IP-C-PDL: A chaos-embedded trust-region interior-point framework for constrained economic dispatch in large-scale power systems
An International Journal of Optimization and Control: Theories & Applications (IJOCTA) · 13.11.2026 · 026310188
DOI: 10.36922/ijocta026310188 ↗Kurz gefasst
Fachartikel in der Zeitschrift An International Journal of Optimization and Control: Theories & Applications (IJOCTA), online erschienen am 13.11.2026. Verfasst von Jangkung Raharjo und 2 weiteren Autor:innen.
Abstract (Originalsprache)
Constrained economic dispatch (ED)—the classical, single-bus, active-poweronly dispatch problem for thermal generation, not an optimal power flow or unit-commitment problem—in large-scale interconnected power systems such as the Java–Madura–Bali (Jamali) network is non-convex and ramprate-coupled: deterministic solvers stall at local optima, while stochastic solvers incur unacceptable run-to-run variance. This paper proposes IP-C-PDL (Interior-Point Chaotic Powell’s Dog-Leg), which, within a single interior-point iteration, couples a Powell’s Dog-Leg trust-region step (Cauchy/quasi-Newton selection via accumulated SR1/BFGS curvature) with a stall-adaptive tentmap chaotic perturbation for ergodic basin escape; a multi-start pool, elite archive, and periodic SQP refinement complete the framework using only three user-specified parameters. On twelve CEC benchmark functions (30 runs, 12 methods), IP-C-PDL ranks 1 in 1 of 12 and top-3 in 7 of 12. On six ED scenarios spanning 5-, 16-, and 42-unit systems (50 runs, 10 methods, single-period and 24-hour horizons), it ranks 1 in 4 of 6 (mean rank 2.17/10) with zero constraint violations, yielding USD 22,239.29/day (σ = USD 0.41) on the 42-unit 24-hour case—0.16% better than the second-ranked comparator (t = −48.26, p < 10−42) and 4.67% better than the worst (t = −180.85, p < 10−70)—confirmed by Friedman, ANOVA, and Wilcoxon–Holm–Bonferroni tests. A five-factor sensitivity analysis identifies multi-start pool size as the dominant factor—with a cliff-edge improvement at just two starts—while the framework is insensitive to the remaining four hyperparameters and stabilises by 400 iterations. A policy sensitivity analysis further links dispatch quality to fiscal savings, Biaya Pokok Penyediaan (BPP) subsidy reduction, and avoided carbon emissions in the Indonesian power sector.
Zitieren
APA
Raharjo, J., Ikhsan, R. R. N., & Yustika, L. M. (2026). IP-C-PDL: A chaos-embedded trust-region interior-point framework for constrained economic dispatch in large-scale power systems. An International Journal of Optimization and Control: Theories & Applications (IJOCTA), 0(0), 026310188. https://doi.org/10.36922/ijocta026310188
Vancouver
Raharjo J, Ikhsan RRN, Yustika LM. IP-C-PDL: A chaos-embedded trust-region interior-point framework for constrained economic dispatch in large-scale power systems. An International Journal of Optimization and Control: Theories & Applications (IJOCTA). 2026;0(0):026310188. doi:10.36922/ijocta026310188
BibTeX
@article{Raharjo2026,
author = {Raharjo, Jangkung and Ikhsan, Rifki Rahman Nur and Yustika, Lindiasari Martha},
title = {IP-C-PDL: A chaos-embedded trust-region interior-point framework for constrained economic dispatch in large-scale power systems},
journal = {An International Journal of Optimization and Control: Theories & Applications (IJOCTA)},
year = {2026},
volume = {0},
number = {0},
pages = {026310188},
doi = {10.36922/ijocta026310188}
}
EUVIMED