Short-term scheduling of thermal power systems

An algorithm to solve short-term thermal scheduling problems using the Lagrangian relaxation approach is presented. For individual subproblems, dynamic programming without discretizing generation levels is proved to be an efficient approach. The ramp rate constraint is handled through relaxation. This method preserves the advantages of nondiscretization of generation levels and is proved to be efficient for systems with a few ramp rate constrained units. Initialization of multipliers associated with system demand using the priority-list commitment and dispatch can significantly cut down the computational time. It is shown that the heuristic method developed to obtain feasible solutions is effective, and near optimal solutions are obtained.<>