Reliability Analysis Method Based On Teaching And Learning Optimization Algorithm

Engineering structures are often affected by uncertain factors in the process of design,construction and use,the safety,applicability and durability of the structure will change.Structural reliability analysis takes the influence of various uncertain factors into account to assess structural reliability.At present,structural reliability analysis mainly targets at two levels of the structure,components and systems.The component reliability analysis based on probability theory has been maturely developed.Its main task is to solve the failure probability or reliability index according to the structural performance function.In fact,modern engineering structures become more complex and larger.When the highly nonlinear performance function or structural performance function that can only be expressed implicitly are solved using traditional calculation methods,convergence,accuracy and calculation efficiency are not satisfactory.With the rise of swarm intelligence algorithms,FORM based on the swarm intelligence optimization algorithm has excellent global search capabilities providing the possibility to improve the convergence and the computational efficiency of complex structural reliability analysis.In view of the fact that Teaching-Learning-Based Optimization has the characteristics of fast convergence and simple implementation,this paper incorporates Teaching-Learning-Based Optimization and Teaching-learning-based optimization with variable-population scheme with FORM into analyzing the reliability of structural and finding the failure modes of structural systems.The influence of the parameters on the structural reliability analysis results in the two algorithms is studied.A new calculation method of penalty function coefficients is proposed to improve the accuracy and calculation efficiency of reliability analysis.The results of numerical examples show that the method proposed in this paper has good convergence and high accuracy,which avoids the poor accuracy of reliability analysis due to the premature.Structural system usually involves many failure modes.How to identify the main failure modes and deal with the correlation between failure modes,and calculate the joint failure probability is the main task of structural system reliability analysis.This paper proposes a method employing multi-objective teaching-learning optimization algorithms to identify the main failure modes,andattempts to use the powerful search capability of the TLBO to improve the recognition efficiency.The results of numerical examples show that the method in this paper achieves both the accuracy of recognition and the search efficiency.It is suitable for large redundant structures.In addition,this paper proposes a new VTLBO-based FORM for parallel system reliability analysis.It is used to calculate the failure probability of a parallel system under a single failure mode.At the same time,this method is combined with the matrix-based system reliability method,yielding a new hybrid system reliability analysis method.The results of the numerical examples show that the two reliability methods presented above have good accuracy and computational efficiency,which provides potential value in practical engineering application.