Complex Surface Detection System Based On FBG Sensor Network

Fiber Bragg Grating(FBG)has many advantages,such as small size,strong anti-interference ability,high sensitivity,and easy networking.It has been widely used in construction,transportation,power.In the process of high-precision surface measurement and assembly of large-sized products such as aircrafts and projectiles,deformation of large-sized workpieces due to factors including their own weight and external stress can cause large errors in measurements.Without offset compensation,the reference measurement is inaccurate,which can affect the assembly quality.To address the quality issues encountered in the assembly process of complex surfaces,in this study,FBG sensor structure design,construction of a strain detection system based on FBG sensor network,and development of an FBG demodulator host computer software were investigated,focusing on the characteristics of large-sized complex surface structures.(1)An optical fiber sensing system was built,and the sensing unit and demodulation module were selected based on the strain distribution characteristics and measurement requirements of large-sized workpieces.For application to strain field detection of complex surface structures,an FBG structure was designed based on the coupled mode theory,and an FBG sensor packaging method was developed according to the characteristics of the complex curved surface structure.(2)A mathematical model for strain-space offset compensation was proposed.The relationship between the surface space offset and the strain detected by FBG sensor array,when a complex surface is deformed due to external factors,was established.The space offset was compensated during the assembly process to improve the assembly accuracy.Finally,a deep learning algorithm was studied to optimize the strain-space offset model.(3)A three-dimensional strain field detection system was constructed.The optimal configuration method for FBG sensor was studied and the strain fields of the plane plate and shell structure and aircraft plate and shell structure were monitored.The strain deviation was less than 10 ?? compared with ANSYS simulations.Based on extensive experimental data combined with visual detection technology,the detection accuracy of the vision detection system was determined as 0.05 mm,which verifies the detection accuracy and verifies the relationship between strain and offset.(4)The host computer software of the demodulator was independently developed using Visual Basic.The data was acquired via Modbus communication protocol,and post-processing was performed on the obtained data.Using the demodulated strain information and the polynomial fitting algorithm,the three-dimensional strain field was reconstructed and incorporated in the software to improve the human-computer interaction ability.The simulation calculation and experimental analysis show that the system can detect the three-dimensional strain field of complex surface structures,and the detection results can be beneficial in digital assembly.This system is suitable for strain detection of large-scale complex surfaces such as aircrafts and rail trains.It can play a significant role in the compensation of reference position,analysis of deformation,and prediction of change trends.