Selamat Datang di Website Dosen

The data with free of noise or outliers will not be obtained in the chemical or physical process measurements using due to some kind of noise arising from thermodynamics and quantum effects may not be removed. The extensive use of personal computers in process instrumentation and flexibility of programming software, encourages its use for filtering time series data with satisfactory results. This paper will investigate outlier filtering techniques on time series data of temperature and flow rate of hydrogen gas as result of sensor measurement on the sintering process. The results an optimal parameters on the filtering technique that yields an adequate signal to noise ratio while still maintaining peak signal on the measurement results, Its is very important respect to process safety parameters of hydrogen gas.

Nonlinear system identification is becoming an important tool which can be used to improve control performance and achieve robust fault-tolerant behavior. Among the different nonlinear identification techniques, methods based on neural network model are gradually becoming established not only in the academia, but also in industrial application. An identification scheme of nonlinear systems for sintering furnace temperature in nuclear fuel fabrication using neural network autoregressive with exogenous inputs (NNARX) model investigated in this paper. The main contribution of this paper is to identify the appropriate model and structure to be applied in control temperature in the sintering process in nuclear fuel fabrication, that is, a nonlinear dynamical system. Satisfactory agreement between identified and experimental data is found with normalized sum square error 1.9𝑒 − 03 for heating step and 6.3859𝑒 − 08 for soaking step.That result shows the model successfully predict the evolution of the temperature in the furnace.

Cylindrical uranium dioxide pellets, which are the main components for nuclear fuel
elements in Light Water Reactor, should have a high density profile, uniform shape and quality for
the safety used as a reactor fuel component. The quality of green pellets is conventionally
monitored through a laboratory measurement of the physical pellets characteristics followed by a
graphical chart classification technique. However, this conventional classification method shows
some drawbacks, such as the difficulties on its usage, low accuracy and time consuming, and does
not have the ability to adress the non-linearity and the complexity of the relationship between the
pellet’s quality variables and the pellett’s quality. In this paper, an Improved Levenberg-Marquard
based neural networks is used to classify the quality process of the green pellets. Robustness of this
learning algorithm is evaluated by comparing its recognition rate to that of the conventional Back
Propagation neural learning algorithm. Results show that the Improved Levenberg-Marquard
algorithm outperformed the Back Propagation learning algorthm for various percentage of
training/testing paradigm, showing that this system could be applied effectively for classification of
pellet quality.