دانلود رایگان مقاله انگلیسی طراحی کنترل پیش بینی مدل برای مبدل های DC-DC کاربردی تا یک سیستم فتوولتائیک - الزویر 2018

ABSTRACT

A continuous control set model predictive control (CCS-MPC) is designed for a DC-DC buck converter used in maximum power point tracking (MPPT) of a photovoltaic (PV) module. A modified incremental conductance (mINC) algorithm is used for MPP determination as a reference signal for CCS-MPC. The small-signal model of the PV system is adaptively obtained around MPP through linearization of its average model. The predictive control is designed and applied to a PV system using an online optimization of the cost function including the discretized present and future states. The performance of the proposed m-INC CCS-MPC is evaluated by simulation study that indicates better performance in terms of transient and disturbance rejection compared to conventional PI controller. Finally, the applicability of the proposed m-INC CCS-MPC strategy is assessed with outdoor experimental results and the associated practical advantages against finite control set (FCS) MPC are discussed.

6. Conclusion

A CCS-MPC has been designed for a solar energy conversion system in order to enhance the efficiency of INC method for MPPT. The proposed controller has been applied on a DC-DC buck converter to charge up the batteries from a PV panel. Moreover, the INC algorithm has been modified to comply with the proposed controller requirements. The closed loop performance has been investigated through simulation study to compare the proposed controller with conventional PI. The simulation results show better transient dynamic and disturbance rejection performances for CCS-MPC which led to harvest higher amount of available photovoltaic power. The practical implementation of the introduced algorithm illustrated its performance for MPPT and controlling the delivered power to the energy storage system. The practical issues have been discussed and superior properties of the CCS-MPC explained against the previously reported predictive methods. It is concluded that using such technique is beneficial to have a constant switching frequency, which is mandatory in filter design. Moreover, it could increase the performance of the active switches in comparison with a FCS-MPC algorithm with varying and high switching frequency.