| Abstrak/Abstract |
Nuclear power plants are one of the main focuses
of national research in Indonesia for the 2020-2024 period,
which aims to develop the first nuclear power plant prototype.
As a step to support this goal, this paper introduces a model of
the NuScale reactor that can be combined with a control
prototype and application of an auto-tuning proportional-
integral-derivative (PID) control strategy that is applied to the
reactor power control system. Non-linear modeling is arranged
in a state-space model that describes the parameters of the
reactor primary loop system. State-space computing is
implemented in the Laboratory Virtual Instrumentation
Engineering Workbench (LabVIEW) software which can be
combined with attractive simulator devices according to the
technology in the NuScale reactor. PID auto-tuning control is
used as a power controller by adjusting the reactivity of 4 types
of control rods in the reactor core. PID control parameter is set
based on an algorithm that identifies the ultimate gain, ultimate
period, dead time, and time constant based on Ziegler and
Nichols’ heuristic methods. Results of verification with NuScale
reactor data that have obtained design approval from the United
State Nuclear Regulatory Commission (NRC) show error values
below 1% on main parameters such as power, neutron flux, fuel
temperature, coolant temperature, reactor pressure, and
coolant flow. Reactor control simulation is run using step signal
input and ramp signal in a load-following mode in the power
range of 90% full power to 100% full power. The result of
comparison with manual PID control shows that auto-tuning
PID shows better performance and stability in maintaining
reactor power in start-up and load-following operation. |
