Febrianto, Mochammad Arief (2023) Analisis Prediksi Power Generation Panel Surya Menggunakan Metode K – Nearest Neighbor Dan Backpropagation Neural Network Pada Sistem Real-Time Dashboard Monitoring Berbasis Internet Of Things. Other thesis, Institut Teknologi Sepuluh Nopember.
![[thumbnail of 02311940000099-Undergraduate_Thesis.pdf]](http://repository.its.ac.id/style/images/fileicons/text.png) |
Text
02311940000099-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 September 2025. Download (6MB) | Request a copy |
Abstract
PLTS adalah sumber EBT yang cocok dan sering diaplikasikan di Indonesia. Namun seiring perkembangannya, terdapat masalah yang sering muncul yakni produksi energi dari PV yang tidak stabil, turunnya efisiensi energi karena shading, serta sulitnya maintenance di daerah remote. Untuk mengatasnya, dibuatlah sistem monitoring PV secara real-time dan merancang model prediksi daya PV dalam beberapa jam kedepan. Sensor yang digunakan yakni sensor iradiasi (TSL2591), sensor suhu (DHT11), kecepatan angin (weather station) dan sensor arus (ACS712). Monitoring menggunakan platform pihak ketiga untuk menampilkan data yang dikirim mikrokontroller. Kualitas akuisisi data memiliki indeks 2,6 sehingga termasuk dalam kategori sedang. Dataset yang didapat digunakan sebagai input prediksi daya PV menggunakan metode k-Nearest Neighbor (k-NN), Decomposition (D) dan Backpropagation Neural network (BPNN). Input diprediksi menggunakan SARIMA terlebih dahulu agar mendapatkan nilai input prediksi. Setelah digenerate, input digabungkan dan diproses oleh model tersimpan. Model yang digunakan menggunakan variasi k-NN, k-NN-BPNN dan k-NN-D-BPNN. Model yang dibentuk memiliki hasil MAPE sebesar 0,52% untuk k-NN, 0,95% untuk k-NN-BPNN dan 33,47% untuk k-NN-D-BPNN serta MSE sebesar 59,84 W2 untuk k-NN, 225,94 W2 untuk k-NN-BPNN serta 17.701 W2 untuk k-NN-D-BPNN sehingga model termasuk prediksi yang sangat bagus dan layak. Ketika waktu prediksi ditambah, maka akurasi yang dihasilkan semakin menurun. Oleh karena itu, prediksi perlu dibatasi dalam 3 jam kedepan.
====================================================================================================================================
PLTS is a suitable source of EBT and is often applied in Indonesia. However, along with its development, some problems often arise such as unstable energy production from PV, decreased energy output efficiency due to shading, and maintenance difficulties in remote areas. To overcome this problem, a PV monitoring system was created in real-time and was developed a prediction model for PV power in the next few hours. The sensors include an irradiation sensor (TSL2591), temperature sensor (DHT11), wind speed (weather station), and current sensor (ACS712). Monitoring uses the help of a third-party platform, so it can display data sent by the microcontroller. The quality of data acquisition has an index of 2.6, thus it is categorized in the medium index. The dataset obtained will be used as input for predicting PV power using the k-Nearest Neighbor (k-NN), Decomposition (D), and Backpropagation Neural network (BPNN) methods. The input is predicted using SARIMA, to get the predicted input value. After being generated, the inputs are processed by the stored model. The model used uses variations of k-NN, k-NN-BPNN, and k-NN-D-BPNN. The model formed has a MAPE yield of 0.52% for k-NN, 0.95% for k-NN-BPNN and 33.47% for k-NN-D-BPNN and MSE of 59.84 W2 for k-NN, 225.94 W2 for k-NN-BPNN and 17,701 W2 for k-NN-D-BPNN so that the model is a very good and feasible prediction. When the prediction time is added, the resulting accuracy decreases. Therefore, predictions need to be limited to the next 3 hours.
| Item Type: | Thesis (Other) |
|---|---|
| Uncontrolled Keywords: | Dashboard Monitoring, Forecast, k-NN-D-BPNN, QoS, SARIMA, Dashboard Monitoring, Forecast, k-NN-D-BPNN, QoS, SARIMA |
| Subjects: | T Technology > T Technology (General) > T174 Technological forecasting T Technology > TJ Mechanical engineering and machinery > TJ808 Renewable energy sources. Energy harvesting. |
| Divisions: | Faculty of Industrial Technology and Systems Engineering (INDSYS) > Physics Engineering > 30201-(S1) Undergraduate Thesis |
| Depositing User: | Mochammad Arief Febrianto |
| Date Deposited: | 26 Jul 2023 14:55 |
| Last Modified: | 26 Jul 2023 14:55 |
| URI: | http://repository.its.ac.id/id/eprint/99191 |
Actions (login required)
 |
View Item |