The wind driven optimization(WDO) is a new heuristic algorithm, which is a kind of heuristic process based on population. The algorithm was proposed by Bayraktar Z. et al in 2010. The idea of the wind driven optimization is to simulate the air flow caused by the pressure of different places, so that the pressure can reach equilibrium in the end. In the atmosphere, the flow of air is to try to balance the air pressure. When the pressure imbalance, the air will bear the pressure gradient force, it will flow from the high pressure area to the low pressure area with a certain speed, and finally make the air pressure reach equilibrium. As the wind driven optimization has the advantages of simple structure, less parameters, easy to understand and easy to realize, the algorithm has been concerned by more and more scholars. However, the algorithm also has a fast convergence speed, easy to fall into local optimum, and the later population diversity is not enough to lead to slow convergence speed and other defects, which greatly limits the application range of wind driven optimization. Therefore, the wind driven optimization needs to be further studied and extended both in theory and in application.In this paper, the shortcomings of the wind driven optimization are analyzed. The algorithm is improved from the aspects of the update strategy and encoding method, and the improved algorithm is applied to the practical optimization problems. The main work of this paper includes the following three aspects:(1) A double population strategy is adopted to improve the wind driving optimization, a population is updated by the wind driven optimization, and the other is updated by differential evolution. The two algorithms achieve the coevolution of the population through information sharing mechanism. This method can increase the diversity of population, and then enhance the global search ability of the algorithm, and avoid the algorithm to get into local optimum because of fast convergence.(2) Improved wind driven optimization by changing the individual’s encoding mode. The idea of using complex-valued encoding to wind driven optimization, using the real and imaginary part to represent an independent variable. Because each complex number can express two-dimensional information, which can enhance the information of the population and the diversity of individuals in the group. This paper also introduces a greedy strategy based on the complex-valued encoding wind drive optimization to solve the 0-1 knapsack problem.(3) The idea of quantum encoding theory is applied to the wind driven optimization, a quantum wind driven optimization is proposed. Quantum rotation gate is used to realize the population of update, the quantum not gate to achieve variation of individuals in a population. These two strategies can improve the population diversity and avoid premature convergence. The quantum wind driven optimization is applied to UCAV path planning, which proves the validity and feasibility of the algorithm. |