Measurement of Sports Balls Based on Optimization of Multivariable Linear Partial Least Square Regression Model

Abstract

Aerodynamic lift on sports balls is typically measured in wind tunnels. Wind tunnel measurements may have measurable differences with ball drag occurring in play. Measurements under game conditions have been attempted, but are difficult to interpret from the data scatter and are not controlled. The following considers lift measurements from a ball propelled through static air in a laboratory setting. High speed light gates were used to measure lift. Lift was observed to depend on the ball speed, roughness, stitch height, and orientation.This article explores a non-linear partial least square (NLPLS) regression method for wind tunnel measurements of sports balls based on Landsat Thematic Mapper (TM) data. Two schemes, leave-one-out (LOO) cross validation (scheme 1) and split sample validation (scheme 2), are used to build models. For each scheme, the NLPLS model is compared to a linear partial least square (LPLS) regression model and multivariant linear model based on ordinary least square (LOLS). This research indicates that an optimized NLPLS regression mode can substantially improve the estimation accuracy of wind tunnel measurements of sports balls.