基于GPR与概率神经网络的城市道路路基病害预警模型研究

近年来由道路路基病害所导致的城市道路塌陷事故频繁发生，造成人员伤亡和巨大的经济损失。针对城市道路路基病害所带来的诸多问题与安全隐患，提出了一种基于探地雷达与概率神经网络的城市道路路基病害预警模型。首先在分析路基病害成因的基础上建立了城市道路路基病害评价指标体系；然后依托城市道路路基病害动态演化物理模型，使用探地雷达采集并积累样本数据；最后基于样本数据建立概率神经网络预警模型并进行仿真分析，结果表明预警准确率达到88.9%。该预警模型在郑州市中牟县的滨河路和荟萃路进行了应用试验，预警结果与实际情况基本吻合，表明该预警模型可对城市道路路基病害做出有效预警，为道路养护和工程整治设计提供技术支撑。

Research on Early Warning Model of Roadbed Diseases under Urban Roads Based on GPR and Probabilistic Neural Network

In recent years, accidents of urban road collapsing caused by roadbed diseases of urban roads occur frequently, which not only cause huge economic loss, but also the casualties. For the various problems and potential security risks brought by urban road collapsing, a new early-warning model of roadbed diseases under urban roads based on GPR and probabilistic neural network is put forward. Firstly, the evaluation index system of roadbed diseases under urban roads is established based on the analyzing on the causes of roadbed diseases; then, relying on the dynamic evolution physical model of roadbed diseases under urban roads, the ground penetrating radar is used to collect and accumulate sample data. Finally, an early warning model of probabilistic neural network is set up based on sample data and a simulated analysis is operated. The results show that the early warning accuracy rate reaches 88.9%. An application experiment of the early-warning model is operated to the Binhe road and Huicui road of the Zhongmo county and the experiment result was consistent with the actual situation. The result shows that: this early warning method can effectively predict roadbed diseases of the urban road, and can also provide effective technical support for the road maintenance and engineering treatment design.