Empirical analysis of control models for different converter topologies from a statistical perspective
Converter topologies including SEPIC, ZETA, etc. are controlled via selection of capacitive and inductive components, which assists in improving its conversion efficiency for different type of loads. A wide variety of such models are proposed by researchers, that include, but are not limited to, bio...
Saved in:
| Main Author: | |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
EDP Sciences
2025-01-01
|
| Series: | Science and Technology for Energy Transition |
| Subjects: | |
| Online Access: | https://www.stet-review.org/articles/stet/full_html/2025/01/stet20240208/stet20240208.html |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Converter topologies including SEPIC, ZETA, etc. are controlled via selection of capacitive and inductive components, which assists in improving its conversion efficiency for different type of loads. A wide variety of such models are proposed by researchers, that include, but are not limited to, bioinspired techniques for rating selection, Neural Network based models for load-based component selection, etc. Each of these models vary in terms of their internal operating characteristics, and showcase highly variant quantitative and qualitative performance under different loads. This variation increases the ambiguity of model selection under context-specific applications. To reduce this ambiguity, a detailed discussion about these models in terms of their context-specific nuances, qualitative advantages, deployment-specific limitations, and functional future scopes is described in this text. Based on this discussion, researchers will be able to identify optimal models for their application-specific use cases. It was observed that bioinspired models and incremental learning techniques assists in improving control performance for efficiency-aware use cases. This text also compares the evaluated models in terms of their conversion efficiency, cost of deployment, delay needed for control, scalability and computational complexity under different scenarios. Based on this comparison, researchers will be able to identify optimized models for their performance-specific deployments. This text further proposes evaluation of a novel Converter Control Rank Metric (CCRM) that combines these metrics, and assists in identification of converter control models that showcase high conversion efficiency with low delay, low complexity, high scalability, and low deployment costs. This will allow readers to select and modify optimized models for their context-specific use cases. The main reason of conduction this research is how to dispel the ambiguity in the process of choosing the suitable control models corresponding to the appropriate converter topology that can satisfy a set of particular application requirements. Even though, there are various control models available, these characteristics often become a barrier to their straightforward adoption for applications that are performance critical. This paper fills this void by providing a thorough experimental comparison of these models, resulting in the creation of the CCRM metric. The proposed framework gives researchers and Practitioners a unitary decision making high level tool for finding, tailoring and enhancing the performance of control models necessary for improved conversion efficiency, increased scalability and cost effectiveness in the deployment. |
|---|---|
| ISSN: | 2804-7699 |