Study On Flexible Detection Method Of Composite Thin Wall Parts' Shape

In recent years,the composite material has been widely used in aircraft manufacturing industry because of its excellent comprehensive properties,and it has become the first choice to replace the single material.Aircraft skin,which is made of composite material,is one of the most important parts of the aircraft body.The detection of the skin morphology is also an important part of the quality inspection in aircraft manufacturing.The traditional back molding method,subject to low precision,large artificial factors and some other disadvantage,is no longer suitable for the morphology of composite thin wall parts' detection.The study on the flexible detection method of morphology has important theoretical and practical significance.Firstly,the development status of the composite thin wall parts' flexible morphology detection was reviewed,and the methods which are basely used,including the flexible tooling platform,finite element analysis method,differential evolutionary algorithm and genetic algorithm,were also introduced.Then,the establish method of thin wall parts' deformation model,through identification with ANSYS software,was proposed.The problems of identification data acquisition,the selection of identification optimization's fitness function,mesh size and identification optimization algorithms were discussed.The simulation results showed that the proposed modeling method can meet the accuracy requirements and its identification speed is faster.Based on thin wall parts' deformation model,according to the actual flexible morphology detection problem,the research group proposed a general idea of composite material thin wall parts' morphology detection,which is first support the workpiece with the support scheme optimization result,then,preset the correction force with optimizing,finally,correction online.As this,in view of the shortage of current flexible support optimization method,the research group introduced the piecewise variation to accelerate the convergence speed.At the same time,using the differential evolution algorithm(DE)to optimize the correction force.In addition,according to the existing problems of the online PID correction method,introduced model free adaptive control(MFAC)used on the MIMO system,put forward the fuzzy feedforward decoupling MFAC method to avoid divergence and to raise the convergence speed.The simulation results verified the effectiveness of the proposed method.Finally,designed the software simulation platform of composite material thin wall parts'flexible morphology detection.Based on the study of MATLAB and ANSYS software joint simulation method,the research group proposed a simulation method of parallel computing with Python platform communication.Based on C#and MATLAB hybrid coding technology,completed the preliminary design and implementation of the software simulation platform's function and interface.