology
Polynomial Controller Design

It gets the reference voltage as input of the controller and u as the output of the controller, while
‘u’ is the input of the buck converter system and y is the output of the buck converter system.
More importantly, ‘yref’ serves as the feedback of the system and it will be compared with the
reference voltage. And by comparing, the controller will adjust to make the output voltage close
to the reference voltage. If the output voltage is higher than the reference voltage, then the
controller will make the switch open to discharge the circuit; while the output voltage is lower
than the reference voltage, the controller will close the switch to charge for the circuit. The
controller is a discrete transfer function, designed to get certain poles in the closed system.
As in the second-order discrete system, the transfer function is;

The transfer function can be expressed as a difference equation;

A controller is introduced as a general difference equation of the same order as the system’s
difference equation;

The controller can be given as a Z-transform;

As ;
The closed system is then given as;

When z=1, Kr can be designed to get the correct stationary gain;

This implies that the output signal Y stationary will be equal to the input value ‘Yref(z)’. For
polynomial controller design, look at the system first.
From the system, the transfer function is;

Then start the designation of the polynomial controller, which refers to determining the
polynomials C(z) and D(z).

From the system, when taking samples from the continuous-time model to get the discrete
model. Since the system is a second-order system. The transfer function H(z)=B(z)/A(z) have
second order, where polynomials B(z) and A(z) have second order in the converter system.
Design Pole Placement:
For pole placement P(z), where P(z)=A(z)C(z)+B(z)D(z), to keep the system stable, which
means P(z)=0 has its roots located inside the unit cycle
Get polynomial coefficients of C(z) and D(z):
When choosing the order of the C and D polynomial, it gives the closed system an arbitrary pole
placement. As the rules of polynomial controller given in Scmidtbaer, the polynomial C(z) and
D(z), where C(z) and D(z) satisfies

&

So polynomial C(z) and D(z) have the first order. And the order of polynomial P(z) which is;

It has the rule that the controller should be monic so that assume;

Then it has

And it also has

Where q1, q2, and q3 are the three chosen poles.

Identifying the coefficients for P(z):

By calculation, the solution is as below:

The pole placement is fixed and the sampling interval h=3e-6 is used by Ericsson now. When
placing the poles it has to judge that the control signal does not saturate outside 0-1 for a
reasonable change in load current.

Design Methodology
After obtaining the mathematical system model and polynomial controller model, the simulation
block building in Simulink of Matlab.

Building the system block in Simulink:

At first, building the system block with a voltage source and current source as input, and the
system output is the voltage of the load, at the same the inductor current can be observed

Build the controller block in Simulink:

When taking the output voltage as the input of the controller and comparing it with the reference
voltage, the output voltage is regarded as feedback. Then continue with the controller block.

The input of the controller is the output of the system, while the output of the controller will use
to compare with the PWM wave and then control the switch to adjust the charging and
discharging of the circuit.

Building the PWM block in Simulink:

When the controller output is less than the PWM value, the switch is closed, indicating that the
controller is charging, and when the controller output is more than the PWM value, the switch is
open, indicating that the controller is discharging.
Continuous-time Simulink model without saturation:
The model consists of the polynomial controller and a continuous system.

Discrete-time Simulink model without saturation:

The model consists of the polynomial controller and a discrete system.
Discrete-time Simulink model with saturation:

PWM Simulink model with controller:

The model consists of the polynomial controller and PWM switched system.
Matlab Code
The code demonstrates how to simulate the models:
It depicts parameter setup, system construction, controller design, and simulation result graphing.
 Results and Conclusion
After finishing the model in Simulink, run the simulation and analyze the results. The simulation
time is 100 sample intervals, which is sufficient to view the result. The inductor current exceeds
the output voltage and controller output by a factor of ten. To make the result more visible, the
inductor current is reduced by ten times.
Simulation result for discrete model:

As can be seen, the discrete system's output voltage may fulfill the needed voltage. The
controller output may become and maintain steady rapidly (in 16 sample intervals). Furthermore,
the controller causes the output voltage to stabilize fast (in 16 sample intervals) and to remain at
the desired output value (reference voltage 3.3V). The simulation is based on the polynomial
controller theory, hence it is a numerical demonstration of the theory. When building the
controller, keep in mind that it should not exceed the saturation limits for an acceptable
disturbance. And the simulation outcome meets it.
Simulation result for continuous model:

Because the reference voltage is 3.3V, the output voltage can very easily achieve it. Furthermore,
the output voltage remains constant at 3.3V. It demonstrates that the continuous model with a
polynomial controller works well. The controller output may become and maintain steady
rapidly (in 48 s). Furthermore, the controller causes the output voltage to stabilize fast (about 48
s) and to maintain the desired output voltage (reference voltage 3.3V). It is quite similar to
polynomial controller theory.
Simulation result without disturbance and noise:
Simulation result without disturbance and with noise:
In practice, the circuit contains some noise, such as noise at the switch due to high-frequency
switching, measurement noise, and so on.
To imitate that, I've added some band-limited white noise. The noise level ranges between -0.2 to
0.2, with a frequency of 33MHz. The frequency is an approximation based on a measurement.
And the measures are introduced by switching, yet they are used all the time.

Simulation result with disturbance and without noise:

The current source is utilized as the load in the model, and when the output current varies
abruptly, there is some disruption in the output voltage.
To examine the disturbance and determine if the polynomial controller is capable of restoring the
output voltage to the reference value. The current change range is set at 5A-15A.
Simulation result with disturbance and with noise:
 Conclusions
Following an examination of the simulation findings shown above, the following conclusions
may be drawn:
 The controller is capable of maintaining the output voltage at the reference voltage level.
 Because it has a smaller voltage drop with current disturbance, a bigger capacitor has a
higher quality to deal with the disturbance. In the closed system, the greater capacitor
implies a lesser bandwidth.
 For a greater capacitor load, a controller design that works for a small capacitor load
might result in a non-stable system.
 It is feasible to employ pole placement as a design strategy in a discrete polynomial
controller.
 References
[1]https://ieeexplore.ieee.org/stamp/stamp.jsp?
arnumber=7166010&casa_token=pJz88aHVlgoAAAAA:Tv2g4MbtI7kuPIPPLX6S5KS-
bXf0qjlcTaKdxkEsb_PejFPFTEa-4ON2lZU6f_e-PHnApij7qbNMzA&tag=1

[2] https://www.diva-portal.org/smash/get/diva2:829885/FULLTEXT01.pdf
[3]https://www.academia.edu/25546738/
Polynomial_Control_Method_of_DC_DC_Converters_for_DC_Bus_Voltage_and_Currents_Ma
nagement_and_x2014_Battery_and_Supercapacitors
[4] https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=9160875
[5] https://sci-hub.ee/10.1016/j.jfranklin.2016.04.008