Claim Missing Document
Check
Articles

Pengendalian Tegangan Generator Oksihidrogen Menggunakan Metode Kontrol Pid Hanafi Rusmantika; Dudi Darmawan; Reza Fauzi Iskandar
eProceedings of Engineering Vol 6, No 2 (2019): Agustus 2019
Publisher : eProceedings of Engineering

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

AbstrakDalam jurnal ilmiah ini, penulis melaporkan penelitian penulis mengenai perancanganalgoritma kontrol PID pada sebuah model sistem sel oksihidrogen yang terdiri dari sebuahmodel konverter buck-boost dan model sel elektrolisis yang dirangkai secara lup-tertutup.Dari penelitian penulis, ditemukan konstanta PID Kp, Ki, dan Kberturut-turut adalah0,02982, 39,236842, dan 0,0000056658. Selain itu ditemukan juga rasio redaman ζ dan frekuensi natural ωn d berturut-turut 0,02679, dan 2187,782 Hz. Hasil temuan penulismenunjukkan perilaku sistem pada simulasi lebih lambat dibandingkan dengan prediksi perilaku sistem orde-dua secara analitik. Hal ini dikarenakan adanya zero bidang kanan padafungsi transfer dari sistem serta karakteristik dari rangkaian konverter buck-boost itu sendiri.Selain itu, algoritma kontrol PID berhasil menstabilkan sistem dibuktikan dari hasil uji setpointdanplotNyquistyangdibuatberdasarkanfungsitransferdarisistem.Kata Kunci : PID, oksihidrogen, konverter buck-boost, MATLAB, Simulink AbstractsIn this science paper work, a research on PID control scheme on a closed-loop oxyhydrogen cellsystem, which consists of a buck-boost converter and an electrolysis cell, is reported. From theresearch, it is concluded that the PID constants Kp, Ki, dan Kare respectively 0.02982,39.236842, dan 0.0000056658. Besides, it is also found that the damping ratio ζ and natural frequency ωn d are respectively 0.02679, dan 2187.782 Hz. On this research, the system’sbehavior is found out slower than a second-order system model by analytical approach. This is due to a right-half-plane zero in the system’s transfer function and the unique characteristic ofthe buck-boost converter itself. It is also found that the PID Control Algorhytm can stabilizesystem based on the set-point tests conducted and the Nyquist Plot generated based on the system’stransfer function. Keywords : HHO, Electrolysis, Buck-Boost Converter, PID, Ziegler-Nichols, MATLAB
Pengaruh Jenis Biomassa Dan Kecepatan Aliran Udara Terhadap Kinerja Kompor Gasifikasi Biomassa (the Influence Of The Biomass Type And The Speed Of Air Flow On The Performance Of Biomass Gasification Stove) Chartika Fadilah Prasetiani; Suwandi Suwandi; Reza Fauzi Iskandar
eProceedings of Engineering Vol 6, No 2 (2019): Agustus 2019
Publisher : eProceedings of Engineering

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Abstrak Sumber energi alternatif dari limbah pengolahan perhutanan dan pertanian memiliki potensi yang berlimpah, haltersebut dapat mengurangi ketergantungan masyarakat terhadap bahan bakar fosil seperti minyak tanah dan gasbumi (LPG). Ketergantungan masyarakat terhadap bahan bakar fosil tersebut masih tinggi, sehingga menyebabkanketersedian terbatas dan cenderung berkurang. Oleh karena itu, memerlukan penanganan dalam pemanfaatansumber energi alternatif diantaranya sumber energi yang berasal dari biomassa. Salah satu teknologinya yaitudiaplikasikan pada kompor dengan teknik gasifikasi Top-Lit Updraft. Penelitian ini bertujuan untuk mengetahuikinerja kompor gasifikasi dari variasi jenis biomassa dan kecepatan aliran udara pada reaktor. Pengujian komporgasifikasi biomassa dilakukan sesuai prosedur SNI Tungku Biomassa 7.926, tahun 2013. Setiap pengujian yangdilakukan menggunakan tiga variasi jenis biomassa, yaitu : wood pellet sekam padi, wood pellet kayu jati danwood pellet kayu sengon. Metode pengujian menggunakan enam variasi kecepatan aliran udara primer, yaitu: 1,1m/s; 2,5 m/s; 3,5 m/s; 4,5 m/s; 5 m/s dan 5,4 m/s. Hasil pengujian kompor gasifikasi biomassa yang dilakukandengan bahan bakar wood pellet sekam padi memiliki total waktu operasi kompor tercepat dengan kecepatan aliranudara 5,4 m/s, massa bahan bakar 0,3 kg dengan waktu pengoperasian kompor selama 10,46 menit. Bahan bakarwood pellet sekam padi yang digunakan untuk mendidihkan air sebanyak satu liter membutuhkan waktu 5,56menit, dengan temperatur reaktor (temperatur nyala api) sebesar 295,4 °C. Nilai efisiensi termal tertinggi padabahan bakar wood pellet sekam padi sebesar 47,65 %.Kata kunci: kompor biomassa; T-LUD; wood pellet. AbstractAlternative energy sources from forestry and agricultural processing waste has abundant potential, this can bereduce people's dependence on fossil fuels such as kerosene and natural gas (LPG). Society dependency onfossil fuels is still macro, causing limited availability and tends to decrease. Therefore, it requires handles inthe use of alternative energy sources such as energy sources derived from biomass. One of the technologies isapplied to the stove using the Top-Lit Updraft gasification technique. This research aims to determine theperformance of gasification stoves from variations in biomass type and velocity of air flow in the reactor. Thetesting of biomass gasification stoves was carried out according to the SNI Furnace Biomass 7.926 procedurein 2013. Each test carried out using three variations of biomassa types, namely: rice husk wood pellets, teakwood pellets and sengon wood pellets. The test method uses six variations of the primary air flow velocity,namely: 1.1 m / s; 2.5 m / s; 3.5 m / s; 4.5 m / s; 5 m / s and 5.4 m / s. Result the testing of biomass gasificationstoves with rice husk pellet wood has the fastest operating time of the stove with an air flow rate of 5.4 m / s,a fuel mass of 0.3 kg with a stove operating time of 10.46 minutes. Pellet rice husk wood fuel used to boil oneliter of water takes 5.56 minutes, with the reactor temperature (flame temperature) of 295.4 ° C. The highestthermal efficiency value in rice husk pellet wood fuel is 47.65%.Keywords: biomass stove; T-LUD; wood pellets.
Implementasi Filter Kalman Untuk Mereduksi Derau Pada Respon Dinamik Rangkaian Rlc Guntur Bhatara Sutra; Asep Suhendi; Reza Fauzi Iskandar
eProceedings of Engineering Vol 6, No 2 (2019): Agustus 2019
Publisher : eProceedings of Engineering

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Abstrak Sistem spring-mass-damper dapat disimulasikan pada rangkaian RLC dengan input tegangan AC karena keduanya merupakan sistem orde dua dan memiliki karakteristik dinamik seperti bentuk respon dan respon waktu.Pada penelitian ini nilai tegangan input dan induktor yang digunakan tetap. Nilai output yang diamati hanyategangan pada kapasitor, dan karakteristik dinamik yang diamati hanya bentuk respon critically damped danoverdamped. Output dari rangkaian RLC ini masih memiliki kemungkinan menghasilkan derau. Karena outputakan diamati pada mikrokontroler, maka dibutuhkan sebuah filter digital sehingga filter yang dipilih adalah filterkalman. Tujuan penelitian ini adalah mengamati pengaruh filter kalman terhadap derau dari output dan mengetahuirasio deraunya. Nilai resistor pada rangkaian diubah saat kapasitor bernilai sama dan begitupun sebaliknya,sehingga didapat respon yang diinginkan. Output dari rangkaian difilter oleh filter kalman pada arduino. Setelahitu dibandingkan nilai derau sebelum dan setelah filter kalman diaplikasikan dan diamati rasio deraunya. Padakeadaan transient rasio derau sebelum dan sesudah difilter adalah 0,17V dan 0,37V. Pada keadaan steady staterasio derau sebelum dan sesudah difilter adalah 0,16V dan 0,05V. Dari hasil penelitian yang telah dilakukan, filterkalman dapat mengurangi derau pada keadaan steady state. Kata kunci: critically damped, overdamped, derau, filter kalman. ABSTRACT Spring-mass-damper system can be simulated on RLC circuits with AC voltage as the input because both of them are second system order and have dynamic characteristics such as shape respond and time respond. Inthis research the value of input voltage and inductor was fixed. The output value that will be observe just thevoltage on capacitor, and the dynamic characteristics that will be observe just the shape respond of criticallydamped and overdamped. The output of this RLC circuits still have a chance of making a noise. Because the outputwill be observed at microcontroller, then it needs a digital filter so the kalman filter was choosen as the filter. Theaim of this research is to obvserve the influence of kalman filter to the noise from the output and find out the noiseratio. The resistor value on the cirtuicts will be modified when the capacitor fixed and otherwise, so the aimresponds will be obtained. The output from the circuits will be filtered by the kalman filter on arduino. And thenthe noise value before and after the kalman filter applied will be compared and the noise ratio will be observed.When transient state the noise ratio before and after filtered was 0,17V and 0,37V. when steady state the noiseratio before and after filtered was 0,16V and 0,05V. From the research that has been done, the kalman filter canreduce the noise when steady state.Keyword: critically dampd, overdamped, noise, kalman filter.
Pengaruh Variasi Durasi Dan Interval Penyemprotan Pada Sistem Pertanian Aeroponik Untuk Budidaya Cabai Rawit (capsicum Frutescens L.) Ary Halimawan; Ahmad Qurthobi; Reza Fauzi Iskandar
eProceedings of Engineering Vol 7, No 1 (2020): April 2020
Publisher : eProceedings of Engineering

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Abstrak Cabai rawit (Capsicum frutescens L.) merupakan salah satu bumbu masakan yang cukup populer di Indonesia. Hal ini menyebabkan permintaan akan cabai rawit di pasaran cukup tinggi jika dibandingkan komoditas sayuran lainnya. Namun, sayangnya kebutuhan masyarakat akan cabai rawit tidak selalu dapat terpenuhi dengan baik. Hal ini dikarenakan jumlah persediaan cabai rawit di pasaran selalu berfluktuasi dari waktu ke waktu. Perubahan jumlah persediaan cabai rawit di pasaran juga dipengaruhi oleh pasokan dari petani yang bertindak sebagai pemasok utama cabai rawit di pasaran. Oleh karena itu, diperlukan metode penanaman yang dapat membuat tanaman cabai rawit dapat tumbuh lebih cepat jika dibandingkan dengan metode konvensional yang menggunakan tanah sebagai media tanamnya. Tujuan penelitian ini adalah mengetahui pengaruh dari variasi durasi dan interval penyemprotan terhadap pertumbuhan tanaman cabai rawit yang akan ditanam dengan menggunakan metode aeroponik. Penelitian dilakukan dengan membuat dua buah sistem aeroponik yang identik, salah satu sistem memiliki sprayer yang akan aktif selama 5 detik dan kemudian mati selama 3 menit, sedangkan sistem yang lain akan memiliki sprayer yang akan aktif selama 30 detik dan kemudian mati selama 15 menit. Hasilnya, tanaman pada sistem aeroponik dengan sprayer yang aktif selama 30 detik dan kemudian mati selama 15 menit memiliki rata -rata pertumbuhan batang 3,06 kali lebih cepat dan rata-rata pertumbuhan akar yang 1,47 kali lebih cepat jika dibandingkan dengan tanaman pada sistem aeroponik dengan sprayer yang aktif selama 5 detik dan kemudian mati selama 3 menit. Hal ini dikarenakan tanaman yang ditanam pada sistem aeroponik dengan sprayer yang aktif selama 30 detik dan mati selama 15 menit memiliki bagian akar yang sudah mulai kering sehingga dapat menyerap lebih banyak oksigen jika dibandingkan dengan tanaman yang ditanam pada sistem aeroponik dengan sprayer yang aktif selama 5 detik dan kemudian mati selama 3 menit. Kata kunci: Cabai rawit, Kelembapan, Lebih cepat, Aeroponik Abstract Thai Chilli (Capsicum frutescens L.) is one of the most popular spices in Indonesia. This causes the demand for thai chilli on the market to be higher when compared to other vegetable commodities. However, people’s need for thai chilli cannot always be met properly. This is because the amount of thai chilli in the market always fluctuates from time to time. The fluctuating amount of thai chilli supplies on the market is also influenced by supplies from farmers who act as the main suppliers of thai chilli on the market. Therefore, other planting methods that can make thai chilli plants grow faster are needed. The purpose of this research was to understand the effect of variation in spraying duration and interval on aeroponic system for thai chilli plants growth. The research was doing by making two identical aeroponic systems, one of system will have sprayer turn on for 5 seconds and turn off for 3 minutes, while the other system will have sprayer turn on for 30 seconds and turn off for 15 minutes. The result, plants on the aeroponic system that have sprayer turn on for 30 seconds and turn off for 15 minute have an average stem growth of 3.06 times faster and an average root growth of 1.47 times faster when compared to the plants that planted on the aeroponic system that have sprayer turn on for 5 seconds and turn off for 3 minutes. This is because the plants that are planted in the aeroponic system that have sprayer turn on for 30 seconds and turn off for 15 minutes have dry root’s part so they can absorb more oxygen when compared to plants that planted in the aeroponic system that have sprayer turn on for 5 seconds and then turn off for 3 minutes. Keywords: Thai Chili, Humidity, Faster, Aeroponic
Rancang Bangun Alat Ukur Massa Jenis Zat Cair Portabel Menggunakan Load Cell Dan Sensor Ultrasonik Pada Pipa U Nanda Aulan Tamami; Endang Rosdiana; Reza Fauzi Iskandar
eProceedings of Engineering Vol 7, No 1 (2020): April 2020
Publisher : eProceedings of Engineering

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Abstrak Alat ukur massa jenis zat cair portable ini menggunakan prinsip hukum Pascal sebagai cara kerja dari alat ukur ini, dimana hukum Pascal tersebut bekerja pada pipa U yang di dalam pipa tersebut terdapat cairan dan dibuat system nya tertutup, selain itu alat ukur ini juga menggunakan load cell dan ultrasonik sebagai pengukur massa dan ketinggian air yang terdapat dalam alat ukur massa jenis tersebut. Dimana untuk sensor ultrasonik memiliki rentang pengukuran 3 cm sampai 400 cm sedangkan load cell yang digunakan dapat mengukur beban sampai 2 kg. system pengolah data dari sensor tersebut menggunakan mikrokontroller Atmega 328 dengan driver Arduino uno dan data hasil pengukuran ditampilkan oleh display 16x2. Untuk kalibrasi sensor ultrasonik digunakan penggaris yang diletakkan di dinding alat ukur sehingga kenaikan air dapat dilihat melalui penggaris yang telah diletakkan tersebut, sedangkan massa dari cairan telah diukur sebelum mmasuk alat ukur sebagai kalibrasi dari dari load cell. Untuk sampel yang diukur adalah minyak goreng, oli, air, gliserin, dan tepung gandum. Alat ukur ini memiliki akurasi sekitar 80 sampai 98%, memiliki presisi sekitar 0.031 sampai 0.14, dan memiliki error sebesar 0.025 atau 2.84% Kata kunci : hukum Pascal, massa jenis, ultrasonik, load cell, pipa U Abstract This portable liquid density measuring instrument uses the principles of Pascal law as the workings of this measuring instrument, where Pascal's law works on the U pipe in which there is liquid and the system is closed, besides this measuring instrument also uses load cell and ultrasonik as a measure of the mass and height of water contained in the density meter. Where for ultrasonik sensors has a measurement range of 3 cm to 400 cm while the load cell used can measure loads up to 2 kg. Data processing system from the sensor uses the Atmega 328 microcontroller with the Arduino uno driver and the measurement data is displayed by a 16x2 display. To calibrate the ultrasonik sensor a ruler is placed on the wall of the measuring device so that the increase in water can be seen through the ruler that has been placed, while the mass of the liquid has been measured before entering the measuring instrument as a calibration of the load cell. The samples measured were cooking oil, oil, water, glycerin, and wheat flour. This measuring instrument has an accuracy of around 80 to 98%, has a precision of about 0.031 to 0.14, and has an error of 0.025 or 2.84% Keywords: Pascal law, density, ultrasonik, load cell, pipe U
Implementasi Metode Background Subtraction Pada Citra Area Parkir Sebagai Upaya Untuk Membangun Citra Satuan Ruang Parkir Virtual Ahmad Faridh; Reza Fauzi Iskandar; Hertiana Bethaningtyas Dyah K
eProceedings of Engineering Vol 7, No 3 (2020): Desember 2020
Publisher : eProceedings of Engineering

Show Abstract | Download Original | Original Source | Check in Google Scholar

Abstract

Abstrak Mencari tempat parkir kosong membutuhkan waktu yang cukup lama di area parkir yang luas.untuk mencari tempat parkir kosong membutuhkan waktu yang cukup lama. Untuk mengatasi hal tersebut dalam penelitian ini maka akan digunakan pendeteksian tempat parkir kosong menggunakan metode background subtraction. Background subtraction merupakan proses pendeteksian objek dengan membedakan antara objek dengan latar yang terdapat dalam suatu citra. Selanjutnya background subtraction melalui proses thresholding untuk menarik batas latar menjadi hitam dan objek menjadi warna putih, sehingga akan ada perbedaan antara objek dan latar. Dimana pada proses pendeteksian menggunakan maket dari tempat parkir dengan perbandingan 1 : 60. Dengan melakukan pengujian dengan menggunakan metode background subtraction dipengaruhi oleh cahaya, semakin besar intensitas cahaya maka semakin bagus juga kualitas pendeteksiannya. Untuk mendeteksi juga membutuhkan nilai threshold yang tepat diperoleh dari niliai histogram frame awal. Sehingga pada proses pendeteksian di penelitian ini menggunakan nilai thershold 40 dengan intensitas cahaya 944 lux, agar mendapatkan persentase ketepatan pendeteksian 100%. Dalam penggunaan metode ini harus menggunakan intensitas cahaya yang konstan agar tidak terjadi kesalahan dalam proses pendeteksian. Kata Kunci: background subtraction,thresholding,camera,cell parkir, latar, objek. Abstract Finding an empty parking lot takes a long time in a large parking area. To overcome this problem in this study, the detection of empty parking spaces will be used using the background subtraction method. Background subtraction is the process of detecting objects by distinguishing between objects and backgrounds in an image. Furthermore, background subtraction goes through a thresholding process to draw the background border to black and the object to white, so that there will be a difference between the object and the background. Where in the detection process using a mockup from the parking lot with a ratio of 1: 60. By conducting tests using the background subtraction method influenced by light, the greater the light intensity, the better the detection quality. To detect it also requires the exact threshold value obtained from the initial frame histogram values. So that the detection process in this study uses a thershold value of 40 with a light intensity of 944 lux, in order to get a percentage of 100% detection accuracy. In the use of this method must use a constant light intensity so that errors do not occur in the detection process. Keywords: background subtraction, thresholding, camera, parking cell, background, object.
Development of Non-Intrusive Load Monitoring of Electricity Load Classification with Low-Frequency Sampling Based on Support Vector Machine Edi Leksono; Auditio Mandhany; Irsyad Nashirul Haq; Justin Pradipta; Putu Handre Kertha Utama; Reza Fauzi Iskandar; Rezky Mahesa Nanda
Journal of Engineering and Technological Sciences Vol. 55 No. 2 (2023)
Publisher : Directorate for Research and Community Services, Institut Teknologi Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.5614/j.eng.technol.sci.2023.55.2.1

Abstract

Non-intrusive load monitoring (NILM) is a promising approach to provide energy consumption monitoring of electrical appliances and analysis of current and voltage data with less instrumentation. This paper proposes an electrical load classification model using support vector machine (SVM). SVM was chosen to keep the computational cost low and be able to implement an embedded system. The SVM model was utilized to classify the on/off state of air conditioners, light bulbs, other uncategorized electronics, and their combinations. It utilizes low-frequency sampling data captured every minute, or at a 0.0167 Hz rate. Utilization change in active and reactive power was used as a feature in the model training. The optimal kernel for the model was the radial basis function (RBF) kernel with C and gamma values of 88.587 and 2.336 as hyperparameters, producing a highly accurate model. In testing with real-time conditions, the model classified the on/off state of the electrical loads with 0.93 precision, 0.91 recall, and 0.91 f-score. The results of testing proved that the model can be applied in real time with high accuracy and with an acceptable performance in field implementation using an embedded system.
Pemodelan dan Simulasi MPPT pada Sistem PLTS Menggunakan Metode DNN Edi Leksono; Robi Sobirin; Reza Fauzi Iskandar; Putu Handre Kertha Utama; Mochammad Iqbal Bayeqi; Muhammad Fatih Hasan; Irsyad Nashirul Haq; Justin Pradipta
Jurnal Nasional Teknik Elektro dan Teknologi Informasi Vol 12 No 4: November 2023
Publisher : Departemen Teknik Elektro dan Teknologi Informasi, Fakultas Teknik, Universitas Gadjah Mada

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.22146/jnteti.v12i4.7931

Abstract

The maximum power point tracking (MPPT) feature in solar power plants is an essential function in increasing the efficiency of electricity production. The incremental conductance (InC) algorithm controls MPPT, aiming to maximize the output power of photovoltaic (PV) panels and increase the efficiency of the solar power plant system. Even though the InC algorithm is simple and practical, this algorithm tends to lack support in precise switching speeds, is sensitive to the measurement precision level, and is inadequate to eliminate power oscillations due to tight switching cycles. The deep neural network (DNN) algorithm has the potential to answer the challenges of MPPT dynamics. DNN’s learning capabilities enable the controller to better recognize the dynamics of shifts in maximum power values, thereby providing more appropriate contact actuation. The input for the DNN is the duty ratio produced by the InC algorithm. The DNN algorithm was implemented on three DC-to-DC power converter topologies, namely buck, boost, and buck-boost, to determine MPPT performance under standard tests and actual environmental conditions. DNN has demonstrated the ability to reduce oscillation effects, speed up steady-state time, and increase efficiency. In actual environmental conditions, the results showed that the buck converter consistently produced the highest power, followed by the boost and the buck-boost converters. Regarding performance efficiency, the buck converter achieved the highest efficiency at 94.58%, followed by the boost converter at 90.79%. Conversely, the buck-boost converter had the lowest performance efficiency, with an efficiency of 79.34%.
Optimasi Respon Flutter Aeroelastis Menggunakan Metode Interior Point Dan Kontrol Proporsional-Integral-Derivatif Reza F. Iskandar; Ahmad Qurthobi; Ayu B. Artini
Jurnal Teknik Elektro dan Komputer Vol. 5 No. 1 (2016): Jurnal Teknik Elektro dan Komputer
Publisher : Universitas Sam Ratulangi

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.35793/jtek.v5i1.11539

Abstract

Abstract — Flutter merupakan fenomena kopel antar struktur elastis yang mengalami vibrasi karena adanya gaya aerodinamis yang dihasilkan oleh aliran udara. Beberapa upaya yang telah dilakukan untuk meredam flutter antara lain: mengintegrasikan sirip dan aktuator redam; dan metode suspensi aktif yang bekerja secara real-time. Pada makalah ini, sistem reduksi flutter dilakukan dengan pendekatan respon  boundary. Optimasi respon dilakukan dengan metode interior-point dan disertai penambahan kontrol PID. Dari simulasi diperoleh puncak maksimum flutter turun menjadi 1,8 cm. Dengan penambahan kontrol PID, lonjakan maksimum sistem menjadi 16,8% dan kesalahan keadaan tunak sebesar 0%.   Abstract —Flutter is a coupling phenomenon between elastic structure which vibrated because of aerodynamic forces that produced by air flow. Some research was done to decrease flutter such as integration between fin and damping actuator; and real-time active suspention method. In this paper, flutter reduction system has been done using respon boundary approximation. System optimization and control have been developed with interior point method and PID controller, respectively. From the simulation,obtains flutter maximum peak value decrease into 1,8 cm. By adding PID Controller, obtain maximum overshoot  and steady state error are 16,8% and 0%, respectively.   Kata kunci — Flutter, IPM, Kontrol, Optimasi.
Adaptive Ensemble Learning for Enhancing Building Energy Consumption Prediction: Insights from COVID-19 Pandemic Energy Consumption Dynamics Handre Kertha Utama, Putu; Leksono, Edi; Nashirul Haq, Irsyad; Indrapraja, Rachmadi; Mahesa Nanda, Rezky; Friansa, Koko; Fauzi Iskandar, Reza; Pradipta, Justin
Journal of Engineering and Technological Sciences Vol. 57 No. 2 (2025): Vol. 57 No. 2 (2025): April
Publisher : Directorate for Research and Community Services, Institut Teknologi Bandung

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.5614/j.eng.technol.sci.2025.57.2.2

Abstract

Buildings account for approximately 40% of the total global energy consumption. Therefore, accurate prediction of building energy consumption is necessary to optimize resource allocation and promote sustainable energy usage. A key challenge in developing building energy consumption models is their adaptability to abrupt changes in consumption patterns owing to extraordinary events, such as the COVID-19 pandemic. Therefore, a two-layer ensemble-learning (EL) model incorporating sliding windows as input features is proposed. The model is a two-layer stacking EL consisting of two base learning methods: (1) support vector regression (SVR), and (2) random forest (RF). Temperature and humidity are included to account for the influence of weather conditions on energy consumption. The proposed model is deployed to forecast building energy consumption both before (November 2019) and during (May – October 2020) the COVID-19 pandemic and is compared with a single machine learning model. The results demonstrate that the EL model outperforms the SVR and RF methods, providing excellent prediction accuracy even during the pandemic when significant changes in energy consumption patterns occurred. The findings also highlight the effectiveness of sliding windows as input features for improving model adaptability. Additionally, the analysis reveals that temperature is more prominent than humidity for improving prediction accuracy.
Co-Authors Abdillah Muttaqin Achmad Rifdatul Hisan Adityo Sandhy Putra Adwin Welmark Cristri Ahmad Faridh Ahmad Marjan Ahmad Qurthobi Ahmad Ramdono Anas Rafii Ramadhan Andi Aditya Pratama Andica Dian Isnaini Andrianto, Deny Anggita Bayu Krisna Pambudi Aprilia Susanti Apriliani Putri Prasetyo Arief Budi Nugraha Ary Halimawan Ary Pranajaya Asep Suhendi Auditio Mandhany Aulia Laila Fithri Awaludin, Rahmat Ayu B. Artini Ayu, Christina Putri Bella Pratiwi Binandika Arya Wangsa Chartika Fadilah Prasetiani Christina Christina Danny Dimas Debriano Dewa Buana Muharmadin Dewi Novianti Diah Ayu Sitoresmi Dinda Rizqi Ridha Debiyani Dudi Darmawan Edi Leksono Edi Leksono Endang Rosdiana Erik Deardo Purba Ery Djunaedy Fachri Devanika Fadhli Rahman Fadhly Aziz Putra Pradana Fajri Amenda Putra Falery Widyawan Fani Putri Utami Fathir Endrawan Fathonah, Indra Wahyudin Febriansah Setiawan Firdy Bahari Putra Fransiskus Xaverius Dwiyanto Meas Friansa, Koko Guntur Bhatara Sutra Hanafi Rusmantika Handre Kertha Utama, Putu Hertiana Bethaningtyas Dyah K I Putu Arya Suarsana Ibrahim Fahri Fuady Ifta Rifqi Tyesadha Ihsan Adhi Nugroho Imam Maulana Taufan Indah Wardati Indra Wahyudin Fathonah Indrapraja, Rachmadi Irsyad Nashirul Haq Irsyad Nashirul Haq Isman Kurniawan Ismudiati Puri H Ismudiati Puri Handayani Jaya Wikrama Justin Pradipta Justin Pradipta Justin Pradipta Koko Friansa Liman Hartawan Lukman Nugraha M An Naas M S M Saladin Prawirasasra Mahesa Nanda, Rezky Maidasari Br Manurung Mamat Rokhmat Maulana, Baharuddin Nur Melina Melina Miranti Widyastuti Mochammad Iqbal Bayeqi Mohammad Rakha Farizi Muhamad Ramdlan Kirom Muhammad Aslam Muhammad Fatih Hasan Muhammad Pandu Aryo Gumilang Muhammad Wawan Kurniawan Muhammad Zakaria Muhammad Zaky Mubarok, Muhammad Zaky Nanda Aulan Tamami Nashirul Haq, Irsyad Nina Retna Utami Nirwana Prasetia Sipayung Nissa, Nabilah Khoirun Novian Lisdi Wahyoto Novita Fatkhuromah Nurwulan Fitriyanti Octo Emerald Siregar Paramitha Octavia Pristian Firzatama Putri, Vika Ristiana Putu Handre Kertha Utama Putu Handre Kertha Utama Raden Rizki Mulia Putra Rahmat Awaludin Salam Rasmid Rasmid Rayhan Humamuddin Reksa Putra Restianim, Vivien Rezky Mahesa Nanda Rizkar Febrian Rizky Anggara Robi Sobirin Suwandi Suwandi T. Nuzul Akbar Tri Ayodha Ajiwiguna Valentina Adimurti Kusumaningtyas Ventiano Ventiano Yanky Newalse