Page 98 - IJEEE-2022-Vol18-ISSUE-1

94 |
the kinetic energy when decelerating or braking, as it works
to improve fuel consumption and the vehicle’s dynamic
response by determining the operating points of the vehicle's
energy converters (engine, motor, and generator)[7]. In this
paper, the NMPC control strategy was used to manage the
energy in a series-parallel hybrid electric vehicle. This
strategy predicts the future behavior of the vehicle during a
period called the prediction horizon. Therefore, it needs a
mathematical model that describes the operation of the
vehicle, and it also needs to create a cost function that
expresses the control objectives of this strategy. To find the
optimum value of the control inputs for a hybrid electric
vehicle, this is done by solving its optimization problem,
which is represented by the equation of cost subject to the
mathematical prediction model and the physical constraints
of the vehicle, which is solved by one of the methods of
solving mathematical optimization problems. In this study, a
nonlinear MPC controller provided by MATLAB was used
to build this strategy, where this strategy aims to optimize the
vehicle's fuel consumption and make the vehicle achieve the
desired speed by the driver, as well as maintain the state of
charge of the batteries at the desired value. Where this study
can be used to learn how to build the mathematical model of
the series-parallel hybrid electric vehicle as well as how to
use the NMPC control strategy to manage this vehicle.
In [8] model predictive control (MPC) strategy was
applied to a series-parallel HEV where was used linear
mathematical prediction model to express the behavior of the
series-parallel HEV which is used by this strategy to predict
the future vehicle's behavior. The cost function used consists
of three terms. The first term is the square of the difference
between the actual and predicted torques of the wheels, and
the predicted torque of the wheels is estimated using an
adaptive recursive prediction algorithm that depends on the
past and present torque of the drivetrain. The second term
represents the square of the engine's fuel flow rate, and the
last term represents the square of the batteries' equivalent
fuel consumption. The cost function is solved by a Linear
Quadrature Tracking (LQT) approach to obtain the control
inputs values that achieve a minimum fuel consumption,
minimizing the difference between the actual and predicted
torque output of the drivetrain as well as maintaining the
battery charge condition at the required level. In [9] is used
model predictive control (MPC) strategy to manage the first
level of the series-parallel hybrid electric vehicle's control
unit. Since this unit consists of two levels, the first level finds
the optimal values for both the speed and torque of the
engine, which are references for the second control level.
The standard linear MPC is used to solve the optimization
problem in each sampling step by Quadratic Program (QP)
approach to find the optimum values of both the speed and
torque of the engine to get the minimum fuel consumption,
reduce using the friction brake, and keep the state of the
charge of the battery at the required level. In [10]the model
predictive control (MPC) strategy was also applied to
manage a control unit of the series-parallel HEV. Where the
standard linear MPC was used to solve the optimization
problem in each sampling step by MATLAB MPC toolbox
to find the optimal values torque of the engine, motor, and
generator in order to obtain the minimum fuel consumption,
Al-Flehawee & Al-Mayyahi
as well as maintain the state of the charge of the battery at
the required level.
This paper has been organized in the following manner.
Section II describes the architecture of the series-parallel
HEV model used in this paper. Section III illustrates the
dynamic equations for the powertrain and vehicle dynamic
equations and the dynamic equations for the state of charge
of the batteries in addition to describing how to express the
rate of fuel flow equation in terms of both the speed and
torque of the internal combustion engine. While section IV
explains the MPC algorithm in details, and section V deals
with the NMPC control strategy for a series-parallel HEV
and how to formulate the optimization problem of a series-
parallel HEV to achieve the objectives of the NMPC control
strategy in addition that it also contains how to create the
non-linear MPC controller block. Section VI explains
Sequential Quadratic Programming (SQP) Algorithm.
Section VI demonstrates the results of the simulation of the
series-parallel HEV model, where its unit was based on the
NMPC control strategy, while the last section discussed the
results of the simulation and the future work of the study.
II. ARCHITECTURE OF THE SERIES-PARALLEL HEV
MODEL
Building a model of a series-parallel hybrid electric
vehicle to study its controller that was built based on a model
predictive control. The approach used in this model is
classified as a forward-looking approach, as the control unit
depends on the required speed of the vehicle and the current
speed of the vehicle in creating commands to produce
torques through the drivetrain (the engine, motor, and
generator), to obtain the required vehicle traction force[11].
This model consists of three parts, as shown in Fig.2.
Fig.2: Architecture of the series-parallel HEV model
These parts are:-
? References block: - This block contains the driving cycle
and the desired values or the target value for the state of
charge (SOC) of the battery.
? Control unit: - The control unit was built depending on
the NMPC control strategy. The driving cycle and the
target state of charge of the battery are represented some
of the control unit entries which come from the reference
block. There is also feedback that only includes the
manipulated variables (MVs) that were computed in the
previous sampling step. There is also feedback that only
includes the manipulated inputs that are computed for the
current time step. In addition, the rest entries are the state
variables, which come from the powertrain and vehicle
dynamic block. While the engine torque, motor torque,
and generator torque are the outputs of this block and go
to the block of powertrain and vehicle dynamic.