In this paper, we propose a user centric real-time (RT) technique for an optimal control of a distributed photovoltaic stand-alone micro-grid. Our objective is to minimize an average aggregated system cost over a finite time period considering a joint optimization problem of battery energy storage (BES) management, load scheduling, and energy procurement process from a peaker generator (PG). Due to online decision making, the dynamic control actions of the state of energy of the storage battery are correlated over time. Thus, we impose a constraint on the BES state of energy over the considered time period and eliminate the finite BES capacity constraint. In the load scheduling process, we account for the temporal variability and spatial uncertainty of each household load explicitly. We introduce the concept of block duration to optimize the energy procurement cost from a PG. We modify and then transform the problem to utilize the Lyapunov optimization technique. We show that the proposed solution to the joint optimization problem is asymptotically optimal with a bounded performance guarantee and is easy to implement. Simulations in different weather conditions show the effectiveness of our proposed user centric RT solution in terms of the selected performance metrics.