Jurnal Teknik dan Sistem Komputer

Penerapan Genetic Algorithm (GA) untuk Optimasi Parameter Model Thevenin Baterai Lithium

Muhammad Aminuddin — 2026-01-23

Accurate battery modeling is essential for improving the performance of state estimation algorithms such as State of Charge (SoC) and State of Health (SoH). The first-order Thevenin model is widely used due to its simplicity and its ability to capture the dynamic voltage behavior of lithium-based batteries. The accuracy of this model strongly depends on the selection of the internal resistance ( ), polarization resistance ( ), and polarization capacitance ( ). However, these parameters are often assumed to be constant, causing the battery model to lose adaptability when the cell experiences voltage dynamics during operation. To address this limitation, this study proposes the use of a Genetic Algorithm (GA) to adaptively optimize the model parameters. The GA is designed to update the parameters when the initial values begin to produce increasing voltage estimation errors, ensuring that the parameters remain representative of the battery’s actual condition. Meanwhile, offline analysis and Recursive Least Squares (RLS) are employed as comparison and validation methods. Experimental results demonstrate that the parameters optimized using GA significantly improve terminal voltage estimation accuracy, achieving a minimum RMSE of 0.005 V, outperforming both the offline method (RMSE 0.0145 V) and RLS (RMSE 0.0143 V). These findings confirm that GA effectively generates dynamic and accurate parameters, enabling the Thevenin model to better represent the battery’s behavior under varying operating conditions.

Keywords: Thevenin Battery Model, Parameter Identification, Battery, Genetic Algorithm, Voltage Estimation

Rancang Bangun Sistem Pengisian Tangki Air Otomatis Berbasis PLC Siemens S7-200 Dengan Sensor Proximity Kapasitif

Jaenal Rizki — 2026-01-24

This study discusses the design of an automatic water tank filling system based on a Siemens S7-200 Programmable Logic Controller (PLC) using a capacitive proximity sensor as a water level detector. The main objective of this study is to automate the filling of water tanks to make it more efficient, reduce energy waste and reduce errors due to inaccurate manual filling. This system is designed by utilizing a capacitive proximity sensor positioned as a detector of water level changes. The digital signal from the sensor is then processed by the PLC to automatically control the water pump using a ladder diagram program as its logic. The PLC functions as a control core that intelligently controls the output, namely the water pump. The research method includes block diagram design, ladder diagram programming, hardware assembly, and system testing. The test results show that the system is able to work stably with an accuracy level of 98.13%, based on the method of measuring the water level at 8 cm. Thus, this system can be implemented as an effective solution for automatic control in small to medium-scale water distribution systems.

Keywords: PLC Siemens S7-200, capacitive sensor, automation, water tank filling, control system

Teknologi Kendali Otomatis Pemotong Pisang Untuk Produksi Keripik Berbasis Arduino

Irham Iswanda Duhair — 2026-01-27

Banana chip production by micro, small, and medium enterprises (MSMEs) generally relies on manual cutting, which is often inefficient, prone to hand injuries, and produces slices with inconsistent thickness. This research aims to design and develop an Arduino-based automatic banana cutting system to increase productivity, time efficiency, and standardize chip slices. The system is controlled by an Arduino Uno, which receives signals from two limit switches and a stepper motor as a driver. When the driver touches the limit switch, it reverses direction. Cutting results were tested on Kepok bananas, with a cutting thickness of 1-3 mm. Test results showed that this tool is capable of producing cuts with a thickness of 1-3 mm on Kepok bananas, without relying on the operator's hand strength. It provides an automatic mechanism that reduces the risk of hand injuries and direct contact with banana sap. Thus, the application of Arduino-based automatic control technology significantly improves the banana chip cutting process, improving work efficiency, more consistent results, and ensuring operator safety.

Keywords: Banana, Limit Switch, Stepper Motor, Cutting

Implementasi Docker Container Dalam Pembuatan Sistem Anggaran, Keuangan dan Akuntansi Pada Lingkungan Universitas

Pratyaksa Ocsa Nugraha Saian — 2026-01-31

Financial management and accounting are crucial aspects for higher education institutions that deal with complex administrative processes, including Satya Wacana Christian University in Salatiga. The financial application previously used at SWCU still had several limitations, such as difficulties in generating timely reports, duplicate transaction codes, and the inability to adapt to changes in policies or Standard Operating Procedures (SOPs). This study aims to develop a prototype of a financial and accounting management system based on microservices architecture utilizing Docker Container technology. The development method applied is Prototyping, which includes requirement analysis, system design, implementation, and evaluation of the prototype up to the third iteration. The results show that the adoption of microservices simplifies system management by dividing the application into eleven independent services. The implementation of Docker supports this architecture by providing separate containers for each service, thereby facilitating maintenance, development, and system scalability. The study concludes that a microservices-based approach using Docker can serve as an effective solution for building financial and accounting systems within universities.

Keywords: Docker, Microservices, Financial System, Accounting System

Analisis K-Means Clustering pada Pendidikan Hibrid untuk Kearifan Lokal Siswa SMK

Sulomo Agil Budiyanto — 2026-02-07

This study aims to develop an inclusive hybrid character education model based on local wisdom using K-Means Clustering analysis to group students’ character profiles according to honesty, discipline, cooperation, responsibility, tolerance, care, pride in tradition, and mutual cooperation. The study employs an exploratory quantitative approach involving 115 students through questionnaires and observations. The results indicate that hybrid learning based on local wisdom enhances students’ character development and supports personalized learning strategies. The K-Means Clustering method successfully grouped students into three character clusters: high (23 students), moderate (39 students), and low (54 students). Differences among clusters were particularly evident in the indicators of discipline, responsibility, cooperation, and mutual cooperation. Cluster quality evaluation showed a Sum of Squared Error (SSE) value of 536.64 with a stable decreasing pattern, while a Silhouette Score of 0.123 indicates that the cluster structure is reasonably formed despite natural overlaps in human behavior. Visualization using Principal Component Analysis (PCA) revealed that the high character cluster is more distinctly separated compared to the moderate and low clusters. The findings imply that data-driven character mapping can serve as a basis for teachers to design more appropriate hybrid learning tailored to students’ characteristics, enabling character development to be conducted in a more focused, gradual, and contextually grounded manner in accordance with the school’s local wisdom.

Keywords: Character; Hybrid; K-Means Clustering; Local Wisdom