Model Predictive Control Via PV-Based VAR Scheme for Power Distribution Systems With Regular and Unexpected Abnormal Loads

摘要： This paper develops a model predictive controller (MPC) via photovoltaic (PV)-based volt-ampere reactive scheme to minimize the power loss and stabilize voltage ?uctuation when PV cells are connected to the power distribution line. The nominal power load data from California independent system operator is used to simulate dynamics of the system with DistFlow equations. Since power consumptions in fact may deviate from nominal values, an estimator is further developed to reconstruct the state variables and power loads from measurements. The integration of the MPC and estimator forms a closed-loop control framework and enables the system to quickly recover from undesired disturbances by effectively changing the real and reactive powers provided by PV cells. We use a bidirectional, single branch distribution circuit to demonstrate the performance of proposed scheme. The results show that our MPC indeed reduces power loss and keeps the voltage within a desired bound. Additionally, the estimator successfully detects and correctly estimates the abnormal change of power consumption and directs the MPC to compensate such unexpected power loads promptly.