The primary purpose of this thesis is to propose a computer method for optimal operations planning of a thermal power-generating system.
Two major decisions must be made when scheduling the operations of a thermal power-generating system over a short time horizon.
The "Unit Commitment" decision indicates which generating unit combination is to be in use at each time period over the scheduling horizon. This decision must take into consideration the system capacity requirements and the economic costs of starting up or shutting down available units. The "Economic Dispatch" decision is the optimal allocation of the system load among the generating units in operation at time period. The optimal allocation of load among the operating units depends upon the relative efficiencies of the units. Optimal operations planning of the units requires the simultaneous consideration of unit commitment and economic dispatch decisions.
In this paper, we present a forward dynamic programming method for the short time horizon power-generating scheduling problem. Nonlinear programming is used to make an economic dispatch decision with quadratic objective function and linear constraints.
The objective of the model is to search the economic decision making at each time period through the given time horizon considering the start-up cost which depends on the amount of time the unit has been shut down prior to start-up with heuristic approach.
The model is applied for the Korean Electric Company. We present the computational results which indicate a considerable amount of fuel cost can be saved.