Optimization of Unit Commitment Problem and Constrained Emission using Genetic Algorithm

Abstract — This paper presents a solution to Optimal Unit commitment (UC) of thermal units based on a evolutionary algorithm known as Genetic algorithm (GA). In this proposed method of GA for the UC problem, the scheduling variables are coded as integers, so that the minimum up/down time constraints can be handled directly. To verify the performance of the proposed algorithm, it is applied to systems with 10 generating units in one-day scheduling period. The test results reveal that not only does the GA consider the constraints very well, but also minimizes the operating cost and emission cost. It can also find solution very close to optimum value within a reasonable time. It is shown that the algorithm is capable to find better solutions in comparison with the conventional methods and most of the other computing techniques. The results also prove the efficacy and correctness of the method.