A framework for sensor driven condition based generator maintenance scheduling was proposed in Part I of this paper. In Part II, we extend the previous model by incorporating the unit commitment and dispatch into the optimal maintenance scheduling problem. We reformulate this extended maintenance scheduling problem as a two-stage mixed integer program. We use this reformulation to construct an algorithm that obtains the global optimal solution to the proposed generator maintenance problem. Finally, we test and analyze the proposed model through extensive experiments conducted on IEEE-118 bus system. For every experiment, we present a benchmark analysis against the maintenance models used in current industry practice and power systems literature. Experimental results indicate that the proposed maintenance schedules provide considerable improvements in both cost and reliability.