Research on correlation analysis and multivariate nonlinear regression modeling method for the tensile properties of mugwort petiole

The defoliation quality of mugwort defoliation equipment is an important factor to measure the defoliation efficiency, and the tensile properties of mugwort petiole will have an impact on the defoliation quality, such as the crushing rate and the abscission rate. In order to reduce the crushing rate and improve the abscission rate during mechanical harvesting of mugwort leaves, the tensile properties of mugwort petiole need to be studied. The tensile properties of mugwort petiole are closely related to its macroscopic and microscopic physicochemical parameters. Therefore, this paper takes Qichun mugwort as the research object, carries out the research on correlation analysis and multivariate nonlinear regression modeling of the tensile properties of mugwort petiole. First, the macroscopic physical parameters of mugwort, the microstructural parameters and the chemical components of mugwort were tested, and the tensile force of mugwort petiole was measured. Then, through the grey correlation and the Pearson correlation analysis, the weight values were determined with tensile properties of the mugwort petiole and each influential factors, and the multivariate nonlinear regression model of the tensile properties of mugwort petiole and its influencing factors was established by the weight values of each influencing factors. Finally, the regression model was verified. The results showed that the model had a goodness of fit of 0.725 to the experimental data, with an average absolute percentage error of 1.9%, a root mean square error of 16.3% and an average absolute error of 13.4%. The accuracy of the tensile properties model is high, which can provide data reference and theoretical basis for the design of mugwort defoliation equipment.