The top of reflection spectrum is osci llatory,side lobes appear on both sides of reflection spectrum,linearity of dela y curve is not good,and dispersion curve is not smooth for linear chirped fiber grating.In order to solve the above problems,piecewise apodized optimization was adopted,and effecti ve apodization functions and optimal piecewise apodization ratio are explored on the basis of characteristic analysis of linear chirped fiber grating.The characterization and evaluation of delay and reflection spectrum are presented.From the comprehensive index,an adaptive genetic algorithm (IGA) is explored from the a spects of genetic operator and algorithm used for the reconstruction of fiber grating parameters.Experimental results show that the linear region of delay curve increases,the dispersion curve has been more stable,and compress ion level of bandwidth effectively reduces after piecewise optimization.The IGA reconstruction met hod is very effective,and has higher precision.Errors of the length,the modulation depth and the chirp coeffic ient of fiber grating respectively are 0.04%,0.532% and 0.31% after IGA reconstruction.Delay and reflection spectru m are all both consistent with the actual results.In addition,this IGA reconstruction method can also meet the requirements of many other practical applications.