Perancangan Sistem Pengendalian Dissolved Oxygen (DO) Pada Polishing Unit Berbasis Logic Solver Di PT. Saka Indonesia Pangkah Limited (SIPL)

Pardosi, Rico Pardona (2020) Perancangan Sistem Pengendalian Dissolved Oxygen (DO) Pada Polishing Unit Berbasis Logic Solver Di PT. Saka Indonesia Pangkah Limited (SIPL). Undergraduate thesis, Institut Teknologi Sepuluh Nopember.

[img] Text
02311640000060-Undergraduate_Thesis.pdf - Accepted Version
Restricted to Repository staff only

Download (5MB) | Request a copy

Abstract

Telah dilakukan perancangan sistem pengendalian Dissolved Oxygen (DO) pada Polishing Unit berbasis Logic Solver di PT Saka Indonesia Pangkah Limited (SIPL). Perancangan ini dilatarbelakangi oleh pemberian nutrisi pupuk urea dan Triple Super Phosphate (TSP) untuk mikroorganisme masih dilakukan secara manual oleh operator di lapangan. Sehingga melalui penelitian ini diharapkan mampu menciptakan simulasi pengendalian otomatis berdasarkan nilai DO sebagai varibel proses serta Urea dan TSP sebagai variabel yang dimanipulasi menggunakan pengendali logic solver. Penelitian ini terdiri dari dua bagian yaitu perancangan model dinamik plant (polishing unit) menggunakan Jaringan Syaraf Tiruan (JST) tipe Backpropagation dan perancangan simulasi sistem pengendalian loop tertutup pada simulink yang terdiri dari (logic solver, control valve, dan JST Polishing Unit). JST polishing unit dilatih lalu didapatkan model terbaik berstruktur 4-24-3 dengan 4 masukan (COD influent, OIW influent, Urea, dan TSP), 24 node lapisan tersembunyi dan 3 keluaran (OIW effluent, COD effluent, DO). Nilai MSE dan RMSE dari JST yang dilatih adalah 0.00485 dan 0.06964 yang diperoleh pada iterasi ke-2. Dari hasil simulasi pada simulink dengan memberi sejumlah skenario pada tabel kondisi logic solver maka diperoleh aksi berupa pemberian nutrisi urea dan TSP yang dikeluarkan oleh control valve. Nilai yang digunakan untuk mencapai set point DO (2 mg/L) antara lain: saat COD dan OIW influent melebihi baku mutu, COD melebihi baku mutu dan OIW tidak melebihi baku mutu serta eror DO dibawah nol. Dari hasil simulasi tersebut, nilai variabel proses (DO) dapat dikontrol mendekati set point dan dihasilkan respon sistem pengendalian yang stabil. ================================================================================================================== A Dissolved Oxygen (DO) control system has been designed based on Logic Solver On Polishing Unit at PT Saka Indonesia Pangkah Limited (SIPL). This design is motivated by the provision of urea fertilizer and Triple Super Phosphate (TSP) for microorganisms still done manually by operators in the field. So, through this research it is expected to be able to create automatic control simulations based on DO values as process variables and Urea and TSP as variables that are manipulated using a logic solver controller. This study consists of two parts, namely the design of a dynamic plant model (polishing unit) using Backpropagation Neural Network (ANN) and the simulation design of a closed loop control system on a simulink consisting of (logic solver, control valve, and ANN Polishing Unit). ANN polishing unit was trained and then obtained the best model with structure 4-24-3 with 4 inputs (COD influent, OIW influent, Urea, and TSP), 24 hidden layer nodes and 3 outputs (OIW effluent, COD effluent, DO). The MSE and RMSE values of the ANN trained were 0.00485 and 0.06964 obtained at the 2nd iteration. From the simulation results on simulink by giving a number of scenarios in the logic solver condition table, the action is obtained in the form of urea and TSP nutrition issued by the control valve. The values used to reach the DO set point (2 mg /L) include: when COD and OIW influent exceeds quality standards, COD exceeds quality standards and OIW does not exceed quality standards and DO errors below zero. From the results of the simulation, the value of the process variable (DO) can be controlled near the set point and a stable control system response is generated.

Item Type: Thesis (Undergraduate)
Uncontrolled Keywords: Jaringan Syaraf Tiruan, Dissolved Oxygen, Logic Solver. Artificial Neural Network, Dissolved Oxygen, Logic Solver.
Subjects: T Technology > TJ Mechanical engineering and machinery > TJ213 Automatic control.
Divisions: Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis
Depositing User: Rico Pardona Pardosi
Date Deposited: 12 Aug 2020 05:39
Last Modified: 12 Aug 2020 05:39
URI: https://repository.its.ac.id/id/eprint/77658

Actions (login required)

View Item View Item