This article investigates the combination of two model-predictive control concepts, i.e., sequential model-predictive control and long-horizon model-predictive control, for power electronics. To achieve sequential model-predictive control, the optimization problem is split into two subproblems. The first one summarizes all control goals, which linearly depend on the system inputs. Sequential model-predictive control generally requires to obtain more than one solution for the first subproblem. Due to the mixed-integer nature of finite control set model-predictive control power electronics, a special sphere decoder is, therefore, proposed in this article. The second subproblem consists of all those control goals that depend nonlinearly on the system inputs and is solved by an exhaustive search. The effectiveness of the proposed method is validated via numerical simulations at different scenarios on a three-level neutral-point-clamped permanent magnet synchronous generator wind turbine system and compared to other long-horizon model-predictive control methods.

Due to the need for highly efficient and compact power electronic converters to operate at higher frequencies, traditional wire-wound magnetics are not suitable. This paper provides a comprehensive review of planar magnetic technologies, discussing their advantages as well as associated disadvantages. An extensive review of the research literature is presented with the aim of suggesting models for planar magnetics. Several strategies are proposed to overcome the limitations of planar magnetics, including winding conduction loss, leakage inductance, and winding capacitance. The goal of this study is to provide engineers and researchers with a clear roadmap for designing planar magnetic devices.

Model predictive control has been widely developed for electrical drive systems. The weighting factors are key parameters affecting the control performance of the motor. This paper proposes a Bayesian ascent and Gaussian process method for the optimized weighting factors calculation. The root mean square error of the current in the d-q axis is taken as a criterion. A data-driven probabilistic proxy model is formed based on the Gaussian process, and the optimal weighting factor is obtained by Bayesian ascent. Grid sampling, random sampling, and sampling in the maximum possible region are integrated to improve sampling reliability. Furthermore, the maximum torque per ampere calibration and motor parameter identification can be implemented by the proposed method as well. Finally, a new perspective of global parameter optimization based on data probability is proposed. The effectiveness of the proposed methods is verified by experimental results.

We propose a novel framework for multistep predictive current control for induction machines and 3 Level-Neutral Point Clamped Inverter. To ensure low computational complexity The multistep prediction horizon is implemented using a sphere decoding algorithm. The DC-link balancing is guaranteed by adding a linearized DC-link balancing term. In order to obtain high accuracy, the linearization is performed during the online stage of the framework. Since the system is not differentiable we furthermore define a linearization framework based on the weak derivative. The framework is verified by means of simulations.

We present a novel model predictive control algorithm for three level neutral point clamped converter back-to-back PMSG wind turbine systems. In order to reduce the computational complexity we utilize a sequential approach. The wind turbine as well as the grid outputs are optimized within the outer loop. In order to reduce the computational effort a modified sphere decoder is applied. The best K solution candidates of the outer loop are then passed on to the inner loop. The DC-link balance is optimized in the inner loop. Since the number of solution candidates is comparably small for the inner loop, an exhaustive search is applied. The framework is verified by means of numerical simulations.

In this paper, a novel control method for back-to-back converters used in grid-to-motor connections is explored.To increase the robustnessoflow DC-link capacitances, a controlmethod based on variable switching point model predictive controlis proposed. While previous model predictive control methods forthe back-to-back converter selected a certain switching state tofulfill all control goals, we use the switching time in additionto the switching statein order to minimise deviations from thetarget voltage. Choosing a variable switching point provides anadditional degree of freedom to the control frameworkand allowsthe system to copewith the large number of control variables. Inthis case, the variable switching point is used to minimize theeffects of low DC-link capacitances on the system. This caneitherbeachieved by selecting a switching point that yields low DC-link capacitor charging or by selecting a switching point thataims to keep the DC-link voltage close to the reference. Theproposed method is verified through numerical simulations andhardware-in-the-loop (HIL) experiments and compared to existingapproaches. The results show that it is possible to control the DC-link using only the switching point of the converter.