The need for the usage of signal processing and pattern recognition techniques to monitor photovoltaic (PV) arrays and to detect and respond to faults with minimal human involvement is increasing. The data obtained from the array can be used to dynamically modify the array topology and improve array power output. This is beneficial especially when module mismatches such as shading, soiling and aging occur in the PV array. A robust statistics-based fault detection algorithm to find faulty modules is presented. Further, topology optimization of PV arrays using module level data is considered. Various topologies such as the series-parallel (SP), the total cross-tied (TCT), the bridge link (BL) and their bypassed versions are considered. The performance associated with these topologies for a possible shading pattern is analyzed and a topology reconfiguration algorithm is employed to find an optimal configuration. The results demonstrate the benefit of having an electrically re-configurable array topology. Results were generated in a SPICE simulator using synthetic and real data obtained from the APS experimental PV array facility.