Study on self-optimizing system of key parameters for complex machine products

Abstract

Modeling and FEA (finite element analysis) are required to be done over again after changing the key parameters. Thus, this paper constructs a distributed system which can change parameters of product, model and do FEA automatically. And it obtains the optimizing parameters at last. This system running on intranet bases on genetic algorithm and CATIA/CAA. Every client reads key parameters from the database in the server, drives the CATIA three-dimensional model by CATIA/CAA and the key parameters, and does FEA. After the result of FEA was sent to the database, the server evaluates the result of FEA and sends the evaluation to the database. Then, the server creates new key parameters using genetic algorithm; the system goes to next circle. The parameters, whose evaluation is the best, are the last result. Based on this system, the self-optimizing of key parameters for TBM (tunnel boring machine) is realized and accelerates the design process