Power system overhead line parameters are highly influenced by complex environmental factors such as temperature, rain and wind. Traditional calculations of available transmission capacity often utilize static transmission line parameters and fixed current-carrying transmission limits, which ignores dynamic changes in transmission line parameters and current-carrying capacity under environmental influences. To address this problem, Mutual Information (MI) quantification methods are used to analyze the extent of the influence of environmental factors on transmission line parameters under different spatial and temporal regions. Specifically, the quantization matrix approach is used to characterize the environment for dynamic correction of transmission parameters. Environmental features are used as inputs for parallel prediction of transmission line parameters using Long-short-term Memory (LSTM) neural network. In order to validate the feasibility and effectiveness of the methodology, the IEEE 14-node system has been selected to perform the transmission parameter calculations under different environmental influences to support the analysis of the proposed model in practical applications.