UNS Conference Portal, The 1st International Conference on Mathematics: Education, Theory & Application

Font Size: 
Identification and Estimation of Variables on Reduced Model Using Balanced Truncation Method
Trifena Punana Lesnussa, Didik Khusnul Arif, Dieky Adzkiya

Last modified: 2016-10-31

Abstract


In this research we will discuss the identification of variables and variable estimation on a reduction model process. Identification of  variables aims to trace relationship between variables on reduced and original system. Estimation of variables aims to estimate state variables at a system that can not be measured directly. Generally a system in nature has a great order. It caused difficulties in analyzing the system. Therefore we need a method to simplify a system of large order into a system with smaller order such that it has a dynamic behavior that is equal or almost equal to the original model. First the model is reduced by using balanced truncation method. The method begins with the construction of balanced system. Next, we identify the variables and then we cut variables that have small influence on the system. Thus we obtain a reduced model. Estimation process is performed on the original and reduced system using a Kalman Fiter algorithm. The expected result of this research is to obtain the relationship between the variables at the original and reduced system, so that we can compare the accuracy of estimation on the original and reduced model.