时间序列预测方及在隧道控制中的应用研究.pdf

摘 要 摘 要 当前金融危机下，随着国家基础设施投入的加大，公路路网建设进入了高峰 时期，投入运营的公路隧道不断增加。如何在保证隧道内安全行驶的条件下最大 限度的减少能耗，已经成为目前公路交通领域面临的严峻问题。为此，本文提出 了一种基于时间序列预测的隧道通风照明控制方案。 首先，根据数据预处理方法，针对隧道数据的特点，总结出了一系列适合隧 道数据的预处理方法。比较分析了灰色预测法，神经网络预测法，针对这些算法 的不足，改进了其中的灰色预测法，并提出了一种适合隧道传感器数据预测的小 波-灰色-神经网络预测法。 其次，通过讨论 PID 控制技术和模糊控制原理，建立了模糊 PID 控制器，仿 真证明了该控制器的优越性。 然后，分析了现有通风控制系统存在的不足，利用时间序列预测方法、模糊 控制理论、PID 控制技术，提出了预先控制的、自学习的隧道通风控制系统方案， 实现了隧道通风控制预警系统。对比分析表明方案取得了明显的节能效果。 最后，介绍了公路隧道照明的控制、布置及灯具选择等常规方式，根据隧道 内驾驶员理想亮度适应曲线，运用隧道照明亮度及长度的考虑因素及计算公式， 利用目标非线性规划方法建立了隧道照明节能控制方案。仿真及节能分析表明本 文提出的方案能很好的实现隧道照明节能的目的。 关键词：时间序列预测 公路隧道 通风 照明 智能模糊控制 ABSTRACT ABSTRACT In the period of current financial crisis,with the state increasing the investment of infrastructure,the establishment of the road system is coming to the peak period,while more and more road tunnels are put into operation.How to save energy as much as possible while making sure the travel is safe has become a serious problem in the field of tunnel traffic.New methods of road tunnel ventilation and lighting technologies based on time series forecasting is proposed in this paper. Firstly, the data collected by the tunnel sensors were disposed of by using integrated methods of data pre-processing after studying the characteristics of the data.The grey prediction method and the neural network forecasting method were comparied and analyzed.The grey forecasting method was improved by knowing the lack of its algorithms.At the same time,a new wavelet-gray-neural network forecasting method which is suitable for the data collected by the tunnel sensors was proposed. Secondly, PID control technology and fuzzy control theory were