A Self-Adapted Across Neighborhood Search Algorithm With Variable Reduction Strategy for Solving Non-Convex Static and Dynamic Economic Dispatch Problems

The economic dispatch problem is a kind of challenging non-convex problem, which minimizes the total operating cost while being subject to a collection of complex equality and inequality constraints. This paper presents a novel meta-heuristic named across neighborhood search (ANS) algorithm to solve both dynamic and static economic dispatch problems. The ANS algorithm is augmented by a solution-difference disturbance mechanism and a parameter self-adaptation strategy. It is generally hard for metaheuristics to handle complex nonlinear equality constraints, because a meta-heuristic's search behavior is essentially stochastic while the equality constraints require the algorithm to exactly locate feasible solutions at the constraint bound. Therefore, a variable reduction strategy (VRS) is employed to deal with the equality constraint when solving the economic dispatch problem. VRS eliminates the equality constraint and reduces the dimensionality of the problem simultaneously, such that significantly improves the optimization efficiency. Extensive experiments and comparisons suggest that the proposed algorithm could generate the state-of-the-art results for both static and dynamic economic dispatch problems.