A New Fast Formulation of Model Predictive Control For CHB STATCOM

Recently, the finite control set model predictive control (FCS-MPC) has obtained a lot of attention for power converter control due to its advantages of high dynamic performance and multi-objective capability. However, it poses a challenge for the current processors directly applying this control to the multilevel inverter in real-time application. Especially for CHB-STATCOM multilevel inverter, there are a huge number of switching combinations and redundancies due to its topology structure. Real-time searching for the optimal switching state among the extremely large candidate pool through exhaustive search algorithm is almost impossible especially when the inverter output voltage levels increases. To end this problem, this paper has presented a new fast FCS-MPC formulation scheme based multilevel CHB-STATCOM. Instead of resorting to existing heuristic optimization algorithms, the FCS-MPC is reformulated mathematically to a matrix problem that can be easily solved on-line. The proposed single step FCS-MPC formulation is validated through simulation based on a seven levels CHB-STATCOM. It has been shown that the proposed single step MPC has the advantages of fast current tracking, minimum CMVphase-shiftedand good voltage balancing capability. Compared with the existing FCS-MPC schemes, the computational burden of the proposed MPC formulation is largely reduced which makes it more suitable for multilevel CHB-STATCOM. Moreover, the proposed FCS-MPC formulation can be easily developed and applied to other multilevel topologies with a large number of voltage levels and redundancy.