An Application of Artificial Intelligent Optimization Techniques to Dynamic Unit Commitment for the Western Area of Saudi Arabia

Unit commitment is the main challenging part of this paper. Operation data are collected form the Western Area of Saudi Arabia, however, the data are average. These data are manipulated to non-linear data that fit with unit commitment. Some missing constraints are imposed from standard IEEE data set. There are total one hundred and thirty five (135) units in the system for around 9,000 MW power demand of the area. So scalability of unit commitment is one of the main issues for this paper. Particle swarm optimization (PSO) method is applied to develop unit commitment system. Standard PSO sometimes does not converge for the one hundred and thirty five (135) units and other practical constraints. Some parts of PSO are modified (e.g., includes bacterial foraging operations) to converge the system. Besides, a repair method is applied to converge the system fast. In modern UC, both cost and emission are minimized, however, in typical UC, only cost is minimized. In this research, firstly cost, secondly emission and thirdly both cost and emission are minimized in the system. Thus the UC is generalized and can be applied to other system data and generates results depending on cost and emission coefficients.