The objectives of this PhD study are fourfold: first, to ascertain the gaps between the dynamic performance, scalability, and design efficiency of existing  DC-AC energy conversion systems and those required for the decentralized and digitized power grids; second, to find a novel solution to fill the gaps; third, to assess the feasibility of the solution for industrial applications; fourth, to conduct the quantitative comparison study between the solution and the legacy technologies. This thesis is written in depth and succinctly on the methodologies and the demonstrative results that achieve the above four objectives. The solution development was carried out through a two-layer study comprising component- and system-level research. The solution includes 1) three new multilevel converter designs utilizing the binary numeral system-based capacitor voltage control and multilevel AC-voltage synthesis and 2) a new model predictive current control (MPCC) scheme exploiting the quadratic integration method. The new converter designs and control scheme effectively enhance the DC-AC power conversion quality and efficiency and the closed-loop control performance of MPCC-based multilevel converters in the aforementioned power grids. The extensive system design and study of a novel medium-voltage-direct-current (MVDC) electric power transmission system utilizing the above novel solution was conducted to assess the feasibility of the solution for industrial applications. Compared to the legacy technologies, this novel system reduces the hardware complexity by 52%, improves the DC-AC energy conversion efficiency by 47.4%, and improves the harmonic performance by 55.47%. The quantitative comparison study focused on validating the solution's novelty and practicality according to the doctoral research proposal addendum.