Finite Control Set Model Predictive Control (FCS-MPC) can improve the control performance of Five-level Active NPC (5L-ANPC) inverter effectively. However, with the increasing of the levels of inverter, the traditional FCS-MPC method has many problems, such as large amount of rolling optimization calculation and difficulty in weight coefficient design. In order to solve those problems, an Optimal Model Predictive Control (O-MPC) method is proposed. First, an optimal control method based on level jump limitation is proposed to reduce the number of switch states. Second, in order to solve the conflict problem of multiple control objectives and reduce the design difficulty of weight factors, a satisfactory optimal control method based on hierarchy is proposed. After the optimization, the calculation time of the controller is greatly reduced and the control performance of the multi-objective control system will be significantly improved. At last, the effectiveness and feasibility of the proposed O-MPC are validated by the experimental results.

The five-level active neutral-point-clamped (5L-ANPC) converter has a great application potential in high power variable-frequency drives (VFDs) with high and medium voltage. However, the traditional method for capacitor voltage buildup is not applicable to the capacitors inside the back-to-back (B2B) 5L-ANPC converter. To solve the problem, the topological features of the B2B-5L-ANPC converter are fully analyzed, and a capacitor voltage pre-charge strategy for this converter is also proposed. With this strategy, all the capacitors in the converter can be charged by controlling the on and off states of switch tubes in the bridge arm, no additional hardware is needed. Moreover, the structural features of the voltage booster circuit were examined in four stages, and used to derive the capacitor voltage buildup formula in each stage. In addition, the relationship between voltage buildup rate, equivalent pre-charge resistor and capacity of capacitor is identified. Finally, Simulation results from a B2B-5L-ANPC simulation model which is constructed according to the actual system parameters and experimental results from a scaled-down prototype are provided to support the proposed control strategy.

Three‐phase AC–DC converter based on integrated transformer is a novel converter which is capable of handling high power levels and can realise power factor correction function. Owing to different parameter tolerances of the primary windings, grid current could be unbalanced, which usually could not be solved through traditional control methods. To handle this problem, a current balancing method is adopted in this study to improve former current control loop. Meanwhile, based on the principle of instantaneous power balance, a dual loop control strategy with an outer voltage loop and an inner current loop is designed. Besides, load current feedforward control is introduced to enhance the dynamic response. Moreover, phase‐shift control is employed as the modulation method, in which duty ratios change along with sinusoidal signals. Simulation and experimental results show that the control strategy could make the three‐phase current of the converter sinusoidal and balanced.

Aiming at the problems of complex realization, single control target and difficulty in selecting the optimal zero sequence voltage when the traditional space vector PWM (SVPWM) algorithm is adopted in five‐level active neutral‐point‐clamp (5L‐ANPC) converter, a predictive control optimal switching sequence selection (PC‐OSSS) method is proposed. The method adopts the SVPWM control strategy base on the line voltage coordinate and then realizes the selection of optimal switching sequence through the multi‐objective predictive control method. Firstly, the principle of SVPWM based on line voltage coordinate is introduced, and all switching sequences that can be output are obtained. Secondly, the mathematical models of the floating capacitor (FC) voltage, the neutral point (NP) voltage and the common mode (CM) voltage corresponding to each switch sequence are established. Finally, the optimal switch sequence is selected by using the predictive control method based on priority ranking. Compared with traditional SVPWM, the outstanding advantage of this method is that the traditional qualitative comparison method is replaced by quantitative calculation. The multi‐objective control performance of the converter is improved while retaining the advantages of the traditional SVPWM fixed frequency. Finally, the practicability and effectiveness of the proposed control strategy are verified by simulation and experiment.

The power grid has become the most important infrastructure system and dominates the economic and social developments in modern society. However, the power network often suffers various failures which can cause huge damages to the power grid. Hence, this paper proposes a community detection method to explore the critical community and identify the critical node, so as to study the vulnerability and resilience of the power grid subjected to malicious attacks. Meanwhile, this paper presents a comprehensive model to quantify the resilience characteristics considering performance level and change speed, and the IEEE 118 network is adopted to verify the feasibility and effectiveness of the proposed schemes. The results show that the attack strategies considering the community structure can cause severe damages to the power grid compared with the traditional methods, and Community Gravity-based Recovery (CGBR) has the better recovery ability among the six recovery strategies. Moreover, the results also demonstrate that Community Priority and Gravity-based Recovery (CPGBR) and CGBR can generate the most complete communities, and they can cause the least resilience damages on the number of surviving nodes (SN) subjected to the attacks on nodes 49 and 69 respectively, and Gravity-based Recovery (GBR) can result in the least resilience damages on the power supplied (PS). Hence, we can discover that there is a tradeoff between decreasing resilience loss and increasing network performance by different choices of recovery strategies. Moreover, the results show that the proposed methods can precisely identify critical nodes and rapidly recover network characteristics, therefore this study has great significances for decreasing resilience loss of the power grid subjected to malicious attacks.