2. Algorithm process

The implementation process of WDO algorithm is as follows: 1) Initialize the number and dimension of air particles, define the maximum number of iterations and related parameter constants, set the search boundary (position and speed), and set the corresponding test function. 2) Initialize the initial information (position and speed) of each particle randomly, calculate the initial pressure value and arrange it in ascending order according to the pressure value. 3) Start the iteration. Update the velocity and position of air particles, calculate the particle pressure value and rearrange the population order in ascending order. 4) Iteration termination. Judge whether the termination condition is met, if not, return step 3), otherwise, the iteration is terminated, and the optimal position searched at last is the optimal solution.

Test the seven test functions listed in Table 2, each with a dimension of 30.

Table 2 Brief introduction of the test functions

During the experiment, the population number N NN of each algorithm was set to 30, and the maximum number of iterations was set to 500. The corresponding search range of the test function was shown in Table 2, and the range of search speed was one hundredth of the location search range. In the WDO algorithm, the constant parameter = 0.4, g = 0.2, T = 3 R, c = \ alpha = 0.4, 0.4 g = 0.2, RT = 3, c = alpha = 0.4, 0.4 g = 0.2, RT = 3, c = 0.4. The iterative convergence curve is shown in the figure below:

[1] Bayraktar Z , Komurcu M , Werner D H . Wind Driven Optimization (WDO): A Novel Natured-inspired Optimization Algorithm for electromagnetics[C]// IEEE. IEEE, 2010. [2] Zhang Tianneng, Zhou Chiwei. Structural Finite Element Model Modification Based on Second-order Taylor Series Expansion and Wind-Driven Optimization Algorithm [J]. [3] ZHU Xiangbing, LI Yuanjiang, WANG Jianhua, WU Hanqing. Wind Drive Optimization Algorithm Based on Levy Flight Mechanism. Computer and Digital Engineering, 2018, 46(10): 1943-1952,1956.