A Model Predictive Control Solution on Wind Farms incorporated with Dual Battery Energy Storage Systems

This paper investigates the application of Model Predictive Control to Dual Battery Energy Storage System so that the reference power from the actual wind farm power can be tracked satisfactorily. Operation of the two batteries is categorized into two modes. In Mode 1, the first battery charges and the second battery discharges; in Mode 2, things function in the opposite way. The power from the battery is treated as the control signal which is an optimized computation results by MPC considering certain practical constraints including the power delivered/extracted from each battery, as well as the state of charge on each battery. The significance of this work contains mainly three aspects. Firstly, the two-mode operation of batteries removes the problem of overloading battery. Secondly, the actual wind power data is used to verify the dual battery integrated wind farm performance using MPC, thereby tracking the reference power by reducing the battery switching times and extending the life of the battery. Lastly, the constrained MPC is solved by following Hildreth’s quadratic programming procedure, for its reputation of providing simple and reliable real-time implementations.