Optimization-Based Parameter Identification for Coupled Biodynamic Model of Seated Posture under Vibration

We recently developed a three-direction (vertical, longitudinal, and lateral) coupled biodynamic model of seated posture under vibration. However, in that study we only tested one algorithm to identify the model parameters. This article investigates four different optimization solvers in Matlab®, i.e., particle swarm optimization (particleswarm), particle swarm and local optimization method (fmincon), genetic algorithm (ga) and local optimization method (fmincon), and local optimization method (fmincon) to identify coupled biodynamic model parameters. Based on the obtained parameters, it further compares experimental and simulation results to determine the best optimization solver in terms of the root mean square error (RMSE), linear regression (R ²), goodness of fit (ε), and Central Processing Unit (CPU) time. The results show that particle swarm optimization is the best one for identifying the biodynamic model’s parameters.