Evaluation of the Optimal Saddle Position by Forward Dynamics Simulation of Pedaling

Abstract

There are many studies for bicycles and pedaling; however, most of the pedaling studies are conducted based on experiments, such as inverse dynamics method. The purpose of this study is to develop a forward dynamics model of pedaling to generate pedaling motion on computer without experimental data. The proposed model was used proportional-derivative (PD) control for joint driving torque and the referred joint angles were optimized by genetic algorithms. Cost function of optimization was defined as minimum of the muscle load and differences between the objective crank angular velocity and that of the simulation. Joint torques and pedal forces was obtained from the simulation and was compared with the actual experimental data. Simulation results were tended to vibrate compared with the actual experimental data. In addition, magnitude of the cost function was investigated when changing saddle height as 0.700, 0.725, 0.750, 0.775 and 0.800 [m]. As a result, the cost function decreased as the saddle height became higher, and the cos function was minimum when the saddle height was 0.775[m].