The scheduling of power system requires maintaining a continuous balance between supply and demand, a proper schedule for the generation of electricity will ensure the reliability of a power system and extend the operational lives of its constituent plants. Power system blackout was a big issue in 2011 E.C, there was a load shedding program all over the country for three months. Drought existed in Gibe III source areas, and water resource was limited to balance the load demand, because Ethiopian electric power does not used small power plants effectively. In this thesis, mathematical models were developed by analyzing different constraint like; power balance, spinning reserve, and capacity limits of generating unit’s constraints. The work is limited in power resource scheduling of Ethiopian power system generation in short-term impacts of the electricity production with time horizon from a day to a week (24 to 168 hours). Generations are scheduled for every day of the week based on the short - term forecasted load in wight moving average method. Power resource scheduling is done using particle swarm optimization (PSO) technique and simulate by MATLAB software. In this thesis, the committed generations decision is stated in database code. The small power plant wind, Repi, and hydro run of river produce maximum power at time t with respect to forecasted load, the generated power in hydro having dam power plants are reduced, the water resource in dams are more saved, the optimal PSO simulation result verifies this issue.