Data-Driven Beetle Antennae Search Algorithm for Electrical Power Modeling of a Combined Cycle Power Plant

Abstract

Beetle Antennae Search (BAS) is a newly developed nature-inspired algorithm, which falls in the class of single-solution driven metaheuristic techniques. This algorithm mimics the searching behavior of the longhorn beetles for food or potential mate using their long antennae. This algorithm is potentially effective in achieving global best solutions promptly. An attempt is made in this paper to implement the data-driven BAS, which exploits the Cascade Feed-Forward Neural Network (CFNN) training for functional approximation. The proposed technique is utilized to model the electrical power output of a Combined Cycle Power Plant (CCPP). The power output of a power plant could be dependent on four input parameters, such as Ambient Temperature (AT), Exhaust Vacuum (V), Atmospheric Pressure (AP), and Relative Humidity (RH). These parameters affect the electrical power output, which is considered as the target variable. The CFNN based predictive model is shown to perform equivalently while compared with published machine learning based regression methods. The proposed data-driven BAS algorithm is effective in producing optimal electric power output for the CCPP.