研究目的
To achieve optimal power generation using ANN for adaptive adjustment of weights of neurons using GA optimized, for fast and accurate tracking of PV power.
研究成果
The hybrid ANN-GA based scheme employing SEPIC converter effectively evaluates global PV power point under non-linear variations, with Bayesian regulation methodology for ANN training and GA based evolutionary algorithm minimizing RMSE. The proposed methodology generates a reference voltage corresponding to optimal duty ratio, demonstrating high correlative degree under non-linear variation in atmospheric conditions.
研究不足
The study focuses on simulated responses and may require further validation through physical implementation. The effectiveness under all possible real-world conditions is not fully explored.
1：Experimental Design and Method Selection:
The study implements a hybrid ANN-GA method for optimal power tracking of PV modules, utilizing Bayesian regulation for ANN training.
2：Sample Selection and Data Sources:
A Photovoltaic (PV) system is considered, with VOC and ISC as supply specifications and IMPP as generated output.
3：List of Experimental Equipment and Materials:
SEPIC converter for power tracking, dSPACE interface for inverter control.
4：Experimental Procedures and Operational Workflow:
The ANN-GA approach generates a reference voltage corresponding to optimal duty ratio, with GA adjusting ANN weights to minimize RMSE.
5：Data Analysis Methods:
Simulated responses under sag, swell, and varying loading conditions are analyzed to show the system's potency.
独家科研数据包，助您复现前沿成果，加速创新突破
获取完整内容
