ANALYSIS AND PREDICTIVE CONTROL OF POWER DISTRIBUTION SYSTEM WITH HIGH PENETRATION OF PV RESOURCES

Abstract

Real-time control is projected as the future of the active distribution networks for hosting extensive integration of renewables and achieving its techno-economic operation. The thesis enhances the four major challenges in the distribution network control at high PV penetration, viz., (a) averting short-term voltage instability caused by severe PV ramps, (b) developing accurate linear model, and achieving (c) voltage control and (d) techno-economic operation, under uncertainty of PV generations and load demands. The thesis shows the first challenge can be resolved by replacing Induction Motor (IM) loads with a suitable variable-speed-drive penetration policy, as the IM load response is found to aggravate the instability. The linear model is enhanced with accurate and efficient computation of voltage sensitivity to tap-position of tap changers. Utilizing the developed linear model, the third and fourth problems are solved with robust constrained model predictive control. The proposed approaches are verified using DigSILENT PowerFactory.