A Novel Invasive Weed Optimization Algorithm (IWO) by Whale Optimization Algorithm(WOA) to solve Large Scale Optimization Problems

Abstract

Abstract In this work, two algorithms of Metaheuristic algorithms were hybridized. The first is Invasive Weed Optimization algorithm (IWO) it is a numerical stochastic optimization algorithm and the second is Whale Optimization Algorithm (WOA) it is an algorithm based on the intelligence of swarms and community intelligence. Invasive Weed Optimization Algorithm (IWO) is an algorithm inspired by nature and specifically from the colonizing weeds behavior of weeds, first proposed in 2006 by Mehrabian and Lucas. Due to their strength and adaptability, weeds pose a serious threat to cultivated plants, making them a threat to the cultivation process. The behavior of these weeds has been simulated and used in Invasive Weed Optimization Algorithm (IWO), as for the Whale Optimization Algorithm (WOA) uses the intelligence of the swarms to reach the goal and achieve the best solution, which simulates the unique hunting behavior of humpback whales, which is called fishing by bubble trap hunting by creating distinctive bubbles along a circle or a path in the form of 9 has appeared for the first time in 2016 by Mirjalili and Lewis. In order to benefit from the intelligence of the flocks and to avoid falling into local solutions, the new hybridization between the IWO and WOA algorithm was proposed to launch the new hybrid algorithm (IWOWOA). The new hybrid algorithm (IWOWOA) was applied on 23 functions of large scale optimization problems, The proposed algorithm showed very high efficiency in solving these functions. The proposed algorithm was able to reach the optimal solutions by achieving the minimum value of most of these functions. This algorithm was compared with the basic algorithms IWO, WOA and two algorithms that follow the swarm system these algorithms are particle swarm optimization (PSO) and chicken swarm optimization (CSO) [7], they have been statistically tested by calculating the mean arithmetic μ and standard deviation σ for these functions.