Research On Tooling Strain Monitoring And Forecasting Technology During Aircraft Assembly

With the increasing importance of large aircraft in production and life,in order to meet the requirements of high flight safety,stable functions and long service life,the assembly accuracy of the whole aircraft has a high standard,and the aircraft is assembled in coordination under the premise of accuracy assurance.Due to assembly operations,assembly processes,and other reasons,during the aircraft assembly process,assembly stress inevitably occurs,leading to deviations in key positions of the tooling,which directly affects the quality of aircraft products and seriously affects the accuracy and quality of aircraft manufacturing.The strain monitoring system in the key area of tooling provides basis for tooling maintenance and quality maintenance evaluation,and guides aircraft assembly operations.Therefore,research on the strain monitoring technology of tooling is very important to improve the assembly quality of aviation products.This article has carried out in-depth research on tooling strain monitoring and prediction technology during aircraft assembly.The main research contents are as follows:(1)Strain measurement is based on FBG strain sensor.The basic working principle of FBG sensor is described and a calibration method of strain sensitivity coefficient is designed.The calibration method of strain sensitivity coefficient of fiber optic sensor and the strain prediction model of tooling are tested.After the calibration experiment,the corresponding strain sensitivity coefficient of FBG strain sensor can be obtained,which has high linearity and repeatability,and can be used in the field of high precision measurement.(2)In order to solve the problem of aircraft assembly tooling strain,the assembly operation is analyzed,the cause of strain formation is explored,and the key monitoring area of strain detection is determined.The front beam positioner,front beam auxiliary positioner,actuator positioner and hinge hole positioner are simulated respectively by COMSOL multiphysics software.Combined with assembly specifications and field conditions,the layout of direct measurement points is designed,then the real-time strain detection system of assembly tooling is established,which provides data support for strain measurement and prediction model establishment.(3)In order to solve the problem that some key areas cannot be monitored directly due to the complex structure of tooling parts and the requirements of actual aircraft assembly process,a prediction model of tooling strain based on support vector machine(SVM)is proposed.Using the tool strain real-time detection system to obtain the input and output strain information needed by the model,establish the nonlinear mapping between the input and output data,train the model,and finally achieve the purpose of accurately predicting the induced strain of complex components in the assembly process.(4)In order to improve the accuracy of the strain prediction model,the improved grid search method is used to optimize the parameters of the model,and the two-dimensional grid on the parameter c and g coordinate system is established.Through the continuous search of the grid,the global optimal parameter combination is finally obtained,and the prediction model of tooling strain based on SVM is improved.(5)Based on the above research,the overall tooling verification experiment results show that the error percentage at the out of tolerance position is less than 5%,and the overall strain error is less than 3με,which has high accuracy and strong generalization ability,and meets the actual detection requirements of complex component assembly.It verifies the effectiveness and practical reliability of the proposed aircraft assembly tooling prediction model based on support vector machine.