An Improved Teaching-Learning Optimization Algorithm based on Morlet Wavelet Variation

In order to overcome the weaknesses of the Teaching and Learning optimization (TLBO) algorithm in solving function optimization problems, such as easy to fall into local optimum, slow convergence at the later stage, and low solution accuracy, an improved algorithm with dynamic adaptive teaching factors and Morlet wavelet variation-based algorithms is proposed. Firstly, the improved algorithm introduces a nonlinear dynamic teaching factor to adjust the influence of teachers on students in the iterative optimization process. Secondly, in order to avoid algorithm trapped in local optimum, using Morlet wavelet to the implementation of global extreme value of each dimension in each generation wavelet disturbance, disturbance and the result was recognized as a certain probability is selected for the new position of the individual, make full use of the advantage of global extremum information guide populations to be near optimal solution quickly, by fine-tuning characteristics of wavelet functions help population out of local minima. The simulation results on 18 classical test functions show that the improved algorithm has better performance than TLBO, SLTLBO, CSA and BOA algorithms, and is suitable for solving function optimization problems. It is applied to engineering practice to solve PID parameter optimization and obtains good results.