Abstract
Short-Term Economic Dispatch (STED) is essential for achieving cost-effective and secure operation of power systems, particularly as renewable generation and nonlinear operating characteristics increasingly complicate generation scheduling. This study develops an optimal STED scheme for the hydro-dominant Ugandan power system using the Invasive Weed Optimization (IWO) algorithm. A base-case model of the Ugandan transmission network was developed using generation and transmission data obtained from the Uganda Electricity Transmission Company Limited. An economic dispatch formulation incorporating power balance, generator operating limits, transmission constraints, transmission losses, and valve-point loading effects of thermal generating units was developed. The IWO algorithm was then employed to determine optimal generation schedules under peak (680 MW), shoulder (520 MW), and off-peak (400 MW) loading conditions. Its performance was evaluated against the base-case operation and two established metaheuristic techniques, Genetic Algorithm (GA) and Artificial Bee Colony (ABC). At peak loading, IWO reduced transmission losses from 57.12 MW to 43.30 MW (24.19%) and thermal fuel cost from UGX 22.51 × 10⁶/hr to UGX 16.56 × 10⁶/hr (26.43%). For shoulder loading, losses decreased by 22.33%, while fuel cost declined by 27.42%. Under off-peak conditions, losses and fuel cost were reduced by 25.34% and 26.02%, respectively. IWO consistently achieved lower losses and operating costs than GA and ABC across all loading conditions. The results demonstrate that the proposed IWO-based STED provides an effective and robust approach for economically scheduling Uganda’s hydrothermal generation fleet while satisfying operational constraints. The developed framework can support practical generation scheduling and economic operation of hydro-dominant power systems with similar characteristics.

Key words: Cost-function; economic load dispatch; Grey wolf optimization; Thermal plants; Valve-point effect