Simulasi Permasalahan Multiobyektif Berbasis Agen pada Kasus Economic and Emission Dispatch (EED) dengan Metode Neuro-Fuzzy System di Power Plant

Nugroho, Fressy (2010) Simulasi Permasalahan Multiobyektif Berbasis Agen pada Kasus Economic and Emission Dispatch (EED) dengan Metode Neuro-Fuzzy System di Power Plant. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 2208205708-Master_thesis.pdf Restricted to Repository staff only Download (38MB)

Abstract

Manajemen operasi listrik telah banyak diteliti, salah satunya adalah permasalahan Economic and Emission Dispatch (EED). Permasalahan EED yang timbul adalah optimalisasi sistem pembangkit yang dipengaruhi stabilitas dan kualitas daya sistem pembangkit serta bertujuan meminimisasi emisi sistem pembangkit dan kerugian daya total dalam jaringan pembangkit. Sehingga didapat tiga parameter yang akan dicari pareto optimalnya, yaitu parameter biaya, emisi, dan rugi daya transmisi. Metode ANFIS digunakan karena tidak membutuhkan waktu yang lama untuk menghasilkan keluaran sesuai target dan bersifat adaptif. Target yang hendak dicapai ANFIS merupakan optimalisasi permasalahan EED menggunakan metode Simulated Annealing. Karena tiap ANFIS hanya mempunyai satu keluaran, maka digunakan multiple ANFIS. Berdasarkan penelitian ini, diketahui bahwa penggunaan metode Multiple ANFIS lebih baik dibandingkan metode Simulated Annealing, memberikan total delta biaya bahan bakar sebesar 365.126.452,6 rupiah/MWjam atau 0,0472%, sedangkan total delta emisi sebesar 2.993,879633 ton/MWjam atau 0,023767131% dan total delta kerugian daya transmisi sebesar 6,415025467 MW/jam atau 0,029218795%. Hasil yang diperoleh ini kemudian digunakan untuk membentuk diagram pareto 3 dimensi yang menggambarkan trade-off dari tiga parameter tersebut.
===================================================================================================================================
Electrical operations management has been widely studied, one of which is the problem of Economic and Emission Dispatch (EED). The problem arising in EED is optimizing the power system influenced by power system stability and power quality, aiming to minimize power system emissions and total power losses in power networks. Thus, three parameters are obtained to search for Pareto optimality: cost, emission, and transmission power loss. The ANFIS method is used because it does not take long to produce the targeted output and is adaptive. The target to be achieved by ANFIS is the optimization of the EED problem using the Simulated Annealing method. Since each ANFIS has only one output, multiple ANFIS is used in this research. Based on this research, it is found that using the multiple ANFIS method is better than the Simulated Annealing method, yielding a total fuel cost delta of 365,126,452.6 IDR/MWh or 0.0472%, a total emission delta of 2,993.879633 tons/MWh or 0.023767131%, and a total transmission power loss delta of 6.415025467 MW/h or 0.029218795%. These results are then used to construct a 3-dimensional Pareto diagram illustrating the trade-off among these three parameters.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	multiobyektif, EED, Neuro Fuzzy System, Power Plant. , multiobjective problems, EED, Neuro Fuzzy System, Power Plant.
Subjects:	Q Science > QA Mathematics > QA269 Game theory
Divisions:	Faculty of Electrical Technology > Electrical Engineering > 20101-(S2) Master Thesis
Depositing User:	magang .
Date Deposited:	28 Sep 2026 01:36
Last Modified:	28 Sep 2026 01:36
URI:	http://repository.its.ac.id/id/eprint/144921

Actions (login required)

View Item