Research On Hybrid Reliability Optimization Method

For a complex aircraft mechanism system,the reliability of the structure analysis results often greatly influenced by the uncertainty of the variables.Due to the uncertainty of the vari-ables,some of the variables are suitable for the application of probabilistic methods,some var-iables can only be treated with non-probabilistic methods.Even in some cases,some variables may have the characteristics of non-probability and probability at the same time,which brings difficulties to the reliability analysis of structures.Considering the uncertainty of the variables in the reliability analysis,it is necessary to construct a hybrid reliability model which can be used to measure the reliability of the structure.In this paper,we mainly study the hybrid model with three kinds of variables,which are the truncated random,fuzzy and interval.The main contents include the truncated random,fuzzy and interval three variables in the same function of the variables in the mutual trans-formation,hybrid reliability model and hybrid reliability index is defined and solved.In the aspect of the definition and solution of hybrid reliability,this paper focuses on solving the re-liability index ? by using the SQP method and solving the reliability index ? with the AFORM.In the aspect of reliability optimization method,this paper presents an improved adaptive genetic algorithm.The improved adaptive genetic algorithm is mainly used to dynamically adjust the relevant parameters of genetic operations,to ensure that the excellent individuals in each generation of population are not destroyed,so that the genetic operations can be carried out continuously,and findlly the optimal solution can be obtained.Finally,the improved adap-tive genetic algorithm and hybrid reliability solution method based on refined hybrid reliabil-ity optimization model is generic,and explained by a reliability optimization problem of syn-chronizer system based on this model.However this method provides a new idea for engi-neering designers,and the application of the new method should be perfected many times in order to play its due value.