Optimasi Respon Ganda Pendekatan Fungsi Kerugian (Aplikasi pada Design Economizer PT. ALSTOM ESI)

Mualimin, Mualimin (2007) Optimasi Respon Ganda Pendekatan Fungsi Kerugian (Aplikasi pada Design Economizer PT. ALSTOM ESI). Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 1305201018-Master_Thesis.pdf Restricted to Repository staff only Download (6MB) | Request a copy

Abstract

Salah satu metode yang banyak digunakan untuk optimasi respon ganda adalah strategi pengurangan dimensionalitas dari respon-respon. Strategi ini mengkonversikan permasalahan respon ganda menjadi suatu ukuran agregat tunggal, dan menyelesaikannya menjadi masalah optimasi dengan objektif tunggal. Beberapa metode optimasi respon ganda telah dibahas oleh Khuri dan Conlon tetapi mereka tidak mempertimbangkan ekonomi proses, sedangkan Ames mengabaikan struktur korelasi antar respon-responnya. Permasalahan yang muncul adalah bagaimana mendapatkan nilai variabel input yang mengoptimalkan respon ganda secara simultan, dan bagaimana memperoleh informasi tentang sifat bias kecil, kekokohan tinggi dan hasil prediksi yang berkualitas tinggi. Tujuan yang akan dicapai adalah mendapatkan nilai variabel input yang mengoptimalkan respon ganda secara simultan, informasi tentang sifat bias kecil, kekokohan tinggi dan hasil prediksi yang berkualitas tinggi, membandingkan fungsi kerugian metode Young dengan metode Vining berdasarkan nilai Loss yang minimum. Berdasarkan tujuan tersebut maka hasil optimasi menggunakan fungsi kerugian adalah outside diameter tubing: 2 inch, transfersal spacing: 3,5 inch, dan number of fin: 5 fin/inch, nilai Loss metode Young: 0,7416, dan nilai Loss metode Vining:1,3148, nilai optimum thermal effisiensi:1, nilai optimum operating cost: 12,7174.
=======================================================================================================================================
One of the popular approach to multiresponse optimization is using dimensionality reduction strategy from the responses. This strategy converts multiresponse problem into single agregate measure, and solves it as a single objective optimization problem. Some other methods of multiresponse optimization have studied by Khuri and Conlon but they do not consider the proses economics, while Ames ignore the correlation structure among the responses. Problem which appear is how to get value input variable that can optimize multiresponse simultaneously and how to get information obout small bias, high robustness and high quality of prediction. The purposes that will be got are finding value of input variabel that can optimize multiresponse simultaneously, information about small bias, high robustness and high quality of prediction and comparing between loss function of Young’s method and Vining’s method based on minimum loss value. Based on that purposes, so optimization result by loss function approach are outside diameter tubing: 2 inch, transfersal spacing: 3.5 inch and number of fin: 5 fin/inch, loss value of Young’s method: 0.7416 and loss value of Vining’s method: 1.3148, optimum value of thermal efficiency: 1, optimum value of operating cost 12.7174.

Item Type:	Thesis (Masters)
Additional Information:	RTSt 519.538 Mua o 2007 3100008031291
Uncontrolled Keywords:	Loss functions, robustness, quality of prediction
Subjects:	Q Science > QA Mathematics > QA279 Response surfaces (Statistics). Analysis of covariance.
Divisions:	Faculty of Mathematics and Science > Statistics > 49101-(S2) Master Thesis
Depositing User:	Anis Wulandari
Date Deposited:	13 Jul 2026 02:42
Last Modified:	13 Jul 2026 02:42
URI:	http://repository.its.ac.id/id/eprint/134601

Actions (login required)

View Item