Clustering based unit commitment with wind power uncertainty

• Scenario generation and reduction techniques to simulate wind uncertainty.
• New approach is developed for creating clusters of unit status.
• Improved particle swarm optimization for accelerating optimization process.
• Cases include 10 thermal, 2 hydro, 1 pump storage unit and large scale wind farm.

Wind power generation is continuously increasing around the world, but due to uncertainty in wind power generation, the unit commitment problem has become complex. In this paper, scenario generation and reduction techniques are used to consider wind power uncertainty on system operation. Also, a new approach is developed for creating clusters of unit status associated with a probability of occurrence from an initial set of large wind power generation scenarios. And then a model of wind-hydro-thermal coordination problem along with the pumped storage plant is established. Combination of proposed weighted-improved crazy particle swarm optimization along with a pseudo code based algorithm and scenario analysis method is utilized to solve above problem. The effectiveness and feasibility of the proposed method is tested on systems with and without pumped storage plant integration. The results are analyzed in detail, which demonstrate the model and the proposed method is practicable in solving the unit commitment.