Abstract: Ladle furnace (LF) refining is a critical process in the steelmaking-continuous casting (SCC) process, where the end-point temperature directly determines the continuity of molten steel casting at the caster. Owing to the strong uncertainty inherent in the LF refining process, deviations between the actual and planned operation times of LF refining (LF refining time disturbance) frequently occur when targeting the desired end-point temperature, which hinders the smooth operation of the steelmaking-continuous casting process. To address this issue, this study first establishes an end-point molten steel temperature prediction model based on CatBoost optimized by grid search (GS-CatBoost), which is used to predict the end-point molten steel temperature under planned production parameters. Secondly, a knowledge model response to LF refining time disturbance is constructed. Based on the predicted end-point temperature, the deviation degree between actual and scheduled operation time is calculated, and corresponding countermeasures are proposed with the support of an LF refining time disturbance response knowledge graph. Furthermore, the proposed end-point temperature prediction model is trained and validatedtested using actual historical industrial data from a domestic steel plant. The results show that the hit ratio reaches 90.99% within an error range of ±5 ℃, with a Coefficient of Determination of 0.9202 and a Root Mean Squared Error of 3.1231. Based on the prediction of the end-point molten steel temperature, corresponding process operations and examples of response strategies for different levels of time disturbances are proposed. In addition, the countermeasures generated by the LF refining time disturbance response knowledge model will be further integrated with scheduling algorithms for practical industrial applications.