The soft yoke mooring system is one of the main mooring methods of FPSO. The motion state of the mooring system under the combined action of wind and wave and other environmental loads shows the dynamic characteristics of high aggregation of multiple degrees of freedom and modal mixing. In this paper, in view of the traditional clustering algorithm that cannot obtain the characteristics of small clusters characterizing extreme operating conditions, a BCALoD clustering algorithm based monitoring data extraction method for soft yoke mooring systems was proposed for compression and dynamical analysis of the massive data. The clustering results show that the cluster centers of multiple data clusters correspond to the mooring attitude under normal operating conditions, but the cluster centers of relatively few data clusters contain not only the mooring attitude under extreme operating conditions but also the mooring attitude under normal operating conditions. The horizontal and vertical mooring forces were calculated for each clustering condition by extracting the time course data of each clustering center based on the original data and combining with the multibody dynamic modelling of soft yoke mooring system. The results of the force analysis based on the clustering show that the mooring forces of the clustered normal and extreme conditions are as expected, and that the extreme force states under the normal behavior of the mooring attitude are found in particular, which characterize the inadequacy of the design consideration of the soft yoke mooring system. It is shown that the clustering method in this paper has the ability to effectively extract the dangerous conditions of the soft yoke mooring system, and that it is important to pay attention not only to the forces in extreme attitudes, but also to the extreme forces generated by the coupling of attitude motion.