论文部分内容阅读
在大量已有湿陷性黄土地区强夯资料的基础上,采用频数统计的方法,根据变权的概念确定权重,从而构建了一个基于模糊相似优先比的湿陷性黄土强夯有效加固深度预测范例推理模型。该模型是将已有强夯实例作为源范例,将待分析实例作为目标范例,选取相应的评价指标作为模糊因子,通过源范例与目标范例之间模糊因子的相似度计算,得到目标范例与源范例之间的相似性序列,找到与强夯有效加固深度目标范例最相似的源范例,实现有效加固深度的预测。实例分析表明,有效加固深度预测值与实测值误差在10%以内,预测精度较高,具有一定的推广价值。
Based on the data of dynamic compaction in a large number of existing collapsible loess regions, the method of frequency statistics was used to determine the weight according to the concept of variable weight, so that a depth prediction based on fuzzy similarity priority ratio was proposed to effectively reinforce collapsible loess Example reasoning model. The model uses the existing dynamic compaction as the source sample, selects the sample to be analyzed as the target sample, selects the corresponding evaluation index as the fuzzy factor, calculates the similarity between the fuzzy sample and the target sample, and obtains the target sample and the source The sequence of similarities between the examples can be used to find the source sample that is most similar to the example of dynamic compaction depth target, and to predict the effective depth of reinforcement. The case study shows that the effective prediction depth is within 10% of the actual measured value, and the prediction accuracy is high, which has a certain promotion value.