Reinforcement Learning with Probabilistic Boolean Network Models of Smart Grid Devices

Figure 7

Maximum reward for normal operation, adjusted scale.