A Multiobjective Unit Commitment Problem: Minimization of Productions Costs and Gas Emissions

Given the increasing public awareness of environmental impacts, governments have made regulation on pollutants more stringent. Therefore, the Unit Commitment Problem (UCP), which traditionally minimizes the total production costs, needs to consider the pollutants emissions as another objective. This way, the UCP becomes a multiobjective problem with two competing objectives. The approach proposed to address this problem combines a Biased Random Key Genetic Algorithm (BRKGA) with a non-dominated sorting procedure. The BRKGA encodes solutions by using random keys, which are represented as vectors of real numbers in the unit interval. The non-dominated sorting procedure is then employed to approximate the set of Pareto solutions through an evolutionary optimization process. Computational experiments have been carried out on benchmark systems with 10 up to 100 generation units for a 24 hours scheduling horizon. The results obtained show the effectiveness and efficiency of the proposed BRKGA to find good solutions to the UCP. The diversity and welldistribution characteristics of the non-dominated solutions obtained are demonstrated. Furthermore, from the comparison with alternative multiobjective methods it is shown that the method proposed obtains better results in most cases. keywords: Unit Commitment, Genetic Algorithm, Multi-objective Optimization, Pareto Optimal Curves.