Research On Improvement Of Teaching-Leaming-Based Optimization Algoritom

The greatest advantages of teaching-learning-based optimization (TLBO) lie in its simple, principle and realization, in addition to its less controlled parameter, and the computational efficiency is high traditional methods. So it has been widely applied in many areas, such as function optimization, neural network optimization, engineering optimization, since the TLBO was put forward. However, every optimization is not always omnipotent, the phenomenon of premature convergence also exists in TLBO. Therefore, it is necessary for its theory and application to make further improvement and research.The text put forward the relevant improvement tactics aimed at unconstrained optimization problems,0-1 nonlinear programming problems, basing on the analysis of standard TLBO algorithm. And it is effective by numerical experiment. The main work of this thesis as follows:1 Two improved TLBO algorithm are proposed:one is TLBO algorithm with the strategy of adaptive teaching factor, the other is TLBO algorithm with the strategy of linear decreasing learning the weight. The experimental results demonstrate than both of the two proposed algorithm are better that the standard TLBO algorithm on convergence speed and optimization accuracy.2 As for the disadvantage in local optima and premature convergence of TLBO algorithm in solving complex optimization problems, a modified TLBO algorithm by using simulated annealing is proposed. The experimental results demonstrate that improved TLBO algorithm is an effective global optimization algorithm.3 For solving 0-1 nonlinear programming problems, a kind of chaotic TLBO algorithm was proposed. It is shown in numerical experiments that this algorithm is simple and easy to implement with fast convergence and high accuracy.Finally, we make a summary of this thesis briefly, and put forward some problems that are worthy to be studied in the future.