针对产品设计时间预测中存在的小样本、不确定性数据和异方差噪音等问题,将模糊回归理论与par--SVM相结合,基于Necessity模型构造约束条件,提出了par-F-SVM,并给出了相应的设计活动时间智能预测方法和相关参数的优选算法.最后通过注塑模具设计的实例分析表明了所提出的基于par-F-SVM的时间预测方法是有效、可行的. Aiming at the problems of small sample, uncertainty data and heteroscedastic noise in product design time prediction, the fuzzy regressive theory is combined with par - SVM to construct the constraint conditions based on the Necessity model, and the par-F-SVM is proposed. And gives the corresponding optimization algorithm of intelligent forecasting time and relevant parameters of design activity.At last, the case analysis of injection mold design shows that the proposed time prediction method based on par-F-SVM is effective and feasible.