Đồ án tốt nghiệp - Application research Fuzzy-pid control on CVT based on PSO algorithm to OTSR

HO CHI MINH CITY UNIVERSITY OF TECHNOLOGY AND EDUCATION
FACULTY OF INTERNATIONAL EDUCATION

GRADUATION PROJECT

APPLICATION RESEARCH FUZZY-PID CONTROL
ON CVT BASED ON PSO ALGORITHM
TO OPTIMIZE THE SPEED RATIO

TRỊNH XUÂN MẠNH

Student ID: 21145027

LÊ CÔNG TUẤN

Student ID: 21145619

BÙI XUÂN BÁCH

Student ID: 21145577

Major: AUTOMOTIVE ENGINEERING
Supervisor: DƯƠNG TUẤN TÙNG, PhD.
Ho Chi Minh City, June 2025

HO CHI MINH CITY UNIVERSITY OF TECHNOLOGY AND EDUCATION
FACULTY OF INTERNATIONAL EDUCATION

GRADUATION PROJECT

APPLICATION RESEARCH FUZZY-PID CONTROL
ON CVT BASED ON PSO ALGORITHM
TO OPTIMIZE THE SPEED RATIO

TRỊNH XUÂN MẠNH

Student ID: 21145027

LÊ CÔNG TUẤN

Student ID: 21145619

BÙI XUÂN BÁCH

Student ID: 21145577

Major: AUTOMOTIVE ENGINEERING
Supervisor: DƯƠNG TUẤN TÙNG, PhD.
Ho Chi Minh City, June 2025

THE SOCIALIST REPUBLIC OF VIETNAM
Independence – Freedom – Happiness


Ho Chi Minh City, June 09, 2025

GRADUATION PROJECT ASSIGNMENT
Student name: TRỊNH XUÂN MẠNH

Student ID: 21145027

Student name: LÊ CÔNG TUẤN

Student ID: 21145619

Student name: BÙI XUÂN BÁCH

Student ID: 21145577

Major: AUTOMOTIVE ENGINEERING

Class: 21145FIE

Supervisor: DƯƠNG TUẤN TÙNG, Ph.D

Phone number: (+84) 914 805 623

Date of assignment: 25/01/2025

Date of submission: 10/06/2025

1. Project title: APPLICATION RESEARCH FUZZY-PID CONTROL ON CVT BASED ON
PSO ALGORITHM TO OPTIMIZE THE SPEED RATIO.
2. Initial materials provided by the supervisor:
- Science journals, references related to the topic
- Modeling system in MATLAB/Simulink from the previous group studying PID control
3. Content of the project:
- Research the Fuzzy-PID control optimized by the PSO algorithm to enhance the speed
ratio of the CVT system in various operating conditions
- Build the modeling system in MATLAB/Simulink and run the cycle tests to evaluate the
control method and compare it with the traditional PID control
- Validate the simulation data by experiment at Bosch Transmission Lab
4. Final product:
- MATLAB/Simulink model of a CVT-equipped vehicle powertrain
- Fuzzy-PID controller optimized by PSO for speed ratio control
- Comparison of control performance (PID and PSO Fuzzy-PID) under SSS, UDDS, and
HWFET cycles
- Experimental validation using the Bosch Transmission Lab Test Bench
- Full report with theory, modeling, simulation-experiment data, and analysis
CHAIR OF THE PROGRAM

SUPERVISOR

(Sign with full name)

(Sign with full name)

THE SOCIALIST
...

...
tion factor for error):
,
num2str(optimized_params.Ke, %.4f)])
disp([Kec (Quantization factor for error rate): ,
num2str(optimized_params.Kec, %.4f)])
disp([Kup (Scaling factor for ΔKp):
,
num2str(optimized_params.Kup, %.4f)])
disp([Kui (Scaling factor for ΔKi):
,
num2str(optimized_params.Kui, %.4f)])
disp([Kud (Scaling factor for ΔKd):
,
num2str(optimized_params.Kud, %.4f)])
disp([Best ITAE achieved:
,
num2str(optimized_params.ITAE, %.6f)])
% Plot convergence curve
figure(Position, [100, 100, 600, 300])
plot(1:max_iter, convergence_curve, b-, LineWidth, 2)
xlabel(Iteration, FontSize, 11)
ylabel(ITAE Performance, FontSize, 11)
title(PSO Convergence Curve, FontSize, 12)
grid on
% Save results
save(optimized_FuzzyPID_params_math.mat,
optimized_params, convergence_curve)
end
function itae = calculate_ITAE(params)
% Mathematical model replaces Simulink
% CVT system simulation with Fuzzy-PID controller
% System parameters (can be adjusted according to actual
characteristics)
Ke = params(1)
Kec = params(2)
Kup = params(3)

163

Kui = params(4)
Kud = params(5)
% Simulation time
t = 0:0.01:10 % 10 giây với bước 0.01s
% Input signal (desired transmission ratio - step from 2.5
to 0.5 in 1s)
target_ratio = 2.5 * ones(size(t))
target_ratio(t = 1) = 0.5
% Initiating state variables
actual_ratio = zeros(size(t))
error = zeros(size(t))
error_integral = 0
error_prev = 0
% Simple CVT Model Parameters
tau = 0.5 % System Time Constant
K_sys = 1.0 % System Gain Factor
% System Simulation
for i = 2:length(t)
% Error calculation
error(i) = target_ratio(i) - actual_ratio(i-1)
% Fuzzy-PID controller (simplify)
% Calculation of PID components
P = Kup * Ke * error(i)
I = Kui * Ke * error_integral
D = Kud * Kec * (error(i) - error_prev)/0.01
% Control Signal Synthesis
u = P + I + D
% Control Signal Limiting
u = max(min(u, 10), -10)

% System model (first-order with noise)
164

actual_ratio(i) = actual_ratio(i-1) + 0.01 * (K_sys*u actual_ratio(i-1))/tau

% Add small random noise
if i length(t)/2
actual_ratio(i) = actual_ratio(i) + 0.01*randn()
end
% Error integral
error_integral = error_integral + error(i)*0.01
error_prev = error(i)
end
% Calculate ITAE
itae = trapz(t, t.*abs(error))
% Penalties for unstable cases
if any(isnan(actual_ratio)) || any(abs(actual_ratio)) 10
itae = 1e10
end
end

165

APPENDIX B
SIMULATION AND EXPERIMENTAL DATA TABLE – SSS CYCLE

166