Using a modified whale optimization algorithm to solve dynamic arrival flights sequencing problem

Abstract For solving dynamic arrival flights sequencing problem, a Modified Whale Optimization Algorithm (MWOA) is proposed. The method can avoid WOA’s disadvantage in several aspects, such as local optimum stuck and solution accuracy degradation. To balance the exploration and exploitation abilities, a nonlinear dynamic strategy based on Branin function for updating the control parameter is given. Chaotic mutation based on a sine function is applied to avoid its falling into local optimum. Mirror selection strategy is adopted in iteration to increase convergence speed. The early maturity detection through Gauss vibration is added to improve local mining and global searching abilities. Twenty-five well-known benchmark test functions are used to test the algorithm. The results show MWOA outperforms other six state-of-the-art optimization algorithms which include Particle Swarm Optimization(PSO), Whale Optimization Algorithm(WOA), Improved PSO(IPSO), β-hill Climbing Modified WOA(BMWOA), Harris Hawks Optimization(HHO), WOA-PSO in terms of solution accuracy andconvergence speed. Then, all the above-mentioned seven algorithms, traditional first-come-first-serve(FCFS) sequencing strategy, Immune Teaching-Learning Based Optimization(ITLBO), PSO based on the Random Key representation(PSORK) and Genetic Algorithm-PSO(GA-PSO) are applied to solve the arrival flights sequencing model. The results show MWOA outperforms the nine algorithms and FCFS strategy. The landing efficiency improved is 16.90%, 2.26%, 2.26%, 2.89%, 2.75%, 2.40%, 1.99%, 2.26%, 7.26% and 3.55% respectively compared with FCFS, PSO, WOA, IPSO, BMWOA, HHO, WOA-PSO, GA-PSO, ITLBO and PSORK. Furthermore, the proposed algorithm also has best performance in terms of optimality, reliability, robustness and will-time delay. The Wilcoxon tests show MWOA is significantly different to other algorithms.