Multiobjective thermal power dispatch using opposition-based greedy heuristic search

This paper proposes an opposition-based greedy heuristic search (OGHS) strategy to solve multi-objective thermal power dispatch problem as a non-linear constrained optimization problem considering operating cost and pollutant emissions as competing objectives. The optimization problem is solved to find global solution, in case any one objective function is non-convex and non-differentiable. To generate initial population opposition-based learning is applied to select good candidates by exploring the search space extensively. Further, opposition-based learning is exploited for migration to maintain the diversity in the set of feasible solutions. Proposed method applies mutation strategy by perturbing the genes heuristically and seeking better one. This concept introduces parallelism and makes the algorithm always greedy for better solution. The greediness and randomness pulls the algorithm towards the global solution. The algorithm is also self sufficient without the need of tuning any parameter that effects acceleration of the algorithm. Fuzzy-theory is employed for decision-making that selects best solution from available non-inferior solutions. Feasible solution is also achieved heuristically that modifies the generation-schedule and avoids violation of operating generation limits. Proposed method has been implemented to analyze economic and multi-objective thermal power dispatch problems considering ramp-rate limits, prohibited-operating-zones, valve-point-loading effects, multiple-fuel options, environmental effects, and exact transmission losses encountered in realistic power system operation. The validity of proposed method is demonstrated on medium and large power systems. Proposed optimization technique is emerged out to compete with existing solution techniques. Wilcoxon signed-rank test for independent samples also proves the supremacy of proposed algorithm OGHS.