Modeling and model predictive control of batteries to reduce a variability of load

Abstract

Model Predictive Control (MPC) has shown great success in different industrial applications, as it allows to have constraints on both, state and inputs. This ensures that processes will be run under tight performance specifications. The main goal of this project is to develop an MPC controller which will determine the optimal operation of the battery, i.e. when and how the battery should be charged or discharged, by controlling input current for the battery, to reduce the amount of the money to pay for the required energy. For this task, the battery model with aging effects, which is able to discharge and charge, will be also developed. Also, an MPC controller for efficient charging of the battery is developed, which allows calculating charging strategy, so without changing the charging time, the lifetime of the battery is prolonged. Source code 1 is publicly available and can be used for future work in other studies.