Most existing Successive Cancellation List Bit-flipping(SCLF)algorithms for polar codes rely on static Critical Sets(CS)or fixed channel reliability ordering，which leads to highly redundant flipping candidate sets，inaccurate error localization，and consequently increased decoding latency and computational complexity. To address this problem， an improved SCLF decoding algorithm combining Assigned Sets(AS)and dynamic Log-Likelihood Ratio(LLR)threshold decision is proposed. Specifically，the proposed algorithm first calculates the AS using the path metrics obtained during the decoding process of the Cyclic Redundancy Check Aided Successive Cancellation List(CA-SCL)algorithm，thereby eliminating invalid flipping positions. Subsequently， a dynamic LLR threshold is introduced to perform secondary screening on the candidate flipping bits by calculating their intersection，which enables accurate error localization with instantaneous channel state information. Simulation results demonstrate that at a Block Error Rate(BLER)of 1×10.4，the proposed SCLF decoding algorithm achieves gains of 0. 24 dB，0.19 dB，0.1 dB，and 0.08 dB compared with the CA-SCL，CS-SCLF，D-Post，and AS-SCLF decoding algorithms，respectively. While maintaining significant gains in error-correction performance，the proposed algorithm effectively reduces the average number of decoding iterations in the low signal-to-noise ratio(SNR)region， achieving a favorable trade-off between decoding performance and computational complexity.