Modular multilevel converter (MMC) has attracted much attention for years due to its good performance in harmonics reduction and efficiency improvement. Model predictive control (MPC) based controllers are widely adopted for MMC because the control design is straightforward and different control objectives can be simply implemented in a cost function. However, the computational burden of MPC imposes limitations in the control implementation of MMC because of many possible switching states. To solve this, we design machine learning (ML) based controllers for MMC based on the data collection from the MPC algorithm. The ML models are trained to emulate the MPC controllers which can effectively reduce the computation burden of real-time control since the trained models are built with simple math functions that are not correlated with the complexity of the MPC algorithm. The ML method applied in this study is a neural network (NN) and there are two types of establishing ML controllers: NN regression and NN pattern recognition. Both are trained using the sampled data and tested in a real-time MMC system. A comparison of experimental results shows that NN regression has a much better control performance and lower computation burden than the NN pattern recognition.

The integration of artificial intelligence (AI) techniques in power converter-based systems has the potential to revolutionize the way these systems are optimized and controlled. With the rapid advancements in AI and machine learning technologies, this paper presents the analysis and evaluation of these powerful tools as well as in computational capabilities of microprocessors that control the converter. This paper provides an overview of AI-based controllers, with a focus on online/offline supervised, unsupervised, and reinforcement-trained controllers. These controllers can be used to create surrogates for inner control loops, complete power converter controllers, and external supervisory or energy management control. The benefits of using AI-based controllers are discussed. AI-based controllers reduce the need for complex mathematical modeling and enable near-optimal real-time operation via computational efficiency. This can lead to increased efficiency, reliability, and scalability of power converter-based systems. By using physics-informed methods, a deeper understanding of the underlying physical processes in power converters can be achieved and the control performance can be made more robust. Finally, by using data-driven methods, the vast amounts of data generated by power converter-based systems can be leveraged to analyze the behavior of the surrounding system and thereby forming the basis for adaptive control. This paper discusses several other potential disruptive impacts that AI could have on a wide variety of power converter-based systems.

This article proposes a machine learning (ML)-based emulation of model predictive control (MPC) for modular multilevel converters (MMCs). In particular, the artificial neural network model, trained offline by the data collected from the traditional fast MPC method, is used to control the MMCs with high accuracy. With this offline training, the majority of computational burden is transferred from online to offline. Therefore, the proposed ML MPC can replace the role of the traditional MPC. The experimental results show that the proposed ML-based MPC has the same performance as the conventional MPC but a significantly computationally efficient structure. The finding from the letter provides ground for many other applications for ML-based emulation of complex controllers in power electronic systems.

The capacitor voltage ripples of the modular multilevel converter (MMC) are increased under unbalanced grid fault conditions. Since high capacitor voltage ripples deteriorate their lifetimes and may even cause tripping of the MMC system, it is important to restrict them. To this end, it is well known that injecting double fundamental frequency circulating currents can reduce the capacitor voltage ripples. However, finding a proper circulating current reference to achieve desired ripples analytically is complicated. This paper proposes an alternative method to quickly calculate the proper circulating current references without analytical computations, which is achieved by an artificial neural network (ANN) trained to approximate the relationship between circulating current references and capacitor voltage ripples. The training data is firstly extracted from a detailed simulation model of the MMC. Afterwards, the ANN is trained by the input-output data to obtain the mapping relationship, which is then used to derive the desired circulating current references. Both the simulation and the experimental results verify the practicability of the proposed method, where the operating region can be extended 30% at a minimum in all testing conditions.

As high-speed railways develop rapidly worldwide, the problems of neutral section and power quality are becoming increasingly prominent. These problems affect the normal operation of the power grid and the electric locomotives. The co-phase power supply schemes were proposed by researchers, which can solve these problems well. Among them, the MMC-based scheme (MMC-5 L) is a representative one. The existing scheme, when applied to the auto-transformer (AT)-fed system which is usually adopted by high-speed railways, needs to install an auto-transformer on the output side. To eliminate the auto-transformer and to obtain the more suitable and superior co-phase power supply schemes to AT-fed system, two improved schemes are proposed in this paper based on the novel MMC-based scheme (MMC-6 L) we proposed before. While establishing the mathematical model, creating the control system, and carrying out simulation verification of the proposed schemes, the proposed schemes are compared with MMC-6 L and MMC-5 L in terms of topological characteristics, control system complexity, power loss, and functionality. The results show that while the proposed schemes can fulfill the functions of the co-phase power supply scheme well like the MMC-5 L and MMC-6 L, they are more suitable and superior for the AT-fed system in terms of control system complexity and power loss. Among them, MMC-5L1CB performs best in power loss, and MMC-4L2CB performs best in control system complexity.

This article proposes a novel application method, inverse application of artificial intelligence (IAAI) for the control of power electronic converter systems. The proposed method can give the desired control coefficients/references in a simple way because, compared to conventional methods, IAAI only relies on a data-driven process with no need for an optimization process or substantial derivations. Noting that the IAAI approach uses artificial intelligence to provide feasible coefficients/references for the power converter control, rather than building a new controller. After illustrating the IAAI concept, a conventional application method of artificial neural network is discussed, an optimization-based design. Then, a two-source-converter microgrid case is studied to choose the best droop coefficients via the optimization-based approach. After that, the proposed IAAI method is employed for the same microgrid case to quickly find good droop coefficients. Furthermore, the IAAI method is applied to a modular multilevel converter (MMC) case, extending the MMC operation region under unbalanced grid faults. In the MMC case, both simulation and experimental online tests validate the operation, feasibility, and practicality of IAAI.