基于弹性网络回归的跨座式单轨旅游交通系统车桥耦合冲击系数研究

Research on Vehicle-bridge Coupling DAF of Straddle-type MTT Based on Elastic Net Regression

  • 摘要:
    目的 跨座式MTTS(单轨旅游交通系统)属于大型游乐设施,目前该系统车桥耦合振动及冲击系数的理论研究相对较少,相关规范取值不尽合理,应建立MTTS冲击系数的合理表达式,完善行业设计规范。
    方法 首先,通过项目现场试验与动力学数值仿真方法,系统分析了MTTS的动力响应特性和冲击系数影响因素,验证了模型的合理性与精确度。其次,结合联合仿真模型揭示了多因素耦合作用对冲击系数的动态影响机理。在此基础上,综合运用Pearson相关系数法和随机森林算法筛选关键回归参数,并基于Elastic Net(弹性网络)回归理论处理影响因素间的多重共线性问题,进一步筛选回归变量,构建了多因素耦合作用下的冲击系数预测模型。最后,通过仿真数据与专业规范的交叉验证,提出了MTTS冲击系数建议计算表达式。
    结果及结论 所建预测模型能预测MTTS冲击系数92.50%的变化信息,并通过了95.00%的置信度和0.05的显著性检验,精度较高且较为合理,可为小型车辆尺寸特征的大型游乐设施类MTTS提供设计参考。

     

    Abstract:
    Objective The straddle-type MTTS (monorail tour transit system) is classified as large-scale amusement facilities. Currently, there is relatively limited theoretical research on the vehicle-bridge coupling DAF(dynamic amplification factors) of MTTS, and the values adopted in relevant standards are not entirely rational. Therefore, it is necessary to establish a reasonable expression for the MTTS DAF to improve industry design standards.
    Method First, through onsite project tests and dynamic numerical simulation methods, the dynamic response characteristics and factors influencing MTTS DAF are systematically analyzed, verifying the model's rationality and accuracy. Second, by combing a co-simulation model, the dynamic influence mechanism of multi-factor coupling effects on DAF is revealed. Building on this, key regression parameters are filtered by comprehensively using the Pearson correlation coefficient method and the random forest algorithm. Based on Elastic Net regression theory, the multi-collinearity issue among influencing factors is addressed to further screen regression variables, and a DAF prediction model under multi-factor coupling is constructed. Finally, a recommended calculation expression for the MTTS DAF is proposed through a cross-validation of the simulation data and professional standards.
    Result & Conclusion The established prediction model can predict 92.50% of the variation information for the MTTS DAF, passing the 95.00% confidence level and the 0.05 significance test. The model has high accuracy and quite rationality, thus providing a design reference for large-scale amusement facilities like MTTS with small vehicle dimensional characteristics.

     

/

返回文章
返回