Abstract A typical single-lined spectroscopic binary (SB1) consists of two stars with almost the same age but markedly different masses and physical properties. SB1s are essential for obtaining the statistical properties of stars and constraining stellar evolutionary models, particularly when orbital parameters are well determined. In this study, we first carry out a data-cleaning step on a sample of LAMOST Medium-Resolution Spectroscopic Survey stars with a sufficient number of observations and perform an F 2 test, obtaining 7775 radial velocity variables. After removing false positives and crossmatching with variable and double-lined spectroscopic binary (SB2) catalogs, we exclude known astrophysical variables and SB2s, leaving 6106 SB1 candidates. We estimate initial periods using the Heck–Manfroid–Mersch method combined with Monte Carlo simulations and then fit each source with both an orbital model and a trend model, selecting between them via the Snedecor F- test of nested models. For orbital models, we accept the least- χ 2 solution if it satisfies F 2 10, phase _ gap _ max < 0. 3, σ v / K < 0. 25 and passes the Markov Chain Monte Carlo–based robustness tests. For trend cases, we discard zeroth-order fits and require F 2 < 3. 1. In total, we obtain 623 accepted solutions, of which 337 SB1s are new relative to existing catalogs.
叶 et al. (2026) studied this question.