Abstract: Directed against the deviation in the process of fiber weighing,the structure and the working principle of the fiber feeding machine were analyzed,and the nonlinear correction mathematical model was established by getting the fiber weighing data based on LabVIEW and using the regression analysis and least square method. The fitting coefficients were sent to the programmable logic controller (PLC) weighing control module by LabVIEW,and the real-time automatic correction of the weighing process was realized. The correction weighing system was verified based on static load test and the stone mastic asphalt(SMA) mixture pavement performance test. Results show that after using the control model,the fiber weight deviation fells to 1.4% from 7.3% before correction and reaches the plus or minus 2% error design requirements. The volume of air voids(VV) of SMA mixture is reduced by 17%,the rut deformation is reduced,the dynamic stability increased by 4.4%,and the flexural strength and flexural failure stiffness modulus increased by 7% and 10% respectively after correction. Other pavement performance is improved than those before correction and is the same as the performance of the laboratory test under the optimum fiber content. It provides certain help for avoiding the phenomenon with less or more fiber stabilizer in the SMA mixture and improving the SMA mixture quality.