ABSTRACT
An accurate and reliable estimation of photovoltaic models holds immense significance within the realm of energy systems. In pursuit of this objective, a Boosting Flower Pollination Algorithm (BFPA) was introduced to facilitate the robust identification of photovoltaic model parameters and enhance the conversion efficiency of solar energy into electrical energy. The incorporation of a Gaussian distribution within the BFPA serves the dual purpose of conserving computational resources and ensuring solution stability. A population clustering strategy is implemented to steer individuals in the direction of favorable population evolution. Moreover, adaptive boundary handling strategies are deployed to mitigate the adverse effects of multiple individuals clustering near problem boundaries. To demonstrate the reliability and effectiveness of the BFPA, it is initially employed to extract unknown parameters from well-established single-diode, double-diode, and photovoltaic module models. In rigorous benchmarking against eight control methods, statistical tests affirm the substantial superiority of the BFPA over these controls. Furthermore, the BFPA successfully extracts model parameters from three distinct commercial photovoltaic cells operating under varying temperatures and light irradiances. A meticulous statistical analysis of the data underscores a high degree of consistency between simulated data generated by the BFPA and observed data. These successful outcomes underscore the potential of the BFPA as a promising approach in the field of photovoltaic modeling, offering substantial enhancements in both accuracy and reliability.
ABSTRACT
In underwater acoustic target parameter estimation, the cross-spectrum method estimates the velocity of the target by the interference structure of the cross-correlated fields in two ranges. However, with the increase in interference in the marine environment, the performance of conventional cross-spectrum target velocity estimation methods gradually declines or even fails. To solve this problem, a cross-spectrum histogram method that combines the multi-line spectrum characteristics of the target is proposed in this paper. First, each frequency point of all line spectra is compensated by the cross-spectrum compensation method. Then the target velocity estimation value is obtained by performing probability distribution statistics on the velocity estimation results for all frequency points. Under different signal-to-noise ratios, simulations and experiments are carried out to verify the effectiveness of the proposed method. The results show that the proposed method is still effective in estimating the target velocity when the conventional method fails.