基于预测信息集对关系的公路软基沉降组合预测模型研究

Research on Combination Prediction Model for Settlement of Highway with Soft Foundation Based on the Set Relationship Between Prediction Information

组合预测能避免单个预测方法在信息处理分析过程中的片面性,提高预测结果的精度和稳定性。基于集对分析理论,利用预测信息间的集对关系表征预测信息及预测方法的有效性,提出一种新的软基沉降组合预测模型。首先,选取合适的公路软基沉降预测模型作为基础预测模型,并采用合理指标构建集对;然后,基于集对元素的分布特征智能识别集对元素间的同异反关系,并采用联系数定量表征各预测模型间的集对联系;最后,利用联系数构建集对联系矩阵,确定信息组合所需的权重系数,实现基于信息同异反关系的软基沉降组合预测。实际软基工程的应用表明该组合预测模型能有效识别单个预测方法包含信息的有效性,并能基于此获得更为准确可靠的软基沉降预测结果。

Combined prediction method can avoid the one-sidedness of single prediction method in the process of information processing and analysis,and thus improve the accuracy and stability of prediction results.Based on the Set Pair Analysis theory,a new combined prediction model for settlement of highway with soft foundation based on the set relationship between prediction information to characterize the validity of prediction information and the prediction methods.First,several reasonable prediction models were selected as the base models,and the set pairs were constructed using a reasonable indicator.Then,the identity-discrepancy-contrary relationship criterion was intelligently identified through the distribution characteristics of the set pair elements,and the connectional number was adopted to reflect the relationship between every two prediction methods.Finally,the connectional matrix was constructed using the connectional number,and the weight coefficient required for the information combination was determined,thus the combination prediction for soft foundation settlement based on the identity-discrepancy-contrary relationship of prediction information was realized.The application of actual soft foundation engineering shows that the combined prediction model can effectively identify the validity of information contained in each prediction method,and thus more accurate and reliable prediction results for soft foundation settlement can be obtained.