【摘要】 经研究发现进气歧管的翘曲变形会影响其进气性能。本文基于进气歧管的翘曲变形问题，以优化歧管的翘曲变形量为目标，提出一种随机森林智能优化算法，并设计25组正交实验样本和4组噪音样本作为随机森林回归模型的原始数据，再利用Bootstrap采样对29组原始数据进行有放回的重复随机抽取新的数据样本作为训练集，并将训练的预测值与实际值进行比较以验证回归模型的可靠性，并得到预测翘曲变形量为1.075 5 mm,经仿真验证实际翘曲变形量为1.084 mm,误差为0.8%,且与参考模型翘曲变形量相比较下降了13.6%。结果表明，随机森林回归模型在注塑成型的寻优应用中的精度是比较高的。更多还原

【Abstract】 Studies had found that the warping deformation of the intake manifold could affect its intake performance.Based on the warpage deformation problem of the intake manifold, this paper proposed a random forest intelligent optimization algorithm with the goal of optimizing the warpage deformation of the manifold, and designed 25 sets of orthogonal experimental samples and 4 sets of noise samples as the original data of the random forest regression model.Then it used Bootstrap sampling to repeatedly randomly sample 29 sets of original data as the training set, and compared the trained predicted value with the actual value to verify the reliability of the regression model.The predicted warpage deformation is 1.075 5 mm, and the actual warpage deformation is 1.084 mm with an error of 0.8%,and the warpage deformation is reduced by 13.6%comparing with the reference model.The results show that the accuracy of the random forest regression model in the optimization application of injection molding is relatively high.更多还原