Abstract:Modulation strategy plays an important role in the good performance of modular multilevel converter (MMC). Since MMC has a special structure as well as a large number of submodules (SMs), it is almost impossible to apply the conventional multilevel space vector modulation (SVM) algorithm in it. In this study, a simplified SVM implementation method is proposed by grouping SMs. With the presented method, the high-level SVM process for MMC can be easily done by implementing multiple two-level or three-level SVM algorithm, while the complicated offline or online efforts/tasks about mapping switching states or sequences in the conventional multilevel SVM are not necessary any more. In order to improve the harmonic characteristics, the sampling time of different subunits (SUs), each of which is formed by grouping a few SMs, is staggered evenly, which helps to achieve a higher equivalent switching frequency and less low-order harmonics. Further, the capacitor voltage balance of SMs is improved by rotation strategies between SUs or inside each SU even without any capacitor voltage detecting or sorting. Since the presented method is independent of the number of SMs, it can be very easily expanded into a high-level MMC. These propositions are verified by experiment.

Improvement On Response Performance Of Min-Max Modular Classifier By Symmetric Module Selection

On Efficient Selection of Binary Classifiers for Min-Max Modular Classifier

Learning from imbalanced data sets with a Min-Max modular support vector machine

Efficient Classification of Multi-Label and Imbalanced Data Using Min-Max Modular Classifiers

A Multi-Class Large Margin Classifier

Weighted Modular 2 DPCA-Based Face Recognition from a Single Sample Image Per Class

Learning Imbalanced Data Sets with a Min-Max Modular Support Vector Machine

On Effective Decomposition of Training Data Sets for Min-Max Modular Classifier

Module Combination Based on Decision Tree in Min-Max Modular Network

Structure pruning strategies for min-max modular network

A general procedure for combining binary classifiers and its performance analysis

Emergent on-line learning in min-max modular neural networks

Multi-view Face Recognition with Min-Max Modular SVMs.

A MARKET CLEARING PRICE PREDICTOR BASED ON MODULAR NETWORKS

An Algorithm for Pruning Redundant Modules in Min-Ma Modular Network

A Confident Majority Voting Strategy for Parallel and Modular Support Vector Machines

Decomposition and Parallel Learning of Imbalanced Classification Problems by Min-Max Modular Neural Network

A Simplified Space Vector Modulation Implementation for Modular Multilevel Converter by Grouping Submodules

Dynamic Base Classifier Pool for Classifier Selection in Multiple Classifier Systems.

Dynamic Classifier Combination Method Based on Minimum Cost Criterion

Modified Minimum Squared Error Algorithm for Robust Classification and Face Recognition Experiments