جستجوی مقالات مرتبط با کلیدواژه « mayfly algorithm » در نشریات گروه « فناوری اطلاعات »

In any economy, it is essential to monitor the rate of population change closely. Governments employ various strategies and programs to regulate population growth since different population growth rates have distinct economic consequences. This paper reveals a global trend of reduced desire to have children, with variations across countries. The paper aims to predict the population growth rate in England by employing Artificial Neural Networks (ANN) in combination with various meta-heuristic algorithms, including the Sparrow Search Algorithm (SSA). The selection of SSA and other algorithms is based on factors such as accuracy and computational efficiency. A set of 18 economic indicators serves as input variables, and a Genetic Algorithm (GA) is used for feature selection. The data used for analysis spans the most recent ten years and is presented on a monthly basis. The results indicate that SSA exhibits the lowest prediction errors for the population growth rate among the applied algorithms in this paper. The primary contribution of this study lies in the application of hybrid algorithms that combine SSA-ANN with other algorithms, such as LA. The paper also emphasizes the inclusion of influential and impactful indices as input variables to enhance prediction accuracy.

The combined economic emission dispatch (CEED) is an important consideration in every power system. In this paper, a modified Mayfly Algorithm named Modified Individual Experience Mayfly Algorithm (MIE-MA) is used to solve the CEED optimization problem. The modified algorithm enhances the balance between exploration and exploitation by utilizing a chaotic decreasing gravity coefficient. Additionally, instead of the MA relying solely on the best position, it calculates the experience of a mayfly by averaging its positions. The CEED problem was modelled as a nonlinear optimization problem constrained with four equality and inequality constraints and tested on a grid-connected microgrid that consists of four dispatchable distributed generators and two renewable energy sources. The performance of the MIE-MA on the CEED problem was compared to Particle Swarm Optimisation (PSO), an MA variant that incorporates levy flight algorithm named IMA and Dragonfly Algorithm (DA) using the MATLAB R2021a software. The MIE-MA achieved the best optimum cost of 11306.6 $/MWh, compared to 12278.0 $, 12875.8$, and 17146.4$ of the DA, IMA and PSO respectively. The MIE-MA also achieved the best average optimum cost over 20 runs of 12163.48 $, compared to 12555.36 $, 13419.67 $ and 17270.08 $ of the DA, IMA, and PSO respectively. The hourly cost curve of the MIE-MA was also the best compared to the other algorithms. The MIE-MA algorithm thus achieves superior optimal values with fewer iterations.