A niching differential evolution algorithm for the large-scale combined heat and power economic dispatch problem

Abstract

The combined heat and power economic dispatch (CHPED) problem, which aims at minimizing system cost to supply system demand under various constraints, is one of the most challenging problems in the power system. Numerous researches are concentrated on solving the CHPED problem with small size, ignoring implementation of the large-scale CHPED problem with lots of local optimal solutions. This paper proposes a niching differential evolution algorithm (NDE), which incorporates niching methods into differential evolution algorithm (DE), to solve the large-scale CHPED problem. In NDE, a modified DE/best/1 strategy inspired by the neighborhood concept of niching methods, called DE/best/1/nh, is developed to generate diverse solutions. A two-phase selection that utilizes the deterministic crowding of niching methods and greedy selection is designed to remain promising solutions. The two strategies help to balance the global and local search capabilities of the algorithm. The performance of NDE is evaluated by a comparative analysis with other existing approaches on the five CHPED systems with 7, 24, 84, 96 and 192 units. The numerical results demonstrate the feasibility and effectiveness of NDE in solving the large-scale CHPED problem.