“…Investigations on identification of hysteretic restoring-force characteristics of structures have been conducted in the last four decades (Toussi and Yao, 1983;Cifuentes, 1984;Leontaritis and Billings, 1985;Masri et al, 1987Masri et al, , 1993Loh and Chung, 1993;Kitada, 1998Kitada, , 2000Kitada et al, 2000;Li et al, 2004a,b;Saadat et al, 2004;Zhang and Sato, 2006;Ikhouance and Rodellar, 2007;Tasbihgoo et al, 2007;Worden and Manson, 2010;Brewick et al, 2016;Pelliciari et al, 2018). Although these investigations are versatile, the problem of physical parameter system identification of 3D building structures with mass and/or stiffness eccentricity is a tough problem (e.g., Omrani et al, 2012;Nabeshima and Takewaki, 2017;Shintani et al, 2017Shintani et al, , 2019Fujita and Takewaki, 2018). Compared to the previous researches using neural networks (Saadat et al, 2004;Tasbihgoo et al, 2007;Brewick et al, 2016), wavelet transforms (Kitada, 1998(Kitada, , 2000Saadat et al, 2004), support vector regressions (Zhang and Sato, 2006), fast Bayesian bootstrap filter (Li et al, 2004b), multi-stage iterative approach (Li et al, 2004a), the present approach is suitable for vertical and horizontal framewise simultaneous identification of hysteretic restoring-force characteristics of structural frames with flexible floors within a small amount of computational load.…”