The roles of reactive power control in a distribution system become essential due to the high penetration of distributed generations (DG) these days. Proper reactive power control can reduce real power losses and regulate the voltage profile in a power system. However, intermittent characteristics of DGs (e.g., renewable energies from wind and solar power) impose uncertainty on power generation in the power system. Therefore, this paper presents a novel fast probabilistic power-flow (FPPF) method based on the Gram-Charlier series expansion to deal with such uncertainty. The FPPF method only deals with stochastic variations of random variables with respect to the expected values, thus reducing the number of iterations. Moreover, the chaotic particle swarm optimization is used to adjust generator voltages, transformer taps, and static compensators to minimize the real power losses while the stochastic voltages satisfy the operational limits. Applicability of the proposed method is verified through simulation using an autonomous 25-bus (Penghu) system and the IEEE 118-bus system. Comparative studies considering traditional probabilistic power-flow methods are performed as well.
- Particle swarm optimization (PSO)
- VAR control
- probabilistic power flow
- renewable energy