Thoyyibi, Muthiutthoriq (2026) Rancang Bangun Sistem AI Berbasi Knowledge Graph-RAG Untuk Desain High Pressure Feedwater Heater Dan Sistem Instrumentasi-Kontrol Dengan Pendekatan Teknoekonomi. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 5009221020_Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
5009221020_Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (7MB) | Request a copy |
Abstract
Peningkatan efisiensi termal PLTU merupakan jalur langsung menuju SDG 7 (energi bersih dan terjangkau) dan SDG 9 (industri, inovasi, dan infrastruktur), dan high pressure feedwater heater (HPH) menjadi salah satu komponen penentunya. Perancangan HPH tipe shell-and-tube heat exchanger (STHE) masih banyak dikerjakan secara manual dan terfragmentasi, sehingga rawan inkonsistensi antara perhitungan termal, keputusan optimisasi biaya, dan rekomendasi sistem kontrol. Penelitian ini mengembangkan asisten AI terintegrasi yang menggabungkan Large Language Model (LLM), Knowledge graph (KG), physics engine Bell-Delaware multi-zona, optimisasi metaheuristik, dan generasi P&ID otomatis dalam satu alur perancangan, dengan studi kasus HPH-3 PLTU Paiton Unit 9. Basis pengetahuan dibangun dari 165 dokumen di Neo4j dan ChromaDB, diakses melalui arsitektur Graph-RAG. Perhitungan termal memisahkan zona desuperheating, condensing, dan subcooling melalui pendekatan diskritisasi enam puluh node, sedangkan optimisasi geometri memakai PSO, GA, dan DE dengan vektor keputusan
==================================================================================================================================
Improving power plant thermal efficiency is a direct pathway toward SDG 7 (affordable and clean energy) and SDG 9 (industry, innovation, and infrastructure), and the high pressure feedwater heater (HPH) is one of its decisive components. The design of shell-and-tube heat exchanger (STHE) type HPH units is still largely manual and fragmented, making it prone to inconsistencies among thermal calculations, cost optimization decisions, and control system recommendations. This research develops an integrated AI assistant combining a Large Language Model (LLM), a Knowledge graph (KG), a multi-zone Bell-Delaware physics engine, metaheuristic optimization, and automated P&ID generation within a single design pipeline, using HPH-3 of Paiton Power Plant Unit 9 as the case study. The knowledge base was built from 165 documents in Neo4j and ChromaDB, accessed through a Graph-RAG architecture. Thermal calculations separate the desuperheating, condensing, and subcooling zones via a sixty-node Marching approach, while geometry optimization uses PSO, GA, and DE with decision vector
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Bell-Delaware, high pressure feedwater heater, knowledge graph, large language model, shell-and-tube heat exchanger. |
| 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: | Muthiutthoriq Thoyyibi |
| Date Deposited: | 01 Aug 2026 04:51 |
| Last Modified: | 01 Aug 2026 04:51 |
| URI: | http://repository.its.ac.id/id/eprint/141690 |
Actions (login required)
 |
View Item |