Abstract: Thirty-nine convection parameters are calculated by using daily routine upper-air sounding data from the Xingtai and Zhangqiu aerological observation stations and hourly observation data from the Liaocheng Airport from June to August during 2011-2015. Based on the significance test of the correlation coefficient between convection parameters and strong convective weather samples, as well as the analysis of numerical distribution of convection parameters in the strong convective weather samples and the non-strong convective weather samples, 12 convection parameters at Zhangqiu station and 8 convection parameters at Xingtai station are finally selected by the methods of boxplot and skill scores. Through analyzing the indication to the strong convective weather events of different types of convection parameters, and according to the different thresholds of convection parameters in each month, different criteria for the strong convective weather are determined respectively. Thus, the strong convective weather forecast model for summer short-time heavy rains in Liaocheng is established by the membership function conversion method. The fitting test and forecasting test of the forecast model are verified with a good result in which the accuracy rate is over 60%.