Dynamic economic dispatch (DED) problem is one of the optimization issues in power system operation. In this paper, an improved chaotic particle swarm optimization (ICPSO) algorithm is proposed to solve DED with value-point effects. In proposed ICPSO, chaotic mutation is embedded to overcome the drawback of premature in PSO. What's more, enhanced heuristic strategies are proposed to handling the various constraints of DED problem effectively. Comparing with penalty function method, the proposed constraints handling method can guide the population to feasible region without violating any constraints. Moreover, the effects of two crucial parameters on the performance of proposed ICPSO are also studied. Finally, the ICPSO algorithm is validated for two test systems consisting of 10 and extended 30 generators. While compared with other method reported in this literature, the experimental results demonstrated the high feasibility and effectiveness of proposed algorithm. (C) 2010 Elsevier Ltd. All rights reserved.