Real-time energy storage management with renewable energy of arbitrary generation dynamics

We design a real-time control policy for energy storage management with renewable energy integration, where the reviewable energy has arbitrary sample path statistics and may be non-ergodic over time. We formulate the control optimization problem to minimize a long-term time averaged cost. By relaxation and reformulation, we apply the Lyapunov optimization technique to design the real-time control algorithm that jointly optimizes the decisions for storage from two energy sources and supply to the consumer. We bound the performance of the proposed algorithm to that under the optimal non-causal T-slot lookahead control policy. We provide a close-form solution to our control optimization which renders our policy implementation with minimum complexity.