A Comparative Study of Unit Commitment Problem by Dynamic Programming and Genetic Algorithm

Unit commitment (UC) is a famous problem in electrical power systems which desire the minimization of total power generation cost at a specific period by determining acceptable scheduling of the generation units. Several methodologies and algorithms have been proposed to solve the UC problem. This research provides a blend of unit commitment problem implemented with genetic algorithm and the dynamic programming and also describes the best technique in solving the same. This problem involves an optimization task that involves schedule ON/OFF of generating units that satisfy various constraints for minimum cost generation of load demand at an hour. This problem improves network reliability and yields a technique to decrease the generation cost. The short-term unit commitment approach requires a quick technique to reduce system changes and scheduled errors. The definite solution for this problem can be achieved by the entire combination of all appropriate generating different combinations focusing on cost optimization. Genetic algorithms are stochastic search algorithms; they will maintain the population solutions to a problem in an encoded form so that the information will evolve in time. The optimal outcomes from the unit commitment problem for this technique are compared with the standard IEEE 10-unit system data.