Interlock Optimization Of An Accelerator Using Genetic Algorithm
Journal Title: International Scientific and Vocational Studies Journal - Year 2017, Vol 1, Issue 1
Abstract
Accelerators are systems where high-tech experiments are conducted today and contain high-tech constructions. Construction and operation of accelerators require multidisciplinary studies. Each accelerator structure has its own characteristics as well as similar features of accelerator structures. Control systems come to the forefront as one of the most important structures that make up accelerators. Since control systems have critical importance for accelerators, in such systems when a problem occurs, there is a danger of environmental and human safety as well as machine system. For that reason interlock systems are being developed in different structures. In the literature, FPGAs and PLCs in such interlock systems have been shown to be suitable for use in accelerators [1,7]. In this work, we describe an interlock system that evaluates the operation and protection modes of devices used in an electron accelerator. In order to ensure that this system can operate at minimum cost and maximum safety, the defined system is divided into 3 subsystems. The error messages from the control devices in the accelerator control systems are the input to the interlock system. The purpose of the interlock system that evaluates error messages is to ensure that the accelerator closes safely. The purpose of this study is to specify which of the 3 interlock subsystems which are defined for minimum cost and maximum security should be connected to the fault outputs from the control devices. As an evaluation criterion, 6 features are defined for the control devices and each control device is weighted according to the importance of the task. In the solution of the problem, genetic algorithms were used for assigning 74 controller outputs to 3 interlock subsystems. Thanks to the Genetic Algorithm used in the study, 94.3% success rate was obtained in terms of cost and safe system.
Authors and Affiliations
İbrahim Burak Koç, Anas Al Janadi, Volkan Ateş
Keywords
Related Articles
- EP ID EP35716
- DOI -
- Views 358
- Downloads 0
Complementary Coded Waveforms Sets in Marine Radar Application
Complementary coded waveform and mismatched filter pairs sets are used. On the contrary with Golays matched waveform filter pair the mismatched waveform filter pair does exist for all N (number pulses in waveform). Using...
The Mesozoic Stratigraphy and Ammonite Faune for Niksar-Erbaa (Tokat) Territory
The basement of the study territory covering the Niksar-Erbaa territories and their immediate vicinity, located on the north and above the North Anatolian Fault, forms the rocks belonging to Permo-Triassic aged Karakaya...
A Comparison of Five Methods for Missing Value Imputation in Data Sets
The missing values in the data sets do not allow for accurate analysis. Therefore, the correct imputation of missing values has become the focus of attention of researchers in recent years. This paper focuses on a compar...
Forecasting of Electricity Generation Shares by Fossil Fuels Using Artificial Neural Network and Regression Analysis in Turkey
This study is conducted to get predictions for the generation of electricity by annual production shares and decide the most suitable method for future periods. Between 2010-2017 in Turkey, the relation of generation sha...
Thermal Stress Control in Functionally Graded Plates with Artificial Neural Network
In this study, trained models were obtained by using Artificial Neural Network (ANN) in order to determine the equivalent stress levels of one dimensional functionally graded rectangular plates. In this training set, a...