Jurnal Elektro

Comparison of Actual Results and PVSyst Simulation in the Design of Off-Grid Solar Power Generation System (PLTS) in Karuni Village, Southwest Sumba

Marsul Siregar — 2024-04-29

This research aims to compare the actual production with the simulations using the PVSyst software for the Off-Grid Solar Power Plant (PLTS) in Karuni Village, Southwest Sumba. The Off-Grid PLTS in Karuni Village is a vital solution for improving remote areas' electricity access. Actual energy production data from the PLTS were obtained from monitoring systems, while simulation results were obtained through PVSyst. The analysis results indicate a difference of approximately 10% between the actual and simulated results. It observed that it is influenced by variability in local weather conditions, maintenance, system management levels, and limitations of the simulation model. The implications of this research emphasize the importance of using accurate data in simulations, improving PLTS system maintenance, and developing more sophisticated simulation models. Recommendations for further research include further analysis of factors influencing the differences in results. This study provides valuable insights into the planning and management of Off-Grid PLTS. It offers perspectives on enhancing the accuracy of future PLTS system planning and management.

Effect of Combining of Shaping at Magnet Edge and Dummy Slotting at Stator Core on the Cogging Torque Reduction in Permanent Magnet Generator

Lucio Alfreido Correira — 2024-04-29

Effect of combining the shaping at magnet edge and dummy slotting at stator on the cogging torque reduction in permanent magnet generator was studied. The cogging torque is the most issues in permanent magnet generator application, since it effects to reduce the generator performance. A permanent magnet generator with 24 slots and 20 poles structure is presented. For purpose of study, in the beginning the permanent magnet generator with conventional structure has been investigated as the basic reference of initial structure. In order to improve the performance of the permanent magnet generator, the magnet edge was shaped but the stator core is kept remain conventional structure. It can be found that the magnet edge effects to inprove a little bit of the performance of the permanent magnet generator. In the next investigation, the magnet edge slotting and dummy slotting in stator core of the permanent magnet generator was combined. Using the FEMM 4.2, the cogging torque value of the proposed of permanent magnet generator were analysed and compared. It was found that cogging torque reduction of the permanent magnet generator with combining the dummy slotting at stator core and magnet edge shaping can reduce the cogging torque from 0.0028554 N.m (intial base line) to 0.0000165 N.m (proposed model). In other word, the cogging torque can be to around 99.42 % compared with the base line. The magnet flux density at core of the permanent magnet generator proposed can be reduced to 35 %.

Perbaikan Kurva Beban Harian pada Industri Kecil: Studi Kasus PT. X

Arka D. Soewono — 2024-04-29

This paper presents the improvement of daily load and load balance curve of home-scale garment industry with the study case in PT. X. Research is started with load measurement on every machine, proceeded by plotting schedule of each machine. The next step is rescheduling daily load and per-phase load balancing. Initial data shows unbalance load sharing between phase with the average load of 4916 kW and standard deviation of 2077. After being rescheduled and phase rearrangement, the interphase load sharing is more balance, with the average load of 4903 kW and standard deviation of 952.

Sistem Pemilahan Barang Berdasarkan Deteksi Label Menggunakan Vision Sensor

Carolus Henry — 2024-04-29

Sorting goods based on the results of checking packaging labels is an important process in controlling production quality in industry. Many industries still carry out manual sorting and label checking processes, which results in low productivity levels and is susceptible to human error. This research develops an automation system for sorting goods based on label inspection using vision sensors, programmable logic controller (PLC), and robot arm. The system controlled by a PLC will read and detect damage to packaging labels by using VeriSens vision sensor and sort them using a robot arm according to predetermined stock keeping unit (SKU) categories, namely SKU 1, SKU 2, SKU 3, SKU 4, and rejected goods. The pneumatic system is used as an actuator to push goods onto the conveyor, moving the robot arm with three degrees of freedom and vacuum. Detection is carried out by applying the edge detection concept to read text, images and code that are reprocessed with the VeriSens Application Suite software. The success rate of the goods sorting system reached 90% with a reading speed of 0.389 seconds and a work process duration ranging from 21.54 seconds to 28.99 seconds.

Perbandingan Algoritma Machine Learning menggunakan Orange Data Mining untuk Klasifikasi Jenis Kendaraan pada Sistem Tilang Digital

Egipta Pranadjaya — 2024-04-29

This paper discusses the application of the Orange Data Mining application to compare several machine learning algorithms for classifying vehicle types in digital ticket systems. This research compares and analyzes the logistic regression algorithm, Support Vector Machine (SVM) and Neural Network (NN) to solve vehicle classification problems in digital traffic tickets. The research results show that in the training process and based on the dataset used, the algorithms that have the highest level of accuracy are Logistic Regression, Neural Network and Support Vector Machine. Meanwhile, during the testing process, all algorithms in the model were able to carry out classification with 100% accuracy