Hidayat, Ghozi Athallah Ridho (2026) Rancang Bangun Sistem Artificial Intelligence Berbasis Knowledge-Graph dan Large-Language-Model untuk Mendesain Cylindro-Conical-Mixer serta Sistem Instrumentasi dan Kontrol dengan Pendekatan Tekno-Ekonomi. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5009221005_Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5009221005_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (14MB) | Request a copy |
Abstract
Kompleksitas perancangan peralatan proses industri yang masih bergantung pada pengalaman subjektif dan iterasi manual berisiko menimbulkan kesalahan serta inefisiensi. Penelitian ini mengembangkan sistem artificial intelligence (AI) hibrida untuk mendukung perancangan cylindro-conical mixer secara terstruktur dan tervalidasi. Sistem mengintegrasikan Knowledge Graph (KG) dan Large Language Model (LLM) melalui pendekatan Retrieval-Augmented Generation (RAG), dipadukan modul perhitungan numerik berbasis Python dan optimisasi Genetic Algorithm (GA). KG merepresentasikan pengetahuan teknis terstruktur, sedangkan LLM menangani interpretasi bahasa alami dan generasi penjelasan. Sistem menghasilkan rekomendasi parameter desain, perhitungan sizing, neraca massa-energi, serta visualisasi flowsheet secara otomatis. Validasi dilakukan secara multidimensi, meliputi perbandingan perhitungan sizing dan neraca massa-energi dengan simulasi Aspen HYSYS dan dokumen rancang pabrik sejenis, validasi optimasi GA terhadap MATLAB, evaluasi jawaban kualitatif menggunakan framework RAGAS (Retrieval-Augmented Generation Assessment), serta pengukuran precision-recall-F1-score terhadap struktur flowsheet yang dihasilkan. Hasil menunjukkan deviasi kuantitatif keseluruhan di bawah 1%, skor RAGAS di atas 90% pada ketiga metrik, dan F1-score sempurna untuk kesesuaian struktur flowsheet. Penggunaan KG juga terbukti meningkatkan spesifisitas dan keterlacakan penjelasan sistem dibandingkan tanpa basis pengetahuan terstruktur. Sistem AI ini dapat menjadi alat bantu perancangan yang akurat dan efisien, sekaligus berkontribusi pada SDG 9 (Industry, Innovation and Infrastructure) melalui inovasi rekayasa berbasis AI, serta SDG 7 (Affordable and Clean Energy) dan SDG 12 (Responsible Consumption and Production) melalui efisiensi energi dan sumber daya hasil optimisasi
======================================================================================================================================
The complexity of industrial process equipment design, which still relies on subjective experience and time-consuming manual iteration, risks introducing errors and inefficiency. This research develops a hybrid artificial intelligence (AI) system to support the structured and validated design of a cylindro-conical mixer. The system integrates Knowledge Graph (KG) and Large Language Model (LLM) through a Retrieval-Augmented Generation (RAG) approach, combined with a Python-based numerical calculation module and Genetic Algorithm (GA) optimization. The KG represents structured technical knowledge, while the LLM handles natural language interpretation and explanation generation. The system generates design parameter recommendations, sizing calculations, mass-energy balance analysis, and flowsheet visualization automatically. Validation spans multiple dimensions: comparing sizing and mass-energy balance results with Aspen HYSYS simulation and a reference plant design, validating GA optimization against MATLAB, RAGAS (Retrieval-Augmented Generation Assessment) based qualitative answer evaluation, and F1-score assessment of the generated flowsheet structure. Results show overall quantitative deviations below 1%, RAGAS scores above 90% across all three metrics, and a perfect F1-score for flowsheet structural conformity. The use of KG also improves the specificity and traceability of system explanations compared to approaches without structured knowledge grounding. This AI system can serve as an accurate and efficient design assistant, while contributing to SDG 9 (Industry, Innovation and Infrastructure) through AI-driven engineering innovation, as well as SDG 7 (Affordable and Clean Energy) and SDG 12 (Responsible Consumption and Production) through the energy and resource efficiency achieved via optimization
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Knowledge Graph, Large Language Model, Retrieval-Augmented Generation |
| Subjects: | T Technology > TJ Mechanical engineering and machinery > TJ212 Control engineering systems. Automatic machinery (General) |
| Divisions: | Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis |
| Depositing User: | Ghozi Athallah Ridho Hidayat |
| Date Deposited: | 01 Aug 2026 02:46 |
| Last Modified: | 01 Aug 2026 02:46 |
| URI: | http://repository.its.ac.id/id/eprint/141150 |
Actions (login required)
 |
View Item |