Unit commitment with non-smooth generation cost function using binary particle swarm optimization

This paper deals with generation unit scheduling, well known as unit commitment (UC). Rather than using quadratic generation cost function, this paper utilizes non-smooth generation cost function (NSGCF). As NSGCF is difficult to handle using conventional technique such as quadratic programming, the metaheuristic technique, particle swarm optimization in this case, is used to solve the economic dispatch which is part of UC. In this UC, minimum up time, minimum down time, start up cost and shunt down cost are considered. In addition, spinning reserve is taken into account. Because UC is mix-integer optimization problem, Binary Particle Swarm Optimization (BPSO) is applied to complete UC. To show the effectiveness of the approach, system with 6 generators is utilized.