INVESTIGATION ON RESOURCE OPTIMIZED AND FAULT TOLERANT HYBRID MULTILEVEL INVERTERS

Abstract

Multilevel inverters (MLIs) have revolutionized power conversion technology, gaining prominence in high-power applications due to their superior performance compared to traditional two-level voltage source inverters (2L-VSIs). While 2L-VSIs rely on the series connection of power semiconductor devices to meet the increasing voltage and power demands of multi-megawatt industrial systems, MLIs offer distinct advantages. These include improved output voltage and current waveforms, lower harmonic distortion, reduced dv/dt stress, smaller common-mode voltage, minimized filter size, and enhanced efficiency. Such attributes make MLIs an ideal choice for medium voltage and high power applications, such as industrial motor drives, renewable energy systems, and grid-connected inverters, where robust and efficient power conversion is critical. MLIs achieve high-quality stepped voltage waveforms by employing multiple power semiconductor switches and capacitor voltage sources, offering greater flexibility in waveform shaping. This capability not only reduces harmonic distortion and switching losses but also minimizes the need for extensive filtering, thereby enhancing overall system efficiency and reliability. These features enable MLIs to address critical challenges like voltage regulation, harmonic mitigation, and nonlinear load compensation in modern power systems. Despite their advantages, MLIs are continuously advancing with the development of novel and hybrid topologies aimed at achieving higher voltage levels while minimizing device count, simplifying system complexity, and enhancing efficiency, thereby expanding their applicability in industrial and emerging applications. MLIs, while offering numerous advantages, encounter significant challenges that limit their widespread adoption in industrial and grid-connected systems. The inherent increase in component count, including switches, gate drivers, and auxiliary circuitry, inherently increases system complexity and costs as the number of voltage levels increases. These challenges are further compounded by the limitations of traditional modulation techniques, which often struggle to meet the stringent requirements of high-level MLIs, such as achieving a balance between low switching losses and high-quality output. A significant limitation of MLIs is their dependence on isolated dc sources, particularly in topologies designed for higher voltage levels. While multiple-source configurations are ideal for applications like photovoltaic systems, where isolated dc sources are readily available, single-source MLIs i provide a simpler solution and are more suitable for back-to-back regenerative applications. However, single-source MLIs face challenges in maximizing dc bus utilization. The maximum dc bus utilization in a hexagonal space vector structure-based VSI is achieved during square wave or sixstep operation, corresponding to a fundamental phase peak voltage of 0.637 times the dc link voltage. In contrast, with conventional modulation techniques, the linear modulation range is limited to a fundamental phase peak voltage of 0.577 times the dc link voltage. Traditional overmodulation techniques, often used to extend the linear modulation range, typically introduce lower-order harmonics that degrade the output waveform quality. Hybrid topologies featuring a primary unit with an active dc source and secondary units fed by floating capacitors offer improve dc bus utilization by extending the linear modulation range. Generally, the voltage balance of floating capacitors is achieved using pole voltage redundancies. However, at higher modulation indices, maintaining this balance becomes more challenging, requiring advanced control strategies to ensure stable and efficient operation. In the absence of redundant states, techniques such as model predictive control or fuzzy logic control must be used to manage the voltage balancing effectively. Fault tolerance poses a further challenge as the number of switches and components increases in hybrid MLIs. Switch faults can lead to complete system shutdown, which is unacceptable in critical applications like grid-connected systems and industrial drives. Conventional fault-tolerant strategies often involve reducing the modulation index or oversizing switches, both of which compromise efficiency and escalate costs. Specifically, lowering the modulation index limits the system’s ability to achieve the required output voltage levels efficiently, particularly in grid applications. As a result, developing fault-tolerant methods that maintain high modulation indices is crucial for improving both performance and cost-efficiency, thereby broadening the scope of hybrid MLIs in various applications.