Abstract
Ant colony system (ACS) model is more suitable for solving combinatorial optimization problem, so ACS has been applied to the hard combinatorial Unit commitment problem (UCP). Here, a parallel can be drawn of ants finding the shortest path from source (nest) to its destination (food) and solving UCP to obtain the minimum cost path (MCP) for scheduling of thermal units for the demand forecasted. Multi-stage decisions give ant search a competitive edge over other conventional approaches like dynamic programming (DP) and branch and bound (BB) integer programming techniques. Before the artificial ants starts finding the MCP, all possible combination of states satisfying the load demand with spinning reserve constraint are selected for complete scheduling period which is called as the ant search space (ASS). Then the artificial ants are allowed to explore the MCP in this search space. The proposed model has been demonstrated on a practical ten unit system and a brief study has been performed with respect to generation cost, solution time and parameter settings on a numerical example with four unit system.
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have