摘要： The phenomenology involved in severe accidents (SA) in nuclear reactors is highly complex. Currently, integrated analysis programs used for severe accident analysis heavily rely on custom empirical parameters, which introduce considerable uncertainty. Therefore, in recent years, the field of severe accidents has shifted its focus toward applying uncertainty analysis methods to quantify uncertainty in safety assessment programs, known as "best estimate plus uncertainty (BEPU)." This approach aids in enhancing our comprehension of these programs and their further development and improvement. This study concentrates on a third-generation pressurized water reactor (Gen-III PWR) equipped with advanced active and passive mitigation strategies. Through an Integrated Severe Accident Analysis Program (ISAA), numerical modeling and uncertainty analysis were conducted on severe accidents resulting from large break loss of coolant accidents (LBLOCA). Seventeen uncertainty parameters of the ISAA program were meticulously screened. Using Wilks' formula, the developed uncertainty program code, SAUP, was employed to carry out Latin hypercube sampling (LHS), while ISAA was employed to execute batch calculations. Statistical analysis was then conducted on two figures of merit (FOMs), namely, hydrogen generation and the release of fission products (FP) within the pressure vessel. Uncertainty calculations revealed that hydrogen production and the fraction of fission product released exhibited a normal distribution, ranging from 182.784 kg to 329.581 kg and from 15.6% to 84.3%, respectively. The ratio of hydrogen production to reactor thermal power fell within the range of 0.0578 to 0.104. A sensitivity analysis was performed for uncertain input parameters, revealing significant correlations between the failure temperature of the cladding oxide layer, maximum melt flow rate, size of the particulate debris, and porosity of the debris with both hydrogen generation and the release of fission products.