Claim Missing Document
Check
Articles

Found 14 Documents
Search

Image-Based Position Control for Three-Wheel Omni-Directional Robot Sukamta, Sri; Nugroho, Anan; Subiyanto, Subiyanto; Rezianto, Rizal; Soambaton, Muhammad Febrian; Ardiyanto, Agus
Jurnal Ilmiah Teknik Elektro Komputer dan Informatika Vol. 10 No. 3 (2024): September
Publisher : Universitas Ahmad Dahlan

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26555/jiteki.v10i3.29601

Abstract

Robotic technology continues to advance, but real-time position tracking of robot movements still faces challenges, especially in dynamic and irregular environments. Sensors often fail to maintain accuracy due to environmental disturbances. This study introduces an innovation using omnidirectional wheels on a three-wheeled robot and image-based position control to improve maneuverability and precision. This method utilizes an infrared camera mounted on the ceiling to detect the robot's position. Image processing algorithms are used to determine waypoints and direct the robot. Omni-directional wheels allow the robot to move freely in all directions, which is important for accurate navigation on complex trajectories. The robot was tested on the "X" and "Square" trajectories in a 1.8-meter x 1.8-meter room to rotate vertical, horizontal, and diagonal movements. The test results showed that on the "X" trajectory, the second movement had the most significant error with a Mean Absolute Error (MAE) of 134.96, while the third movement had the slightest error with an MAE of 52.49, with an average error of 91.36. The first and third movements in the “Square” trajectory showed more significant errors than the second and fourth movements, with MAE of 105.37 and 100.47, respectively. The second and fourth movements had MAE of 67.20 and 59.65, with an average error of 83.17. These results indicate that the image-based control system and omnidirectional wheels improve accuracy compared to conventional methods, which is important for robot navigation applications. Practical implications of this technology include potential applications in the robotics and automation industry. Future research should focus on developing more precise control algorithms and sensors to improve accuracy. Directions for future research include exploring more sophisticated image processing techniques and applying this technology to various industrial scenarios. 
Implementasi Multilayer Perceptron Artificial Neural Network untuk Prediksi Konsumsi Energi Listrik PT PLN (Persero) UP3 Salatiga Saputra, Roni; Sunardiyo, Said; Nugroho, Anan; Subiyanto, Subiyanto
Elektrika Vol. 15 No. 2 (2023): October 2023
Publisher : Universitas Semarang

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.26623/elektrika.v15i2.6411

Abstract

Electricity is energy that flows through cable networks and has become an important part of the progress of human civilization in various fields. The high demand for electrical energy for consumers requires providers of electrical energy to provide a reliable but economical supply of electrical energy. Therefore, strategies and methods are needed to adjust the supply and demand of electrical energy. This can be achieved by carrying out proper and appropriate operational planning. One of the important steps in planning the operation of an electric power system is predicting the demand for electrical energy. However, in the existing research there are still deficiencies in the form of a high error rate. The purpose of this study was to determine the implementation of the multilayer perceptron artificial neural network to predict electricity in 2022-2026 at PT PLN (Persero) UP3 Salatiga. The study used time series data on electricity consumption for the previous 5 years. Based on the research that has been done, the best network variation is TRAINGDA 4 hidden layer with 20 hidden layer nodes, this network model at the training stage produces output with MAD of 2,624,072 kWh and MAPE of 2.79%, and at the stage testing produced an output with MAD of 3,728,386 kWh and MAPE of 3.24%. Keywords: Multilayer perceptron artificial neral network, Forecasting, Electricity consumption.ABSTRAK  Listrik merupakan energi yang mengalir melalui jaringan kabel serta sudah menjadi bagian yang penting dalam kemajuan peradaban manusia di berbagai bidang. Tingginya kebutuhan energi listrik pada konsumen mengharuskan penyedia energi listrik menyediakan suplai energi listrik yang handal tetapi tetap ekonomis. Oleh karena itu, diperlukan strategi dan metode untuk penyesuaian antara supplay dan demand energi listrik. Hal tersebut dapat dicapai dengan melakukan perencanaan operasi yang baik dan tepat, salah satu langkah perencanaan operasi sistem tenaga listrik yang penting yaitu prediksi kebutuhan energi listrik. Namun dalam penelitian yang ada masih terdapat kekurangan berupa tingkat kesalahan yang masih cukup tinggi. Tujuan dari penelitian ini adalah untuk mengetahui implementasi multilayer perceptron artificial neural network untuk melakukan prediksi listrik pada tahun 2022-2026 pada PT PLN (Persero) UP3 Salatiga. Penelitian menggunakan data time series konsumsi energi listrik 5 tahun sebelumnya. Berdasarkan penelitian yang telah dilakukan didapatkan variasi jaringan terbaik yaitu TRAINGDA 4 hidden layer dengan 20 node hidden layer, model jaringan ini pada tahap training menghasilkan output dengan nilai MAD sebesar 2,624,072 kWh dan MAPE sebesar 2.79%, serta pada tahap testing menghasilkan output dengan nilai MAD sebesar 3,728,386 kWh dan MAPE sebesar 3.24%.
PEMANFAATAN TEKNOLOGI COMPUTER VISION BERBASIS YOLO UNTUK MENDETEKSI KERUMUNAN DI SMKN 4 MALANG Nugroho, Anan; Indaryanto, Faizal; Suni, Alfa Faridh; Arfriandi, Arief; Wibawanto, Hari; Oktaviyanti, Dwi; Savitri, Dina Wulung
Jurnal Abdi Insani Vol 11 No 1 (2024): Jurnal Abdi Insani
Publisher : Universitas Mataram

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29303/abdiinsani.v11i1.1437

Abstract

In limited face-to-face learning, teachers and students must implement health protocols to prevent the spread of the Covid-19 virus. However, implementing health protocols in schools as a new habit in the midst of a pandemic is certainly not easy. There are many reports related to the number of violations of health protocols in schools during face-to-face learning. Therefore, by innovating and utilizing existing technology in the 4.0 era can help us to detect social boundaries. The purpose of this service activity is so that teachers and students can learn YOLO-based computer vision technology at SMKN 4 Malang as a means of preventing the spread of Covid-19. In addition, teachers and students can also learn to make simple applications based on YOLO. This activity begins with the socialization of YOLO-based computer vision technology as a crowd detection tool to the school. Then the design and manufacture of crowd detection tools by the service team, training in making crowd detection applications, and ending with a discussion between the trainees and the service team. The results of the service show that this activity has succeeded in developing a crowd detection tool that can help calculate the number and distance of people who do not apply health protocols at SMKN 4 Malang. This tool makes it easier for schools to monitor the activities of school residents in implementing health protocols. This service activity is very useful and in demand by teachers and students. This is evidenced by the enthusiasm of the participants in participating in the training of YOLO-based crowd detection tools.
Peningkatan kompetensi elektronika siswa SMKN 4 Semarang melalui pelatihan rakit drone quadcopter Nugroho, Anan; Wahyudi, Wahyudi; Rusiyanto, Rusiyanto; Naryanto, Rizqi Fitri; Ghulamzaki, Muhammad; Mubarak, Raihan Fa'iq; Hidayatulloh, Dimas Restu
SELAPARANG: Jurnal Pengabdian Masyarakat Berkemajuan Vol 8, No 3 (2024): September
Publisher : Universitas Muhammadiyah Mataram

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.31764/jpmb.v8i3.25385

Abstract

AbstrakPentingnya penguasaan teknologi bagi siswa Teknik Elektronika Industri di SMKN 4 Semarang dapat diwujudkan melalui pelatihan perakitan quadcopter yang mendukung pemahaman dan praktik langsung dalam menghadapi perkembangan teknologi. Tujuan dari kegiatan pengabdian ini adalah agar siswa dapat belajar dan memiliki pengetahuan dalam bidang elektronika, khususnya pemahaman terkait quadcopter. Metode pelaksanaan meliputi merumuskan masalah, menentukan komponen, penyampaian materi, praktik merakit, dan praktik menerbangkan quadcopter. Pelatihan perakitan quadcopter diadakan pada hari kamis, 25 April 2024, di SMKN 4 Semarang, dengan jumlah peserta sebanyak empat puluh siswa. Pelatihan berhasil membuka wawasan bagi siswa SMKN 4 Semarang untuk mengembangkan kompetensi elektronika dengan mengetahui komponen, cara merakit, dan cara menerbangkan quadcopter.  Kata kunci: elektronika; kompetensi; quadcopter; pelatihan; perakitan AbstractThe importance of mastering technology for Industrial Electronics Engineering students at SMKN 4 Semarang can be realized through quadcopter assembly training that supports understanding and direct practice in dealing with technological developments. The purpose of this service activity is for students to learn and have knowledge in the field of electronics, especially understanding related to quadcopters. The implementation method includes observation, formulating problems, delivering material, practicing assembling, and practicing flying the quadcopter. The quadcopter assembly training was held on Thursday, April 25, 2024, at SMKN 4 Semarang, with forty students participating. The training succeeded in opening insights for SMKN 4 Semarang students to develop electronic competencies by knowing the components, how to assemble, and how to fly the quadcopter.  Keywords: electronics; competency; quadcopter; service; assembly