Adaptive Agent-Based Control for Lithium-Ion Batteries in Naval Microgrids

Abstract This paper presents an adaptive agent-based control scheme for operating distributed energy storage in naval power systems. The adaptive agent-based control scheme enables batteries distributed throughout an electrical grid to autonomously communicate and align each other to operate in the most effective way possible. To achieve this, each battery is assigned an agent capable of controlling the battery’s output with the goal of maximizing the life of the battery while supplying the required power to the load. Along with the goals of the individual agents, system-level alignment strategies are employed to maximize certain aspects of the battery operations. The employed system-level alignment strategies are duration, ready, and aggressive. The duration strategy maximizes the longevity of grid resources of expected battery degradation and electric generator component wear by utilizing the system in a manner that optimizes system-level longevity. The ready strategy maximizes the amount of power being kept in reserve for use when needed while the Aggressive strategy maximizes power output by allowing the batteries to be used to their limits to deliver as much power as possible. Moreover, an adaptive strategy was also created by mapping the previously mentioned static strategies to the levels of power demand. This adaptive strategy allows the batteries to operate in the most efficient manner while adapting to the power demand from the load. To analyze these strategies, the chosen performance metrics were 1) battery and generator remaining useful life (RUL), 2) average battery temperature, and 3) percent of time the requested load profile was achieved. Results demonstrate that the adaptive strategy operates with better overal efficiency by maintaining the 2nd highest battery RUL, the 3rd highest generator RUL and the 2nd lowest average temperatures. It was also the only strategy to fulfill the maximum pulse seen from the load.