Optimal Contribution Selection Pada Pemuliaan Tanaman Menggunakan Metode Single Cut Proximal Bundle (SCPB)

Umama, Rahmalia (2026) Optimal Contribution Selection Pada Pemuliaan Tanaman Menggunakan Metode Single Cut Proximal Bundle (SCPB). Other thesis, Institut Teknologi Sepuluh Nopember.

Text 5002221105-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (3MB) | Request a copy

Abstract

Pemuliaan tanaman bertujuan memaksimalkan genetic gain sekaligus mengendalikan tingkat inbreeding dalam populasi. Nilai genetik individu merupakan hasil estimasi
berbasis data fenotipe yang dipengaruhi variasi lingkungan sehingga mengandung ketidakpastian. Sebagian besar pendekatan yang ada masih mengabaikan ketidakpastian
tersebut dengan memodelkan nilai genetik secara deterministik. Penelitian ini merumuskan model optimasi pemuliaan tanaman yang mempertimbangkan ketidakpastian nilai genetik melalui pendekatan Stochastic Convex Composite
Optimization (SCCO) dan menerapkan metode Single Cut Proximal Bundle (SCPB) untuk menyelesaikannya. Data yang digunakan berupa 1.385 line tanaman kapas dari dataset CSIRO dengan 11.856 SNP setelah dilakukan filtering.
Ketidakpastian nilai genetik direpresentasikan melalui distribusi posterior breeding value yang diturunkan dari model GBLUP dengan estimasi varians menggunakan REML. Fungsi objektif dirumuskan sebagai minimasi ekspektasi negatif genetic gain dikombinasikan dengan komponen expected improvement, sedangkan kendala seleksi direpresentasikan sebagai fungsi indikator konveks. Hasil simulasi menunjukkan bahwa metode SCPB mampu menghasilkan solusi kontribusi individu yang memenuhi seluruh kendala yang ditetapkan, kecuali pada kondisi tidak feasible. Misalnya pada batas
bawah yang terlalu besar (lb = 0,001) dan batas inbreeding yang dibawah baseline inbreeding, yaitu pada beta = 0,05. Jumlah siklus yang semakin besar meningkatkan kestabilan solusi, tetapi berdampak pada waktu komputasi yang lebih lama. Nilai β memiliki pengaruh paling signifikan karena menentukan trade-off antara genetic gain dan tingkat inbreeding sehingga pemilihannya perlu mempertimbangkan keseimbangan antara peningkatan kualitas genetik dan batas inbreeding yang dapat ditoleransi dalam program pemuliaan.
================================================================================================================================
Indonesia is committed to achieving Net Zero Emissions (NZE) by 2060, requiring the development of low-carbon energy sources, one of which is green hydrogen production as an alternative to conventional (grey) hydrogen in the industrial sector. This study aims to design a green hydrogen plant based on water electrolysis with a production capacity of 26,000 tons/year in Palembang, South Sumatra. The production technology was selected using the Analytical Hierarchy Process (AHP) method, with Alkaline Water Electrolysis (AWE) chosen for its high technology readiness level, low investment cost, and long stack lifetime. Raw water is sourced from the Musi River and treated through a series of water treatment processes, fine screening, coagulation-flocculation, ultrafiltration, reverse osmosis, and ion exchange to meet ASTM Type II standards for electrolyzer feedwater. The hydrogen product achieves a purity of $\ge$99.95%, complying with CGA G-5.3-2004 Type I-B standards, with oxygen produced as a commercially valuable byproduct. Based on the design results, the electrolysis electrical energy requirement is 167.38 MW, with a total plant electricity demand of 177.33 MW. Economic evaluation shows a total capital investment (TCI) of IDR 9.73 trillion, comprising a fixed capital investment (FCI) of IDR 8.76 trillion and a working capital investment (WCI) of IDR 973.36 billion. Feasibility analysis yields an NPV of IDR 374 billion, an IRR of 8.9%, a Pay Out Time (POT) of $\pm$9.1 years, a Break Even Point (BEP) of 36.36%, with hydrogen selling price identified as the variable most influential on IRR. Based on technical, economic, and environmental aspects, this green hydrogen plant design is deemed feasible for construction.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Pemuliaan tanaman, estimated breeding value, optimal contribution selection, stochastic convex composite optimization, Single Cut Proximal Bundle Plant breeding, estimated breeding value, optimal contribution selection, stochastic convex composite optimization, Single Cut Proximal Bundle
Subjects:	Q Science > QA Mathematics > QA273.5 Stochastic geometry Q Science > QA Mathematics > QA402.5 Genetic algorithms. Interior-point methods.
Divisions:	Faculty of Science and Data Analytics (SCIENTICS) > Mathematics > 44201-(S1) Undergraduate Thesis
Depositing User:	Rahmalia Umama
Date Deposited:	30 Jul 2026 07:28
Last Modified:	30 Jul 2026 07:28
URI:	http://repository.its.ac.id/id/eprint/139937

Actions (login required)

View Item