Abstract
This paper presents a developed predictor formulation for performing the time-series power-flow simulation of balanced and unbalanced distribution systems. The developed technique uses the changes in reactive and active powers of loads during the 24 hr in order to predict the updated voltage magnitudes and phase angles, respectively. The Lagrange linear interpolation is used as a predictor approach in the developed formulation. The developed technique is compared with the traditional technique that uses the previous solution as predictor values for the next solution. The conventional and developed techniques are implemented into forward/backward sweep for the three-phase radial distribution power-flow algorithm. The iterations number and computation time of the power-flow algorithm are significantly reduced in the case of using the developed technique compared with the previous solution technique. To prove its effectiveness, the developed technique is adopted for IEEE test systems, such as the radial balanced 69-node and unbalanced IEEE 123-node feeder. All calculations are performed using C++ programming.