A Heuristics-guided Evolutionary Approach to Multi-objective Generation Scheduling

A novel approach for multi-objective generation scheduling is presented. The work reported here employs a simple heuristics-guided evolutionary algorithm to generate solutions to this nonlinear constrained optimization problem where the objectives are mutually conflicting and equally important. The algorithm produces a cost-emission frontier of Pareto-optimal solutions, any of which can be selected based on the relative preference of the objectives. Within this framework, an efficient search algorithm has been developed to deal with the combinatorial explosion of the search space such that only feasible schedules are generated based on heuristics. This approach has been evaluated by successful experiments with three test systems containing 11, 19 and 40 generating units. Attaching importance to heuristics results in producing high-quality solutions in a reasonable time for this large-scale tightlyconstrained problem.