Opposition based Laplacian Ant Lion Optimizer

Performance of any nature inspired optimization algorithm is subject to appropriate combination of operators used for exploration and exploitation. The lack of this combination inclines an algorithm towards premature convergence, entrapment of local optima and inability to reach global optima. This paper presents a novel algorithm called opposition based Laplacian antlion optimizer (OB-L-ALO) to accelerate the performance of the original ALO. For achieving acceleration, exploration is to be enhanced. Two strategies are used for this purpose: Firstly, Laplace distribution is used in random walk of ALO instead of uniform distribution which ensures exploration of more search area than the original random walk of ALO. Secondly, Opposition Based Learning model which ensures the exploration of original as well as opposite candidate solutions in the search space at the same time to estimate the better candidate solutions while evolution process is in progress. A comprehensive set of 27 benchmark problems including wide range of different characteristics and different dimensions have been employed for verification of results. Also the influence of Laplace distribution random numbers and opposition based new population generation during evolution process has been analysed by behaviour of trajectories, convergence rate, data distribution of objective function values using boxplot and average fitness improvement for certain test suit. The proposed OB-L-ALO is also employed to the set of unconstrained engineering design problems of Gear Train Design and Optimal Capacity of Gas Production Facilities, showing diversity in solving the real world optimization problems.