Development of a simple force prediction model and consistency assessment of knee movements in ten-pin bowling

The aim of this research is to use LabVIEW to help bowlers understand theirjoint movements, forces acting on their joints, and the consistency of their knee movements while competing in ten-pin bowling. Kinetic and kinematic data relating to the lower limbs were derived from bowlers’ joint angles and the joint forces were calculated from the Euler angles using the inverse dynamics method with Newton-Euler equations. An artificial-neural-network (ANN)-based data-driven model for predicting knee forces using the Euler angles was developed. This approach allows for the collection of data inbowling alleys without the use of force plates. Correlation coefficients were computed after ANN training and all values exceeded 0.9. This result implies a strong correlation between the joint angles and forces. Furthermore, the predicted 3D forces (obtained from ANN simulations) and the measured forces (obtained from force plates via the inverse dynamics method) are strongly correlated. The agreement between the predicted andmeasured forces was evaluated by the coefficient of determination (R2), which reflects the bowler’s consistency and steadiness of the bowling motion at the knee. The R2 value was beneficial in assessing the consistency of the bowling motion. An R2 value close to 1 implies a more consistent sliding motion. This research enables the prediction of the forceson the knee during ten-pin bowling by ANN simulations using the measured knee angles. Consequently, coaches and bowlers can use the developed ANN model and the analysis module to improve bowling motion.