In this study, the problems of modeling, energy dispatching and Photovoltaic (PV) array energy priorities for a grid connected residential house with PV array and battery storage using model predictive control (MPC) have been investigated. Artificial neural network (ANN) based global solar radiation forecast was used to plan in advance for periods of low sunshine. MPC was able to reduce electricity consumption in the house when solar radiation forecast was unfavorable. Quadratic programming optimization was used to maximize usage of the PV system. Excess energy from the PV array was used to further raise hot water cylinder (HWC) temperature, rather than exporting it to the utility grid. Performance of the overall model predictive control system was verified using simulation results.