An Improved Particle Swarm Optimization Approach for Unit Commit- ment Problem

Unit commitment problem is a large scale nonlinear hybrid integer programming problem. Optimal unit commitment scheduling involves determining on/off states of units and determining generations of units. This paper proposes an improved particle swarm optimization (IPSO) for the solution of optimal unit commitment problem (UCP). In the proposed approach, the on/off states of units are limited into feasible schedules by providing a new method related to a new time order at first. After that, the problem is transformed into a simple economic load dispatch problem. Then this dispatch problem is solved by an improved priority list technique instead of the classical equal lambda-iteration method. All the above improvements are embedding into the framework of particle swarm optimization approach for UCP. It is seen from the results of numerical example that the proposed IPSO approach surely possesses a high quality and a high convergence speed.