Jangkung Raharjo
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Jangkung Raharjo has a total of 75 co-authors. Here are the top co-authors by number of shared publications:
Kharisma Bani Adam
3 shared publicationsArief Goeritno
2 shared publications- Bandiyah Sri Aprillia
Telkom University,Dept. of Electrical Engineering,Bandung,Indonesia
2 shared publications - Hermagasantos Zein
Bandung State Polytechnic
2 shared publications Ignatius Riyadi Mardiyanto
Department of Energy Conversion, Bandung State of Polytechnic, Bandung, 40559, Indonesia
2 shared publications- Inung Wijayanto
School of Electrical Engineering, Telkom University, Bandung, Indonesia
2 shared publications
Latest Publications
- Open Access
- 1 January 2027
- Published by Tech Science Press in Energy Engineering
- Vol. 124 (1) , 1-10
Electricity Load Growth Projection Based on Customer Sector in Ermera District, Timor-Leste, Using Spatial Projection Method
This study presents an analysis of electricity load growth projections in Ermera District, Timor-Leste, using spatial projection methods. Accurate electricity load growth projection is a vital component of long-term power system planning, infrastructure investment, and resource allocation. In developing countries like Timor-Leste, increasing population and economic expansion drive a continuous rise in energy demand. This study develops an electricity load growth projection model based on specific customer sectors (domestic, commercial, industrial, social, and government) in the Ermera District, Timor-Leste, utilizing spatial projection methods integrated with Geographic Information Systems (GIS) and the Gompertz growth equation. Ermera District, characterized by its mountainous topography and prominent coffee production sector, currently faces electrification challenges with a regional electrification ratio of approximately 75%, falling below the national average of 83.4%. Historical data from 2021 to 2024 reveals a significant and synchronized expansion of the electricity customer base, surging from 3367 to 8532 customers. The domestic sector consistently represents the primary driver of total energy consumption, accounting for approximately 85% of the total customer base. Spatial analysis identifies the sub-districts of Ermera, Railaco, and Atsabe as the highest customer-density zones and future socio-economic hotspots. By employing a spatial projection method tuned to the regional context, the mathematical models successfully minimized validation deviations against recorded Electricidade de Timor-Leste (EDTL) data to a Mean Absolute Percentage Error (MAPE) threshold below 10%, highlighting the superior accuracy of spatial modeling over conventional methods in handling topographical variability. The Gompertz equation was applied to model long-term trends, projecting a rapid acceleration phase for the 2025–2030 planning cycle, with predicted cumulative demand escalating from 595,365 kW in 2025 to 1,120,610 kW by 2030. To evaluate network implications, peak demand scenarios were simulated using the Electrical Transient Analyzer Program (ETAP). The power flow simulations exposed recurrent technical limitations across the distribution network, particularly marginal under-voltage conditions (0.38–0.39 kV on the low-voltage side) on peripheral buses along the extended Ermera and Letefoho feeders, alongside transformer loading risks during peak hours. To mitigate these issues and optimize grid reliability, this thesis proposes critical engineering interventions, including the strategic activation of capacitor banks, distribution transformer tap changer optimizations, phase load balancing, and network reconductoring to lower line impedance. Ultimately, these findings offer a comprehensive, spatially explicit framework to support sustainable infrastructure expansion and energy policy formulation in the Ermera District.
- 1 January 2027
- Published by Tech Science Press in Energy Engineering
- Vol. 124 (1) , 1-10
Open AccessElectricity Load Growth Projection Based on Customer Sector in Ermera District, Timor-Leste, Using Spatial Projection Method
This study presents an analysis of electricity load growth projections in Ermera District, Timor-Leste, using spatial projection methods. Accurate electricity load growth projection is a vital component of long-term power system planning, infrastructure investment, and resource allocation. In developing countries like Timor-Leste, increasing population and economic expansion drive a continuous rise in energy demand. This study develops an electricity load growth projection model based on specific customer sectors (domestic, commercial, industrial, social, and government) in the Ermera District, Timor-Leste, utilizing spatial projection methods integrated with Geographic Information Systems (GIS) and the Gompertz growth equation. Ermera District, characterized by its mountainous topography and prominent coffee production sector, currently faces electrification challenges with a regional electrification ratio of approximately 75%, falling below the national average of 83.4%. Historical data from 2021 to 2024 reveals a significant and synchronized expansion of the electricity customer base, surging from 3367 to 8532 customers. The domestic sector consistently represents the primary driver of total energy consumption, accounting for approximately 85% of the total customer base. Spatial analysis identifies the sub-districts of Ermera, Railaco, and Atsabe as the highest customer-density zones and future socio-economic hotspots. By employing a spatial projection method tuned to the regional context, the mathematical models successfully minimized validation deviations against recorded Electricidade de Timor-Leste (EDTL) data to a Mean Absolute Percentage Error (MAPE) threshold below 10%, highlighting the superior accuracy of spatial modeling over conventional methods in handling topographical variability. The Gompertz equation was applied to model long-term trends, projecting a rapid acceleration phase for the 2025–2030 planning cycle, with predicted cumulative demand escalating from 595,365 kW in 2025 to 1,120,610 kW by 2030. To evaluate network implications, peak demand scenarios were simulated using the Electrical Transient Analyzer Program (ETAP). The power flow simulations exposed recurrent technical limitations across the distribution network, particularly marginal under-voltage conditions (0.38–0.39 kV on the low-voltage side) on peripheral buses along the extended Ermera and Letefoho feeders, alongside transformer loading risks during peak hours. To mitigate these issues and optimize grid reliability, this thesis proposes critical engineering interventions, including the strategic activation of capacitor banks, distribution transformer tap changer optimizations, phase load balancing, and network reconductoring to lower line impedance. Ultimately, these findings offer a comprehensive, spatially explicit framework to support sustainable infrastructure expansion and energy policy formulation in the Ermera District.
- Open Access
- 10 November 2026
- Published by AccScience Publishing in An International Journal of Optimization and Control: Theories & Applications (IJOCTA)
- p. 026400306
Structural feasibility handling in non-convex economic dispatch: Integrating fuzzy C-means clustering with a diagnostically-redesigned genetic algorithm
Practical thermal economic dispatch (ED) problems with valve-point effects (VPE), prohibited operating zones (POZ), ramp-rate limits, and unit commitment (UC) decisions are non-smooth, non-convex, and fragmented, which defeats gradient-based solvers. This paper presents a feasibility-aware computational intelligence framework—the Fuzzy C-Means Population-Based Genetic Algorithm (FCM-PbGA)—that embeds commitment-consistent encoding, FCM-guided initialisation, ramp-feasible bounds, POZ-aware projection, and Lamarckian refinement in a single-population evolutionary architecture. The framework is evaluated on the IEEE 15-unit, 24-hour benchmark (30 trials). The initial island-based architecture (Stage 1) achieves $662,173 ± $2,151. Budget-matched re-runs and ablation isolate architecture from budget: fixed six-island partitioning proves counterproductive, and the matched GA outperforms Stage 1. Three redesigns follow: adaptive merging, reheating, POZaware crossover and fractional seeding close 49% of the gap; corrected operators close an additional 36%; and removing islands entirely (v4) reaches $655,073.62 ± $1,469—a statistical tie with the matched GA (p = 0.199, d = −0.34) under published conditions, and a significant lead (p = 5.2 × 10−6, d = −1.34) with correctly loaded POZ sets and strict repair. Zero power-balance, ramp-rate, POZ, and generation-limit violations are verified over 10,800 unit-hours. The study demonstrates a diagnose-then-redesign methodology for constrained engineering scheduling.
- 10 November 2026
- Published by AccScience Publishing in An International Journal of Optimization and Control: Theories & Applications (IJOCTA)
- p. 026400306
Open AccessStructural feasibility handling in non-convex economic dispatch: Integrating fuzzy C-means clustering with a diagnostically-redesigned genetic algorithm
Practical thermal economic dispatch (ED) problems with valve-point effects (VPE), prohibited operating zones (POZ), ramp-rate limits, and unit commitment (UC) decisions are non-smooth, non-convex, and fragmented, which defeats gradient-based solvers. This paper presents a feasibility-aware computational intelligence framework—the Fuzzy C-Means Population-Based Genetic Algorithm (FCM-PbGA)—that embeds commitment-consistent encoding, FCM-guided initialisation, ramp-feasible bounds, POZ-aware projection, and Lamarckian refinement in a single-population evolutionary architecture. The framework is evaluated on the IEEE 15-unit, 24-hour benchmark (30 trials). The initial island-based architecture (Stage 1) achieves $662,173 ± $2,151. Budget-matched re-runs and ablation isolate architecture from budget: fixed six-island partitioning proves counterproductive, and the matched GA outperforms Stage 1. Three redesigns follow: adaptive merging, reheating, POZaware crossover and fractional seeding close 49% of the gap; corrected operators close an additional 36%; and removing islands entirely (v4) reaches $655,073.62 ± $1,469—a statistical tie with the matched GA (p = 0.199, d = −0.34) under published conditions, and a significant lead (p = 5.2 × 10−6, d = −1.34) with correctly loaded POZ sets and strict repair. Zero power-balance, ramp-rate, POZ, and generation-limit violations are verified over 10,800 unit-hours. The study demonstrates a diagnose-then-redesign methodology for constrained engineering scheduling.
- Open Access
- 28 September 2026
- Published by Elsevier in Atmospheric Environment: X
- Vol. 32
Projecting Indonesia’s Coal-Fired Power Plant Aerosol emissions to 2060: A modified Monte Carlo Framework using Himawari-8 satellite data
- 28 September 2026
- Published by Elsevier in Atmospheric Environment: X
- Vol. 32
Open AccessProjecting Indonesia’s Coal-Fired Power Plant Aerosol emissions to 2060: A modified Monte Carlo Framework using Himawari-8 satellite data
- 15 September 2026
- Published by Springer Nature in Environmental Monitoring and Assessment
- Vol. 198 (10) , 1074
Predictive modelling of fine-mode combustion aerosol events near Indonesian coal-fired power plants using AE-filtered Himawari-8 satellite data
- 15 September 2026
- Published by Springer Nature in Environmental Monitoring and Assessment
- Vol. 198 (10) , 1074
Predictive modelling of fine-mode combustion aerosol events near Indonesian coal-fired power plants using AE-filtered Himawari-8 satellite data
- Open Access
- 19 August 2026
- Published by Elsevier in SSRN Electronic Journal
Physics Guided Learning for Real World Battery Health Estimation in Electric Motorcycles
Lithium-ion batteries are central to electric vehicles, and tracking their state of health (SOH) is essential for safe, reliable operation over their lifetime. This work introduces a physics-informed temporal convolutional network (PI-TCN) for SOH estimation of a 72V/20Ah LiCoO2 motorcycle battery pack, using real-world data collected in Indonesia. A purely data-driven TCN is limited to whatever patterns appear in its training set, so it can produce predictions that look locally accurate but drift into physically implausible territory, especially in noisy or data-sparse regions. PI-TCN addresses this by building an exponential capacity fade prior directly into training, paired with a monotone regularization term that keeps the degradation trend physically consistent. Its dilated causal convolutions also capture temporal patterns within and across charging cycles, something conventional machine learning models cannot do. We tested the model under an interleaved protocol, using every fifth cycle from cycle 2 to 700 to cover the full degradation range. PI-TCN reached a mean R2 of 0.9997, an MAE of 0.065%, an RMSE of 0.092%, and a maximum error of 0.468%, beating all four baselines (KNN, extra trees, random forest, and XGBoost) on every metric by avoiding the implausible errors those models are otherwise free to make.
- 19 August 2026
- Published by Elsevier in SSRN Electronic Journal
Open AccessPhysics Guided Learning for Real World Battery Health Estimation in Electric Motorcycles
Lithium-ion batteries are central to electric vehicles, and tracking their state of health (SOH) is essential for safe, reliable operation over their lifetime. This work introduces a physics-informed temporal convolutional network (PI-TCN) for SOH estimation of a 72V/20Ah LiCoO2 motorcycle battery pack, using real-world data collected in Indonesia. A purely data-driven TCN is limited to whatever patterns appear in its training set, so it can produce predictions that look locally accurate but drift into physically implausible territory, especially in noisy or data-sparse regions. PI-TCN addresses this by building an exponential capacity fade prior directly into training, paired with a monotone regularization term that keeps the degradation trend physically consistent. Its dilated causal convolutions also capture temporal patterns within and across charging cycles, something conventional machine learning models cannot do. We tested the model under an interleaved protocol, using every fifth cycle from cycle 2 to 700 to cover the full degradation range. PI-TCN reached a mean R2 of 0.9997, an MAE of 0.065%, an RMSE of 0.092%, and a maximum error of 0.468%, beating all four baselines (KNN, extra trees, random forest, and XGBoost) on every metric by avoiding the implausible errors those models are otherwise free to make.
- Open Access
- 7 August 2026
- Published by Springer Nature in Discover Internet of Things
- Vol. 6 (1) , 143
Real-time IoT monitoring of soil parameters and monthly machine learning-based prediction of soil organic carbon in an urban forest
Soil organic carbon (SOC) is central to soil function and to the global carbon cycle, yet most assessments still depend on laboratory analysis that can only be repeated a few times a year. We describe an Internet of Things (IoT) and machine learning framework that estimates SOC dynamics indirectly, through soft sensing of soil parameters that are measured continuously in a campus urban forest. The sensors do not measure SOC itself; they record soil moisture, pH, electrical conductivity (EC), temperature, and the macronutrient indicators nitrogen, phosphorus, and potassium (N, P, K) in the Telkom University Forest, Bandung, Indonesia. Monthly laboratory SOC measurements, collected from January to August 2025, provided the reference values used to calibrate and test the models. Raw readings were filtered against physical ranges and screened for outliers before being averaged to monthly values, which then served as model inputs. Of the algorithms we compared, Random Forest gave the best result, with an R² of 0.829, a mean absolute error (MAE) of 0.221, and a root mean square error (RMSE) of 0.363 for monthly SOC. While the model demonstrates moderate predictive accuracy, it robustly captures the seasonal dynamics of SOC, aligning with laboratory-validated trends. We therefore present the framework as a tool for early monitoring and trend detection that complements laboratory analysis instead of replacing it, and that brings soil-carbon assessment to a temporal resolution conventional sampling cannot reach.
- 7 August 2026
- Published by Springer Nature in Discover Internet of Things
- Vol. 6 (1) , 143
Open AccessReal-time IoT monitoring of soil parameters and monthly machine learning-based prediction of soil organic carbon in an urban forest
Soil organic carbon (SOC) is central to soil function and to the global carbon cycle, yet most assessments still depend on laboratory analysis that can only be repeated a few times a year. We describe an Internet of Things (IoT) and machine learning framework that estimates SOC dynamics indirectly, through soft sensing of soil parameters that are measured continuously in a campus urban forest. The sensors do not measure SOC itself; they record soil moisture, pH, electrical conductivity (EC), temperature, and the macronutrient indicators nitrogen, phosphorus, and potassium (N, P, K) in the Telkom University Forest, Bandung, Indonesia. Monthly laboratory SOC measurements, collected from January to August 2025, provided the reference values used to calibrate and test the models. Raw readings were filtered against physical ranges and screened for outliers before being averaged to monthly values, which then served as model inputs. Of the algorithms we compared, Random Forest gave the best result, with an R² of 0.829, a mean absolute error (MAE) of 0.221, and a root mean square error (RMSE) of 0.363 for monthly SOC. While the model demonstrates moderate predictive accuracy, it robustly captures the seasonal dynamics of SOC, aligning with laboratory-validated trends. We therefore present the framework as a tool for early monitoring and trend detection that complements laboratory analysis instead of replacing it, and that brings soil-carbon assessment to a temporal resolution conventional sampling cannot reach.
- Open Access
- 5 June 2026
- Published by Universe Scientific Publishing Pte. Ltd. in Clean Energy Science and Technology
Enhanced generator scheduling through the lambda iteration method with clone-based optimization and Lévy flight integration
One of the hardest parts of running a power system is economic dispatch (ED), which means finding the cheapest way to meet demand and operational limits. To alleviate these limitations, this study presents a hybrid methodology known as Vectorized Lambda Iteration Method–Clone-Based Optimization with Lévy flight (VLIM-CBO-LF). The contribution comprises the synchronized amalgamation of three complementary strategies. The VLIM-CBO-LF improves the accuracy of dispatch calculations, clone-based optimization improves local search by using adaptive cloning, selection, and Lévy flight improves global exploration by letting movements happen over long distances from time to time. The suggested method was tested on a 42-unit economic dispatch system over a 24-h period, running it 30 times to make sure it worked each time. The results show that VLIM-CBO-LF has the lowest average daily fuel cost of IDR 3.9096 × 1010. This is better than the artificial bee colony, particle swarm optimization, grey wolf optimization, and whale optimization algorithm. The cost goes down from IDR 1.050 × 10⁹ to 3.666 × 10⁹ per day, but the performance stays the same, with a standard deviation of 6.50 × 10⁷ IDR. The ANOVA test (F = 19.004, p = 1.820 × 10⁻⁷) and the Friedman test (χ2 = 60, p = 9.358 × 10⁻14) both show that the results are statistically significant. The Wilcoxon signed-rank test (W⁺ = 0) shows that the results are strictly better than all other benchmark methods. Future research may investigate the applicability of the VLIM-CBO-LF model to more complex electrical market mechanisms.
- 5 June 2026
- Published by Universe Scientific Publishing Pte. Ltd. in Clean Energy Science and Technology
Open AccessEnhanced generator scheduling through the lambda iteration method with clone-based optimization and Lévy flight integration
One of the hardest parts of running a power system is economic dispatch (ED), which means finding the cheapest way to meet demand and operational limits. To alleviate these limitations, this study presents a hybrid methodology known as Vectorized Lambda Iteration Method–Clone-Based Optimization with Lévy flight (VLIM-CBO-LF). The contribution comprises the synchronized amalgamation of three complementary strategies. The VLIM-CBO-LF improves the accuracy of dispatch calculations, clone-based optimization improves local search by using adaptive cloning, selection, and Lévy flight improves global exploration by letting movements happen over long distances from time to time. The suggested method was tested on a 42-unit economic dispatch system over a 24-h period, running it 30 times to make sure it worked each time. The results show that VLIM-CBO-LF has the lowest average daily fuel cost of IDR 3.9096 × 1010. This is better than the artificial bee colony, particle swarm optimization, grey wolf optimization, and whale optimization algorithm. The cost goes down from IDR 1.050 × 10⁹ to 3.666 × 10⁹ per day, but the performance stays the same, with a standard deviation of 6.50 × 10⁷ IDR. The ANOVA test (F = 19.004, p = 1.820 × 10⁻⁷) and the Friedman test (χ2 = 60, p = 9.358 × 10⁻14) both show that the results are statistically significant. The Wilcoxon signed-rank test (W⁺ = 0) shows that the results are strictly better than all other benchmark methods. Future research may investigate the applicability of the VLIM-CBO-LF model to more complex electrical market mechanisms.
- Open Access
- 1 March 2026
- Published by IOP Publishing in IOP Conference Series: Earth and Environmental Science
- Vol. 1598 (1)
Analysis of Potential Alkaline Water Produced by Electrolysis Method for Agricultural Fertilizers Using Solar Cell Modules
The growing population and rising food demand have driven the agricultural sector to seek sustainable alternatives to chemical fertilizers. In Indonesia, this need aligns with the Minister of Agriculture Regulation No. 64 of 2013, which promotes environmentally friendly farming. Previous studies on ionized water demonstrated its potential to improve plant growth and soil properties, but limited research has explored its integration with renewable energy for agricultural use. This study investigates the production of alkaline water through electrolysis powered by a 20 Wp solar cell module using five test solutions: raw water, NaCl, KCl, KIO 3 , and NH 4 Cl. A Portable Water Ionizer (PWI) was used to analyze pH, TDS, energy performance, and soil response. NaCl yielded the highest pH (9.26) and energy output (24.05 Wh), while alkaline irrigation improved soil pH up to 7.0 without affecting moisture (maintained at 58–60%). In contrast, NH 4 Cl produced strong acidification (pH 3.48). These results demonstrate that electrolyzed NaCl and KCl solutions can serve as low-cost, solar-powered alternatives to improve soil conditions and reduce chemical fertilizer dependency in tropical agriculture like Indonesia.
- 1 March 2026
- Published by IOP Publishing in IOP Conference Series: Earth and Environmental Science
- Vol. 1598 (1)
Open AccessAnalysis of Potential Alkaline Water Produced by Electrolysis Method for Agricultural Fertilizers Using Solar Cell Modules
The growing population and rising food demand have driven the agricultural sector to seek sustainable alternatives to chemical fertilizers. In Indonesia, this need aligns with the Minister of Agriculture Regulation No. 64 of 2013, which promotes environmentally friendly farming. Previous studies on ionized water demonstrated its potential to improve plant growth and soil properties, but limited research has explored its integration with renewable energy for agricultural use. This study investigates the production of alkaline water through electrolysis powered by a 20 Wp solar cell module using five test solutions: raw water, NaCl, KCl, KIO 3 , and NH 4 Cl. A Portable Water Ionizer (PWI) was used to analyze pH, TDS, energy performance, and soil response. NaCl yielded the highest pH (9.26) and energy output (24.05 Wh), while alkaline irrigation improved soil pH up to 7.0 without affecting moisture (maintained at 58–60%). In contrast, NH 4 Cl produced strong acidification (pH 3.48). These results demonstrate that electrolyzed NaCl and KCl solutions can serve as low-cost, solar-powered alternatives to improve soil conditions and reduce chemical fertilizer dependency in tropical agriculture like Indonesia.