Model predictive controller for battery management systems

Advances in battery storage technology and the integration of renewable energy sources offer new opportunities in electricity grid operations to reduce the cost to end users. The introduction of time-of-usage tariff plan enables users to store energy when its price is low and re-use it at other times. Such cost savings can further be improved by harvesting energy from renewable sources such as solar PV Panels and wind mills. While intuitively simple, this problem is challenging because of the uncertainty in the renewable sources; further, frequent charging and discharging battery storage system and/or operating them outside the rated manufacturing specifications will damage the battery cells. Therefore, to increase the life cycle of the battery storage system and maximize the cost savings, it is necessary to carefully schedule these battery actions. We formulate and solve a model predictive control problem to obtain an optimal battery scheduling policy that minimizes battery life cycle cost and grid cost.