Reinforcement Learning with Probabilistic Boolean Network Models of Smart Grid Devices

Figure 8

Maximum expected reward obtained for the failure operation mode of the IPR in one year of operation.