Abstract: Water wave optimization algorithm(WWO) is a novel population-based optimization algorithm.Despite of its advantages with few controllable parameters,simple operations,and easy implementation,it still risks slow convergence rate and low search precision.To mitigate the aforementioned risks,firstly,a concise yet powerful theoretical analysis was carried out to derive the convergence condition that the control algorithm parameters should satisfy.Then,an improved WWO was proposed to meet the above conditions by incorporating an adaptive algorithm parameters tuning strategy,and it’s expected to further enhance the capability of balancing between global exploration and local exploitation via this strategy.Finally,simulation experiments and statistical comparisons between four algorithms(ApWWO,WWO,FA,MVO) on 10 benchmark functions were conducted.The results show that ApWWO performs significantly better than WWO and FA in terms of search accuracy,speed and robustness,as well as outperforms MVO in five test function.Compared with PSO and GA,ApWWO can yield good performance,which can be affected by the problem’s dimension and population size,and ApWWO also performs well on the permutation flow shop scheduling problem.

Key words: Evolutionary algorithm,Water wave optimization algorithm,Adaptive controlling parameter,Permutation flow shop scheduling